Competition, Innovation, and Antitrust. A Theory of Market ... - Intertic
Competition, Innovation, and Antitrust. A Theory of Market ... - Intertic
Competition, Innovation, and Antitrust. A Theory of Market ... - Intertic
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Federico Etro<br />
COMPETITION,<br />
INNOVATION, AND<br />
ANTITRUST<br />
July 19, 2007<br />
Springer-Verlag<br />
Berlin Heidelberg NewYork<br />
London Paris Tokyo<br />
Hong Kong Barcelona<br />
Budapest
<strong>Competition</strong>, <strong>Innovation</strong>, <strong>and</strong> <strong>Antitrust</strong><br />
by Federico Etro
A Francesca, Riccardo e Leonardo
Preface<br />
In 1934 Springer published a book by Heinrich von Stackelberg, “<strong>Market</strong> <strong>and</strong><br />
Equilibrium”, which contained pathbreaking studies on oligopolistic markets.<br />
In particular, it analyzed the behavior <strong>of</strong> a firm acting as a leader with a first<br />
mover advantage in the choice <strong>of</strong> its production level over another firm acting<br />
as a follower. That analysis became the foundation <strong>of</strong> the economic theory <strong>of</strong><br />
market leaders <strong>and</strong> is the starting point <strong>of</strong> my book. In the following pages<br />
I develop a generalization <strong>of</strong> Stackelberg’s idea, with a focus on the underst<strong>and</strong>ing<br />
<strong>of</strong> the behavior <strong>of</strong> market leaders under different entry conditions,<br />
particularly when entry in the market is endogenous. Rather than limiting<br />
the analysis to the effects <strong>of</strong> the market structure on the behavior <strong>of</strong> the<br />
market leaders, I also study the effects <strong>of</strong> the behavior <strong>of</strong> market leaders on<br />
the market structure.<br />
In other words, this book can be seen as an attempt to describe endogenous<br />
market structures where the strategies, the expectations on the<br />
strategies <strong>of</strong> the others, <strong>and</strong> also the entry decisions are the fruit <strong>of</strong> rational<br />
behavior. In the last few decades, economic theory has put a lot <strong>of</strong> emphasis<br />
on the rational behavior in the choice <strong>of</strong> actions <strong>and</strong> strategies <strong>and</strong> on the<br />
rational expectations on these choices. Most fields <strong>of</strong> economic theory have<br />
embraced both these elements adopting the rational expectations approach<br />
in models with perfect competition first <strong>and</strong> imperfect competition later.<br />
The theory <strong>of</strong> industrial organization has embraced these elements with the<br />
adoption <strong>of</strong> game theory as the st<strong>and</strong>ard tool <strong>of</strong> analysis <strong>of</strong> the interactions<br />
between firms. Meanwhile, economists have <strong>of</strong>ten neglected the rational behavior<br />
<strong>of</strong> the firms in their entry decisions, both in partial equilibrium <strong>and</strong><br />
general equilibrium models. For this reason, microeconomic <strong>and</strong> macroeconomic<br />
analyses <strong>of</strong> markets with imperfect competition have been <strong>of</strong>ten limited<br />
to situations in which the number <strong>of</strong> firms was exogenously given. The main<br />
scope <strong>of</strong> this book is to provide a general microeconomic analysis <strong>of</strong> markets<br />
where entry decisions are rational decisions, <strong>and</strong> to underst<strong>and</strong> the effects <strong>of</strong><br />
endogenous entry on the equilibrium behavior <strong>of</strong> the firms <strong>and</strong> on the welfare<br />
properties <strong>of</strong> the equilibrium market structure.<br />
A great deal <strong>of</strong> this work is inspired by <strong>and</strong> based on the revolutionary<br />
contributions <strong>of</strong> game theoretic analysis to industrial economics <strong>and</strong> antitrust<br />
policy in the last three decades. The pathbreaking works <strong>of</strong> Avinash Dixit,
viii<br />
Preface<br />
Michael Spence, Joseph Stiglitz, Paul Milgrom, John Roberts, Drew Fudenberg,<br />
Jean Tirole, Michael Whinston <strong>and</strong> others during the 80s made it clear<br />
how one could study the rational behavior <strong>of</strong> market leaders <strong>and</strong> draw welfare<br />
implications in a solid game theoretic framework. On the policy front,<br />
the main consequence <strong>of</strong> these studies was the development <strong>of</strong> the so-called<br />
post-Chicago approach to antitrust, which emphasized a number <strong>of</strong> situations<br />
in which an incumbent could engage in anti-competitive practices such<br />
as predatory pricing, bundling, vertical restraints, price discrimination, anticompetitive<br />
mergers <strong>and</strong> so on. The game theoretic approach was able to<br />
emphasize that these practices could harm consumers by excluding other entrants<br />
or by facilitating collusion. This approach challenged the former school<br />
<strong>of</strong> thought associated with the Chicago school <strong>and</strong> represented by Richard<br />
Posner, Robert Bork <strong>and</strong> others who were (<strong>and</strong> still are) skeptical toward<br />
antitrust intervention against exclusionary strategies <strong>and</strong> mergers.<br />
The classic book by Jean Tirole “The <strong>Theory</strong> <strong>of</strong> Industrial Organization”<br />
(1988) today remains the best exposition <strong>of</strong> the game theoretic foundations<br />
<strong>of</strong> the modern industrial organization, <strong>of</strong> the strategic interactions between<br />
firms, <strong>and</strong> <strong>of</strong> the policy implications <strong>of</strong> the post-Chicago approach. In the<br />
Introduction to the second part <strong>of</strong> that book, entitled “Strategic Interaction”<br />
<strong>and</strong> entirely dedicated to the strategic behavior <strong>of</strong> firms, Tirole points out a<br />
fundamental distinction for the behavior <strong>of</strong> a market leader facing an entrant:<br />
this leader will be aggressive under strategic substitutability <strong>and</strong> accommodating<br />
under strategic complementarity, 1 unless it tries to foreclose entry.<br />
Since competition in quantities is associated with strategic substitutability<br />
<strong>and</strong> competition in prices with strategic complementarity, Tirole’s taxonomy<br />
<strong>of</strong> business strategies based on this distinction became a classic result <strong>of</strong> the<br />
modern industrial organization <strong>and</strong> affected most <strong>of</strong> its subsequent evolution.<br />
The natural consequence for markets where firms compete in prices is indeed<br />
a simple one: incumbents adopting aggressive pricing strategies or equivalent<br />
strategies must have a predatory intent, otherwise they would adopt accommodating<br />
strategies. Since then, most <strong>of</strong> the antitrust analysis <strong>of</strong> exclusionary<br />
practices was based on related arguments.<br />
This book develops a general characterization <strong>of</strong> the strategic interactions<br />
between firms taking into account alternative entry conditions. The<br />
traditional analysis <strong>of</strong> incumbents <strong>and</strong> entrants that I sketched above has a<br />
main problem: it largely neglects the role <strong>of</strong> the endogenous entry <strong>of</strong> competitors<br />
in constraining the behavior <strong>of</strong> the incumbents. Entry in a market<br />
is endogenous when in equilibrium there are no pr<strong>of</strong>itable opportunities to<br />
1 The strategic variables <strong>of</strong> two firms interacting in a market are defined strategic<br />
substitutes when an increase in the variable chosen by one firm induces the<br />
other firm to adjust its own strategic variable in the opposite direction. They are<br />
strategic complements when an increase in the strategy <strong>of</strong> one firm induces the<br />
other one to adjust its own strategy in the same direction. The terminology is<br />
due to Bulow et al. (1985).
Preface<br />
ix<br />
be exploited by potential entrants. A simple situation in which this occurs is<br />
when entry is simply free. A more general situation emerges when firms or<br />
entrepreneurs are active in different markets <strong>and</strong> the rate <strong>of</strong> pr<strong>of</strong>it mustbe<br />
equalized across these markets. Another <strong>and</strong> more realistic situation in which<br />
entry can be regarded as endogenous is when there are large fixed costs <strong>of</strong><br />
entry or limited sunk costs (traditionally considered barriers to entry) that<br />
constrain endogenously the entry decision <strong>of</strong> the firms. Overall, we do believe<br />
that endogenous entry should be regarded as the st<strong>and</strong>ard situation<br />
in most markets, while exogenous entry only emerges in extreme situations<br />
where entry is not a decision taken by the firms, but it is determined by other<br />
institutional or regulatory authorities.<br />
When entry is endogenous market leaders are always aggressive under<br />
both strategic substitutability <strong>and</strong> complementarity, under both competition<br />
in quantities <strong>and</strong> in prices, <strong>and</strong> even under other forms <strong>of</strong> competition. This<br />
has radical implications for the pricing strategies, for the choice <strong>of</strong> strategic<br />
investments in cost reductions, quality improvements <strong>and</strong> advertising, for<br />
the choice <strong>of</strong> the financial structure, for the decisions to bundle goods or<br />
price discriminate, for the production decisions in the presence <strong>of</strong> network<br />
externalities, two-sided markets <strong>and</strong> learning by doing, for the adoption <strong>of</strong><br />
vertical restraints, for the decision to merger or collude with a rival, <strong>and</strong> for<br />
many other important issues in industrial organization.<br />
Evidently, the endogenous entry approach has crucial consequences on<br />
concrete antitrust policy for the analysis <strong>of</strong> the behavior <strong>of</strong> market leaders<br />
<strong>and</strong> also for merger <strong>and</strong> collusion issues. When entry is endogenous, incumbents<br />
are always aggressive, typically without exclusionary purposes, <strong>and</strong><br />
their strategies hardly harm consumers; mergers in markets where entry is<br />
endogenous take place if <strong>and</strong> only if they create enough cost efficiencies; <strong>and</strong><br />
cartels between a limited number <strong>of</strong> firms facing endogenous entry are ineffective.<br />
The flavor <strong>of</strong> these results goes back to the Chicago view, but our<br />
game theoretic analysis is derived from the st<strong>and</strong>ard post-Chicago approach,<br />
which is augmented with endogenous entry.<br />
The literature on industrial organization is quite fragmented because separate<br />
analysis is usually undertaken for models <strong>of</strong> competition in quantities,<br />
models <strong>of</strong> competition in prices <strong>and</strong> models <strong>of</strong> competition for the market.<br />
A possible advantage <strong>of</strong> the approach I adopt in this book is the provision<br />
<strong>of</strong> a unified framework for the analysis <strong>of</strong> market structures. This framework<br />
encompasses most models <strong>of</strong> competition in quantities, prices <strong>and</strong> models <strong>of</strong><br />
competition for the market, <strong>and</strong> can be used to analyze <strong>and</strong> compare different<br />
market structures in a simpler manner. The book contains a large amount<br />
<strong>of</strong> unpublished material, especially in the theoretical analysis <strong>of</strong> Chapters 2<br />
to 4. The applied analysis in Chapters 5 to 7 is based on policy oriented<br />
work, some <strong>of</strong> which was realized as the chief economist <strong>of</strong> the Task Force<br />
on <strong>Competition</strong> established by the International Chamber <strong>of</strong> Commerce <strong>of</strong><br />
Paris in 2006.
x<br />
Preface<br />
In Chapter 1, I introduce the basic theoretical tools <strong>of</strong> industrial organization<br />
<strong>and</strong> describe the simplest examples <strong>of</strong> competition <strong>and</strong> innovation.<br />
Thestartingpointisamarketinwhichafirm decides how much to produce<br />
on the basis <strong>of</strong> dem<strong>and</strong> <strong>and</strong> cost conditions. In such a context, I describe<br />
the behavior <strong>of</strong> a monopolist <strong>and</strong> compare it with the behavior <strong>of</strong> two firms<br />
in a Cournot duopoly; on this basis I introduce the discussion <strong>of</strong> the fundamental<br />
subjects <strong>of</strong> antitrust analysis as mergers, foreclosure <strong>and</strong> collusion.<br />
Then, I employ the same model to describe competition between multiple<br />
firms within the four main market structures analyzed in this book. In the<br />
first (Nash competition), firms take decisions independently <strong>and</strong> their number<br />
is exogenous. In the second (Marshall competition), the number <strong>of</strong> firms<br />
is endogenized assuming that firms enter in the market if <strong>and</strong> only if they<br />
expect positive pr<strong>of</strong>its. In the third (Stackelberg competition), there is again<br />
an exogenous number <strong>of</strong> firms but one <strong>of</strong> them, the leader, takes its decision<br />
before the others. In the fourth (Stackelberg competition with endogenous<br />
entry), there is still a leader with a first mover advantage, but the number<br />
<strong>of</strong> firms is endogenous <strong>and</strong> again derived assuming that firms enter in the<br />
market if <strong>and</strong> only if they expect positive pr<strong>of</strong>its. The same analysis can<br />
be extended to a model where firms sell differentiated products <strong>and</strong> choose<br />
their prices. I analyze such a model adopting the simplest dem<strong>and</strong> <strong>and</strong> cost<br />
conditions, <strong>and</strong> characterizing the same four different forms <strong>of</strong> competition<br />
as before: with price competition, however, I show that the behavior <strong>of</strong> the<br />
leader is radically different according to whether entry is endogenous or not.<br />
Finally, I provide a simple example <strong>of</strong> competition for the market where firms<br />
invest to increase their relative chances to innovate, I analyze the four different<br />
equilibria, <strong>and</strong> apply the result to discuss the incentives <strong>of</strong> an incumbent<br />
monopolist to invest in R&D.<br />
In Chapter 2, I present a general model <strong>of</strong> competition <strong>and</strong> I show that<br />
most models used in industrial organization are nested in this general model.<br />
Applications include virtually all symmetric models <strong>of</strong> competition in quantities<br />
with homogenous <strong>and</strong> differentiated goods, models <strong>of</strong> price competition<br />
with Logit or isoelastic dem<strong>and</strong>, <strong>and</strong> st<strong>and</strong>ard contests or patent races. I discuss<br />
in some detail how to characterize the Nash equilibrium <strong>and</strong> the Marshall<br />
equilibrium for the general model <strong>and</strong> for its main applications. Then, I extend<br />
these equilibria with a firm, the leader, which undertakes a preliminary<br />
investment affecting competition ex post, as in the literature on strategic investments<br />
started with the contributions <strong>of</strong> Avinash Dixit <strong>and</strong> others. This<br />
general approach allows one to verify what the strategic incentives <strong>of</strong> the<br />
leader are to engage in a number <strong>of</strong> commitments or investments <strong>and</strong> to<br />
be aggressive or accommodating in the market. 2 I perform this analysis in<br />
the presence <strong>of</strong> an exogenous number <strong>of</strong> competitors <strong>and</strong> <strong>of</strong> an endogenous<br />
number, <strong>and</strong> derive the general principle for which market leaders facing en-<br />
2 A more aggressive strategy reduces the pr<strong>of</strong>its <strong>of</strong> the other firms, a more accommodating<br />
strategy increases them.
Preface<br />
xi<br />
dogenous entry always take those strategic decisions that induce them to be<br />
aggressive in the market. Then, I apply these results to specific decisions <strong>of</strong> a<br />
leader: 1) investments in cost reductions; 2) persuasive advertising (<strong>and</strong> other<br />
dem<strong>and</strong> enhancing investments); 3) decisions on the financial structure <strong>and</strong><br />
the optimal equity-debt ratio; 4) preliminary production levels in the presence<br />
<strong>of</strong> network externalities <strong>and</strong> two-sided markets; 5) bundling <strong>of</strong> goods; 6) price<br />
discrimination; 7) delegation <strong>of</strong> pricing decisions to downstream distributors<br />
for interbr<strong>and</strong> competition; <strong>and</strong> 8) horizontal mergers.<br />
In Chapter 3, I generalize the analysis <strong>of</strong> the forms <strong>of</strong> competition in<br />
which a leader has a first mover advantage <strong>and</strong> followers decide their strategies<br />
independently in a subsequent stage. I characterize the Stackelberg equilibrium<br />
<strong>and</strong> the Stackelberg equilibrium with endogenous entry within the<br />
general framework <strong>and</strong> for alternative forms <strong>of</strong> competition in quantities <strong>and</strong><br />
in prices. In particular, I derive the general principle for which market leaders<br />
facing endogenous entry are always aggressive under both strategic complementarity<br />
<strong>and</strong> strategic substitutability: they produce more than the rivals<br />
when competing in quantities <strong>and</strong> they set lower prices when competing in<br />
prices. I also derive the conditions under which a market leader is so aggressive<br />
to adopt an entry-deterring strategy. This happens under constant<br />
or decreasing marginal costs <strong>of</strong> production <strong>and</strong> homogenous goods, independently<br />
from the size <strong>of</strong> the fixed costs <strong>of</strong> production <strong>and</strong> <strong>of</strong> the shape <strong>of</strong> the<br />
dem<strong>and</strong> function, <strong>and</strong> it provides a game theoretic foundation for some <strong>of</strong> the<br />
insights <strong>of</strong> the limit-pricing framework associated with Joe Bain, Paolo Sylos<br />
Labini <strong>and</strong> Franco Modigliani <strong>and</strong> <strong>of</strong> the contestability approach associated<br />
with William Baumol, John Panzar <strong>and</strong> Robert Willig. The latter approach<br />
could be re-interpreted in terms <strong>of</strong> Stackelberg competition in prices with<br />
endogenous entry <strong>and</strong> homogenous goods, but our framework allows us to<br />
extend its spirit to the more general case <strong>of</strong> product differentiation. In such a<br />
case (as when marginal costs are increasing), market leaders prefer to allow<br />
entry while still adopting aggressive strategies under both quantity <strong>and</strong> price<br />
competition. Finally, I show that, when entry is endogenous, the allocation<br />
<strong>of</strong> resources is improved by the presence <strong>of</strong> the leader. The spirit <strong>of</strong> these<br />
results extends to the more complex cases with asymmetries between firms,<br />
multiple leaders or endogenous leadership, <strong>and</strong> to the case <strong>of</strong> multiple strategic<br />
variables. In conclusion, I illustrate how one can apply these results to<br />
differentpolicyquestions:1)Ireconsider the role <strong>of</strong> a collusive cartel in the<br />
presence <strong>of</strong> endogenous entry, <strong>and</strong> argue that this is ineffective unless it has<br />
a leadership role (in which case the cartel coordinates aggressive strategies<br />
between its members); 2) I review the problem <strong>of</strong> the optimal state aids <strong>and</strong><br />
trade policy for firms exporting in a foreign country, <strong>and</strong> I show that the<br />
traditional results break down when the domestic firms are engaged in competition<br />
in a market where entry is endogenous (in such a realistic case, state<br />
aids inducing aggressive export strategies, <strong>and</strong> in particular export subsidies,
xii<br />
Preface<br />
are always optimal); <strong>and</strong> 3) I analyze the role <strong>of</strong> privatizations <strong>and</strong> liberalizations<br />
in markets for private goods.<br />
In Chapter 4, I exploit the results <strong>of</strong> the previous chapters <strong>and</strong> apply<br />
them specifically to models <strong>of</strong> competition for the market. My starting point<br />
is a simple model in which all firms choose an initial investment that delivers<br />
drastic innovations according to a stochastic process. Analyzing the usual<br />
four forms <strong>of</strong> competition, I show the general principle for which incumbent<br />
monopolists that are leaders <strong>and</strong> face endogenous entry in the competition<br />
for the market, invest in R&D more than any other firm. This outcome overturns<br />
a st<strong>and</strong>ard result <strong>of</strong> the theory <strong>of</strong> innovation, due to Kenneth Arrow, for<br />
which incumbent monopolists would have lower incentives to invest in R&D<br />
<strong>and</strong> replace their own technological leadership. The same result on innovation<br />
byleadersisconfirmed in a more realistic version <strong>of</strong> the model in which firms<br />
invest over time, when innovations are non drastic, <strong>and</strong> especially when they<br />
are sequential. The investment <strong>of</strong> the technological leaders in the presence <strong>of</strong><br />
sequential innovations leads automatically to an explanation for the persistence<br />
<strong>of</strong> monopolistic positions, which is associated (somewhat paradoxically)<br />
with free entry in the competition for the market. On this basis, I develop<br />
a theory <strong>of</strong> technological progress driven by market leaders which is closely<br />
related to the original ideas <strong>of</strong> Joseph Schumpeter on the role <strong>of</strong> monopolies<br />
in enhancing growth - ideas that are hardly consistent with the recent literature<br />
on endogenous technological progress, in which leaders do not invest in<br />
R&D because <strong>of</strong> the Arrow effect. Finally, I discuss the relationship between<br />
competition in the market on one side <strong>and</strong> competition for the market on the<br />
other side.<br />
In Chapter 5, I apply my theoretical analysis to antitrust policy, particularly<br />
to issues concerning abuse <strong>of</strong> dominance. First, I review the traditional<br />
approaches to antitrust policy <strong>and</strong> emphasize the strengths <strong>and</strong> the limits <strong>of</strong><br />
the Chicago school <strong>and</strong> <strong>of</strong> the post-Chicago approach. Subsequently, I provide<br />
a first attempt to derive policy implications from the theoretical analysis<br />
on the behavior <strong>of</strong> market leaders in the presence <strong>of</strong> exogenous entry <strong>and</strong> endogenous<br />
entry. I emphasize that any inference on the market power <strong>of</strong> a<br />
leader from its market share can be highly misleading. Moreover, when entry<br />
<strong>of</strong> firms is endogenous, one should be extremely careful in associating aggressive<br />
pricing strategies by market leaders (or related strategies as bundling)<br />
with exclusionary purposes. I also note that when firms compete to obtain<br />
sequential innovations protected by intellectual property rights (IPRs), persistence<br />
<strong>of</strong> technological leadership can derive from endogenous entry in the<br />
competition for the market rather than market power in the competition in<br />
the market. Therefore, antitrust policy should be careful in evaluating dominant<br />
positions in dynamic high-tech sectors, <strong>and</strong> should avoid interfering<br />
with the protection <strong>of</strong> IPRs which is the source <strong>of</strong> investments in R&D <strong>and</strong><br />
technological progress. In conclusion, I apply these ideas to current antitrust<br />
policy with particular reference to the efficiency defense for dominant firms,
Preface<br />
xiii<br />
to the determination or predatory pricing, to bundling as an exclusionary<br />
strategy, <strong>and</strong> to issues <strong>of</strong> IPRs protection.<br />
In Chapter 6, I apply my theoretical analysis to the markets <strong>of</strong> the New<br />
Economy, in particular to the s<strong>of</strong>tware sector, which is characterized by a<br />
number <strong>of</strong> peculiar features analyzed in the book as network externalities,<br />
two-sided markets, high investments in R&D <strong>and</strong> a pre-eminent role <strong>of</strong> the<br />
leader in both the competition in the market <strong>and</strong> for the market. This leader<br />
has been also the subject <strong>of</strong> antitrust investigations in US <strong>and</strong> EU, therefore<br />
I analyze these famous antitrust cases from an economic point <strong>of</strong> view, <strong>and</strong><br />
try to focus on its main aspects: 1) whether Micros<strong>of</strong>t is a monopolist; 2)<br />
whether its bundling strategies are predatory <strong>and</strong> harm consumers; <strong>and</strong> 3)<br />
whether antitrust authorities should force the disclosure <strong>of</strong> its IPRs to promote<br />
competition in the s<strong>of</strong>tware market. I can briefly summarize the results<br />
<strong>of</strong> my investigation as follows: 1) evidence from the competition in <strong>and</strong> for<br />
s<strong>of</strong>tware markets witnesses the lack <strong>of</strong> monopolistic power by Micros<strong>of</strong>t <strong>and</strong><br />
better defines its role as that <strong>of</strong> a Stackelberg leader in a market with endogenous<br />
entry; 2) bundling strategies by Micros<strong>of</strong>t appear as natural aggressive,<br />
or pro-competitive, strategies which may harm competitors but create benefits<br />
to all consumers; <strong>and</strong> 3) forced disclosure <strong>of</strong> the IPRs <strong>of</strong> Micros<strong>of</strong>t for<br />
interoperability purposes may severely jeopardize investment in R&D rather<br />
than promoting it, with negative consequences for the consumers in the long<br />
run.<br />
In Chapter 7, I conclude this book by suggesting ways to investigate the<br />
empirical predictions <strong>of</strong> the theory <strong>of</strong> market leaders concerning the pricing<br />
policy <strong>of</strong> the leaders, <strong>and</strong> their decisions on quality, advertising, distribution,<br />
financing <strong>and</strong> R&D investments as functions <strong>of</strong> the entry conditions. I also<br />
re-interpret my results on the behavior <strong>of</strong> market leaders from the point <strong>of</strong><br />
view <strong>of</strong> business administration recommendations for marketing <strong>and</strong> strategy.<br />
Finally, I suggest avenues for future theoretical research on market leadership<br />
<strong>and</strong> on endogenous entry.<br />
My initial interest in the role <strong>of</strong> market leaders <strong>and</strong> endogenous entry,<br />
especially in the market for innovations, was inspired by discussions with<br />
Michele Boldrin at U.C.L.A. While our later research efforts have taken radically<br />
different directions, I am grateful to him for inspiring motivations.<br />
At U.C.L.A., between 1998 <strong>and</strong> 2000, I also benefited from interaction with<br />
Harold Demsetz, Jack Hirshleifer, David Levine, John Riley, Bill Zame <strong>and</strong>,<br />
most <strong>of</strong> all, with Karina Firme whose wisdom <strong>and</strong> intelligence has enlightened<br />
many <strong>of</strong> my thoughts on these issues (<strong>and</strong> others as well). I presented<br />
a prototype model on the behavior <strong>of</strong> leaders in markets with endogenous<br />
entry for the first time in a seminar at M.I.T. in November 2000. In that<br />
occasion, comments by Robert Barro <strong>and</strong> Daron Acemoglu shaped a lot <strong>of</strong><br />
my subsequent theoretical investigations. I developed the first ideas <strong>of</strong> this<br />
book at N.B.E.R. <strong>and</strong> Harvard University: the rigorous logic <strong>and</strong> the depth<br />
<strong>of</strong> the suggestions <strong>of</strong> Robert Barro have been crucial for my underst<strong>and</strong>ing <strong>of</strong>
xiv<br />
Preface<br />
many topics, <strong>and</strong> my way <strong>of</strong> thinking about economic issues is largely shaped<br />
around his free market ideals. At the time, I also benefited from interesting<br />
discussions with Philippe Aghion, Oliver Hart, David Laibson, Gregory<br />
Mankiw, Ricardo Reis, Silvana Tenreyro <strong>and</strong> Joseph Zeira.<br />
Since then, I have presented parts <strong>of</strong> this book at different conferences,<br />
seminars <strong>and</strong> lectures in many places around the world, including U.C.L.A.,<br />
Harvard University, University <strong>of</strong> Milan, Bicocca, CERGE <strong>and</strong> Charles University<br />
(Prague), European University Institute (Florence), University <strong>of</strong> Vienna,<br />
University <strong>of</strong> Virginia (Charlottesville), the Finnish <strong>Competition</strong> Authority<br />
<strong>and</strong> ETLA (Helsinki), the Roundtable on The Lisbon Agenda <strong>and</strong> the<br />
future <strong>of</strong> Information Technology IPRs (Brussels), the Telecom Conference<br />
on the Economics <strong>of</strong> the Information <strong>and</strong> Communication Technology (Paris),<br />
the Conference on <strong>Competition</strong> <strong>and</strong> Regulation <strong>of</strong> the Athens University <strong>of</strong><br />
Economics <strong>and</strong> Business (Corfù), the DIW Roundtable on <strong>Competition</strong> <strong>and</strong><br />
IPRs (Berlin), the Conference on EU <strong>and</strong> Greek <strong>Competition</strong> Policy (Athens)<br />
<strong>and</strong> others. I am grateful to many participants for important comments, <strong>and</strong><br />
especially to Jacques Bourgeois, Guglielmo Cancelli, David de Meza, Vincenzo<br />
Denicolò, David Encoua, Maxim Engers, David Evans, Leonardo Felli,<br />
Hans Jarle Kind, Joseph Harrington, Massimo Motta, Meir Pugatch, Jennifer<br />
Reinganum, Patrick Rey, David Ulph <strong>and</strong> Martti Virtanen. Between<br />
2002 <strong>and</strong> 2003, while I was economist for the Ministry <strong>of</strong> Economy <strong>of</strong> my<br />
country <strong>and</strong> teaching at Luiss University (Rome), I also benefited from interesting<br />
conversations with Riccardo Faini <strong>and</strong> Domenico Siniscalco on related<br />
policy issues.<br />
A large part <strong>of</strong> the antitrust implications <strong>of</strong> my theories is derived from my<br />
pr<strong>of</strong>essional experience as a consultant on antitrust issues for international<br />
organizations <strong>and</strong> private companies. I am thankful to many brilliant people<br />
from these organizations <strong>and</strong> companies with whom I have collaborated since<br />
2004, especially for providing a unique opportunity to apply, discuss <strong>and</strong> test<br />
many <strong>of</strong> the ideas presented in this book. However, the responsibility for what<br />
follows is only mine <strong>and</strong> should not involve any <strong>of</strong> the institutions I have been<br />
<strong>and</strong> am affiliated with.<br />
Since 2004, I have contributed to organize INTERTIC, the International<br />
Think-tank on <strong>Innovation</strong> <strong>and</strong> <strong>Competition</strong> (website www.intertic.org), <strong>and</strong><br />
I am extremely grateful to its co-founder <strong>and</strong> vice-president, Krešimir Žigić:<br />
interacting with him has been fundamental for many <strong>of</strong> the ideas presented<br />
in this book. Simon Anderson, also vice-president <strong>of</strong> <strong>Intertic</strong>, has been a continuous<br />
source <strong>of</strong> inspiration during the last years: I am extremely grateful<br />
for many <strong>of</strong> his precious comments. Similarly, I need to thank all the other<br />
members <strong>of</strong> <strong>Intertic</strong>, <strong>and</strong> especially Avinash Dixit, Yannis Katsoulacos, Vincenzo<br />
Denicolò, Barbara Spencer, Stephen Martin <strong>and</strong> Dennis Mueller for<br />
their valuable comments. The 2007 <strong>Intertic</strong> Conference, held at the University<br />
<strong>of</strong> Milan, Bicocca (“International Conference on <strong>Innovation</strong> <strong>and</strong> <strong>Competition</strong><br />
in the New Economy”, May 4-5, 2007) put together some <strong>of</strong> the best
Preface<br />
xv<br />
international economists working on issues <strong>of</strong> competition, innovation <strong>and</strong> industrial<br />
policy <strong>and</strong> has been a source <strong>of</strong> deep inspiration; I am thankful to all<br />
the participants, <strong>and</strong> especially to the members <strong>of</strong> <strong>Intertic</strong> <strong>and</strong> to Kris Aerts,<br />
Rabah Amir, Carlo Cambini, Guido Cozzi, Raymond De Bondt, Giovanni<br />
Dosi, Nisvan Erkal, Katerina Goldfain, Heli Koski, Kornelius Kraft, Eugen<br />
Kováč, Daniel Piccinin, Jan V<strong>and</strong>ekerckhove <strong>and</strong> Viatcheslav Vinogradov for<br />
stimulating debate.<br />
I completed this book at the University <strong>of</strong> Milan, Bicocca, one <strong>of</strong> the most<br />
modern <strong>and</strong> advanced challenges <strong>of</strong> graduate <strong>and</strong> postgraduate education in<br />
Italy. At its Department <strong>of</strong> Economics I found the ideal environment to write<br />
these pages. I am very grateful to all <strong>of</strong> my colleagues, especially Luigino<br />
Bruni, Floriana Cerniglia, Emilio Colombo, Mario Gilli, Giovanna Iannantuoni,<br />
Jean Jacques Lambin, Silvia Marchesi, Graziella Marzi, Mariapia Mendola,<br />
Ahmad Naimzada, Piergiovanna Natale, Pier Luigi Porta, Luca Stanca<br />
<strong>and</strong> Patrizio Tirelli, for many comments <strong>and</strong> suggestions on preliminary versions<br />
<strong>of</strong> the book.<br />
A special thanks to Flavia Ambrosanio, Massimo Bordignon, Umberto<br />
Galmarini <strong>and</strong> Piero Giarda from the Catholic University <strong>of</strong> Milan, who directed<br />
me toward the study <strong>of</strong> economic issues more than ten years ago, <strong>and</strong><br />
helped me with generosity <strong>and</strong> precious suggestions since then. I would also<br />
like to thank the Editor <strong>of</strong> Springer, Niels Peter Thomas, who has been extremely<br />
kind in supporting this project from the beginning <strong>and</strong> improving it<br />
in many ways, <strong>and</strong> Irene Barrios-Kezic for outst<strong>and</strong>ing editorial assistance.<br />
Finally, I am extremely grateful to Indira Pottebaum who read the manuscript<br />
many times <strong>and</strong> gave me a lot <strong>of</strong> precious comments.<br />
While preparing this book, I was teaching industrial organization <strong>and</strong><br />
competition policy to advanced undergraduates <strong>and</strong> I am thankful to my<br />
students at the University <strong>of</strong> Milan, Bicocca, for many questions <strong>and</strong> comments<br />
on Chapter 1. This chapter is extremely simplified <strong>and</strong> can be used<br />
for a short undergraduate course on oligopoly theory; an updated version for<br />
teaching purposes can be found at www.intertic.org (where other material related<br />
to this book can be found as well). Also Chapters 5, 6 <strong>and</strong> 7, which are<br />
entirely verbal, should be accessible to anyone who has no formal background<br />
in economic theory, but is interested in antitrust issues <strong>and</strong> in the evolution<br />
<strong>of</strong> the New Economy, the s<strong>of</strong>tware market <strong>and</strong> the Micros<strong>of</strong>t case. Chapters<br />
2, 3 <strong>and</strong> 4, however, are more advanced at a technical level <strong>and</strong> could be<br />
used for a postgraduate course on industrial organization or on the theory <strong>of</strong><br />
innovation. Finally, I tried to write each chapter as a self contained treatment<br />
<strong>of</strong> a particular topic, therefore the reader may also look at a chapter <strong>of</strong> his<br />
or her interest without having to read the previous parts.<br />
My approach to industrial organization issues is largely affected by studies<br />
in other fields as macroeconomics, international economics <strong>and</strong> business<br />
administration, <strong>and</strong> it probably reflects the fact that I have never taken a<br />
course in industrial organization. Also for these reasons, this book should be
xvi<br />
Preface<br />
seen as a complement <strong>of</strong> other graduate textbooks in the field, <strong>and</strong> not as a<br />
substitute. Tirole (1988) is the “first-mover” <strong>and</strong> still the leader in the market<br />
for game theoretic textbooks in industrial organization, but <strong>of</strong> course it does<br />
not include two decades <strong>of</strong> literature (especially on the theory <strong>of</strong> innovation<br />
<strong>and</strong> on the evolution <strong>of</strong> the post-Chicago approach to antitrust). Many other<br />
good <strong>and</strong> diversified textbooks have appeared (or endogenously entered) in<br />
this market in the following years. Shy (1995) <strong>of</strong>fers a wide review <strong>of</strong> basic<br />
models at an advanced undergraduate level. Anderson et al. (1992) <strong>and</strong><br />
Vives (1999) provide more sophisticated analysis respectively <strong>of</strong> the models<br />
with product differentiation <strong>and</strong> <strong>of</strong> the leading models <strong>of</strong> oligopolistic interaction,<br />
but they largely ignore the role <strong>of</strong> market leaders in their frameworks.<br />
Martin (2002) provides an excellent guide to many theoretical <strong>and</strong> empirical<br />
issues, but (as the other cited books) it contains a limited treatment <strong>of</strong> many<br />
aspects that are relevant to the markets <strong>of</strong> the New Economy, as network<br />
externalities, multi-sided markets, Schumpeterian theories <strong>of</strong> innovation <strong>and</strong><br />
the related antitrust issues. Scotchmer (2004) provides a nice overview <strong>of</strong><br />
the theory <strong>of</strong> innovation, but her analysis does not include the most recent<br />
progress in the theory <strong>of</strong> innovation by leaders <strong>and</strong> <strong>of</strong> its consequences for<br />
endogenous technological progress <strong>and</strong> for R&D policy. Motta (2004) is a<br />
useful survey <strong>of</strong> the theoretical <strong>and</strong> applied literature on antitrust policy before<br />
the advent <strong>of</strong> the endogenous entry approach <strong>and</strong> <strong>of</strong> the related policy<br />
implications. Finally, the classic books by Bork (1993) <strong>and</strong> Posner (2001) on<br />
the major achievements <strong>of</strong> the Chicago school could be also used in parallel<br />
to our treatment, which is largely aimed at formalizing some <strong>of</strong> the informal<br />
results <strong>of</strong> the Chicago view on antitrust policy.<br />
A last word on the cover <strong>of</strong> this book, for which I have chosen a painting<br />
by the Dutch artist Jan Vermeer, The Astronomer, now visible at the Louvre<br />
Museum in Paris. This masterpiece, painted in 1668, depicts a researcher<br />
engrossed in scientific investigation, <strong>and</strong> directing his attention toward a celestial<br />
globe, 3 metaphor <strong>of</strong> the sphere <strong>of</strong> knowledge at a time when a radical<br />
change <strong>of</strong> paradigm was taking place in science. The Astronomer seems to be<br />
pondering about the mysteries <strong>of</strong> the universe, <strong>and</strong> is working indoors without<br />
looking through the window at the heavens, but the penetrating light<br />
coming from the window is enlightening him, the globe, the astrolabe <strong>and</strong><br />
the focus <strong>of</strong> his work. Doing scientific research is a bit like touching a piece<br />
<strong>of</strong> the sphere <strong>of</strong> knowledge; the rest, as always, is left for future research. I<br />
hope you will have as much fun reading this book as I did in writing it.<br />
Federico Etro<br />
Department <strong>of</strong> Economics, University <strong>of</strong> Milan, Bicocca<br />
Milan, July 2007<br />
3 See James A. Welu, 1975, “Vermeer: His Cartographic Sources”, The Art Bulletin,<br />
Vol. 57 (4), pp. 529-47.
Contents<br />
Preface .......................................................<br />
vii<br />
1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry ....................... 1<br />
1.1 ASimpleModel<strong>of</strong><strong>Competition</strong>inQuantities.............. 4<br />
1.1.1 Monopoly<strong>and</strong><strong>Antitrust</strong>Issues..................... 5<br />
1.1.2 NashEquilibrium ................................ 8<br />
1.1.3 Marshall Equilibrium . ............................ 9<br />
1.1.4 StackelbergEquilibrium........................... 10<br />
1.1.5 StackelbergEquilibriumwithEndogenousEntry ..... 12<br />
1.2 Increasing Marginal Costs <strong>and</strong> Product Differentiation ...... 15<br />
1.2.1 U-shapedCostFunctions.......................... 16<br />
1.2.2 Product Differentiation ........................... 18<br />
1.3 ASimpleModel<strong>of</strong><strong>Competition</strong>inPrices.................. 20<br />
1.4 ASimpleModel<strong>of</strong><strong>Competition</strong>forthe<strong>Market</strong> ............ 25<br />
1.4.1 TheArrow’sParadox ............................. 27<br />
1.4.2 <strong>Innovation</strong>byLeaders ............................ 31<br />
1.5 Conclusions............................................ 34<br />
1.6 Appendix ............................................. 36<br />
2. Strategic Commitments <strong>and</strong> Endogenous Entry ........... 41<br />
2.1 <strong>Market</strong>Structure....................................... 44<br />
2.2 Nash Equilibrium....................................... 48<br />
2.3 Marshall Equilibrium ................................... 49<br />
2.4 <strong>Competition</strong>inQuantities,inPrices<strong>and</strong>forthe<strong>Market</strong>..... 50<br />
2.4.1 <strong>Competition</strong>inQuantities......................... 50<br />
2.4.2 <strong>Competition</strong>inPrices............................. 54<br />
2.4.3 <strong>Competition</strong>forthe<strong>Market</strong>........................ 58<br />
2.5 StrategicInvestments ................................... 59<br />
2.5.1 The Fudenberg-Tirole Taxonomy <strong>of</strong> Business Strategies 61<br />
2.5.2 StrategicCommitmentswithEndogenousEntry...... 63<br />
2.6 CostReductions<strong>and</strong>Signaling ........................... 66<br />
2.7 Advertising<strong>and</strong>Dem<strong>and</strong>EnhancingInvestments ........... 70<br />
2.8 Debt<strong>and</strong>theOptimalFinancialStructure................. 72<br />
2.9 NetworkExternalities<strong>and</strong>Two-Sided<strong>Market</strong>s ............. 76
xviii<br />
Contents<br />
2.10 Bundling .............................................. 79<br />
2.11 VerticalRestraints...................................... 82<br />
2.12 PriceDiscrimination.................................... 84<br />
2.13 <strong>Antitrust</strong><strong>and</strong>HorizontalMergers......................... 87<br />
2.14 Conclusions............................................ 89<br />
3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry .......... 91<br />
3.1 Stackelberg Equilibrium ................................. 94<br />
3.2 Stackelberg Equilibrium with Endogenous Entry............ 97<br />
3.3 <strong>Competition</strong>inQuantities,inPrices<strong>and</strong>forthe<strong>Market</strong>..... 100<br />
3.3.1 <strong>Competition</strong>inQuantities......................... 100<br />
3.3.2 <strong>Competition</strong>inPrices............................. 106<br />
3.3.3 <strong>Competition</strong>forthe<strong>Market</strong>........................ 108<br />
3.4 Asymmetries,MultipleLeaders<strong>and</strong>MultipleStrategies ..... 109<br />
3.4.1 AsymmetriesBetweenLeader<strong>and</strong>Followers ......... 109<br />
3.4.2 MultipleLeaders ................................. 110<br />
3.4.3 EndogenousLeadership ........................... 113<br />
3.4.4 MultipleStrategies ............................... 114<br />
3.4.5 General Pr<strong>of</strong>itFunctions .......................... 116<br />
3.5 <strong>Antitrust</strong><strong>and</strong>Collusion ................................. 118<br />
3.6 State-Aids<strong>and</strong>StrategicExportPromotion................ 120<br />
3.7 Privatizations.......................................... 123<br />
3.8 Conclusions............................................ 124<br />
3.9 Appendix ............................................. 125<br />
4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry ............ 131<br />
4.1 ASimplePatentRacewithContractualCosts<strong>of</strong>R&D...... 135<br />
4.1.1 EndogenousEntry................................ 138<br />
4.1.2 WelfareAnalysis ................................. 141<br />
4.2 Dynamic<strong>Competition</strong>forthe<strong>Market</strong> ..................... 142<br />
4.2.1 NashEquilibrium ................................ 143<br />
4.2.2 Marshall Equilibrium . ............................ 144<br />
4.2.3 StackelbergEquilibrium........................... 144<br />
4.2.4 StackelbergEquilibriumwithEndogenousEntry ..... 146<br />
4.2.5 Non-drastic<strong>Innovation</strong>s ........................... 148<br />
4.2.6 StrategicCommitments ........................... 150<br />
4.3 Sequential<strong>Innovation</strong>s .................................. 151<br />
4.3.1 EndogenousValue<strong>of</strong><strong>Innovation</strong>s................... 152<br />
4.3.2 EndogenousTechnologicalProgress................. 155<br />
4.4 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> <strong>Competition</strong> for the <strong>Market</strong> . 159<br />
4.5 Conclusions............................................ 162<br />
4.6 Appendix ............................................. 165
Contents<br />
xix<br />
5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance ....................... 171<br />
5.1 TheTraditionalApproachestoAbuse<strong>of</strong>Dominance ........ 174<br />
5.1.1 TheChicagoSchool .............................. 174<br />
5.1.2 ThePost-ChicagoApproach ....................... 176<br />
5.2 The<strong>Theory</strong><strong>of</strong><strong>Market</strong>Leaders<strong>and</strong>EndogenousEntry ...... 178<br />
5.2.1 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> Policy Implications . . . 179<br />
5.2.2 <strong>Competition</strong> for the <strong>Market</strong> <strong>and</strong> Policy Implications . . 186<br />
5.3 ADigressiononIPRsProtection ......................... 189<br />
5.3.1 PatentsinDynamicSectors<strong>and</strong><strong>Innovation</strong>s ......... 190<br />
5.3.2 Open-Source<strong>Innovation</strong>s .......................... 191<br />
5.3.3 ConclusionsonIPRsProtection.................... 194<br />
5.4 Reforming<strong>Antitrust</strong> .................................... 195<br />
5.4.1 EfficiencyDefense ................................ 196<br />
5.4.2 PredatoryPricing ................................ 197<br />
5.4.3 Bundling........................................ 201<br />
5.4.4 IntellectualPropertyRights ....................... 203<br />
5.5 Conclusions............................................ 204<br />
6. Micros<strong>of</strong>t Economics ...................................... 207<br />
6.1 TheS<strong>of</strong>tware<strong>Market</strong> ................................... 208<br />
6.1.1 Network Effects .................................. 210<br />
6.1.2 Multi-sidedPlatforms............................. 212<br />
6.1.3 Micros<strong>of</strong>t........................................ 215<br />
6.2 The<strong>Antitrust</strong>Cases .................................... 218<br />
6.2.1 TheUSCase .................................... 218<br />
6.2.2 TheEUCase .................................... 221<br />
6.3 IsMicros<strong>of</strong>taMonopolist?............................... 223<br />
6.3.1 WhyIsthePrice<strong>of</strong>WindowssoLow? .............. 225<br />
6.3.2 Does Micros<strong>of</strong>t Stifle<strong>Innovation</strong>?................... 228<br />
6.4 Bundling .............................................. 230<br />
6.4.1 Strategic Bundling . . . ............................ 232<br />
6.4.2 TechnologicalBundling ........................... 234<br />
6.5 IntellectualPropertyRights ............................. 235<br />
6.5.1 Patents, Trade Secrets <strong>and</strong> Interoperability .......... 236<br />
6.5.2 Licenses<strong>and</strong>St<strong>and</strong>ards ........................... 238<br />
6.6 Conclusions............................................ 240<br />
7. Epilogue .................................................. 243<br />
7.1 EmpiricalPredictions<strong>of</strong>the<strong>Theory</strong><strong>of</strong><strong>Market</strong>Leaders...... 243<br />
7.2 ImplicationsforBusinessAdministration .................. 252<br />
7.3 ImplicationsforEconomic<strong>Theory</strong> ........................ 252<br />
7.4 Conclusions............................................ 255<br />
8. References ................................................ 257
xx<br />
Contents<br />
Index ......................................................... 275
1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
Most <strong>of</strong> the traditional industrial organization literature has studied the way<br />
market structure affects the behavior <strong>of</strong> firms. This book is also about how<br />
the behavior <strong>of</strong> firms affects the market structure. Therefore we will focus<br />
on market structures where both the strategies <strong>of</strong> the firms <strong>and</strong> their entry<br />
choices are endogenous.<br />
We will study the strategies <strong>and</strong> the entry decisions within a general<br />
framework <strong>and</strong> apply the results to different environments, characterized by<br />
competition in the market <strong>and</strong> competition for the market. The difference<br />
between these two forms <strong>of</strong> competition is simple. When firms compete in<br />
the market, they choose the price <strong>of</strong> their products or the production level,<br />
or even other auxiliary strategies, but the products <strong>of</strong> all the firms are exogenously<br />
given. When firms compete for the market, they invest in R&D to<br />
innovate <strong>and</strong> create new products or better versions <strong>of</strong> the existing products.<br />
Our ultimate objective will be to employ our theoretical results to derive<br />
some insights on policy issues, <strong>and</strong> in particular on antitrust issues. For this<br />
purpose, we will pay a close attention to the behavior <strong>of</strong> market leaders <strong>and</strong><br />
to the interaction between these firms <strong>and</strong> the other firms, the followers.<br />
In this chapter we will study the simplest models <strong>of</strong> competition one can<br />
think <strong>of</strong>. Our purpose is to introduce the reader to the basic tools <strong>of</strong> the<br />
theory <strong>of</strong> oligopoly. Nevertheless, we will also present new insights on the<br />
behavior <strong>of</strong> leaders in markets where entry is endogenous. In the rest <strong>of</strong> the<br />
book we will generalize these results in many directions, but the spirit <strong>of</strong> our<br />
analysis can be grasped from the examples developed in this chapter.<br />
We will focus on four general typologies <strong>of</strong> competition <strong>and</strong> their related<br />
equilibria. The first typology goes back to the early analysis <strong>of</strong> Cournot (1838)<br />
who was the real pioneer <strong>of</strong> the modern economic analysis <strong>and</strong> the first one<br />
to study market structures for homogeneous goods where firms choose their<br />
output <strong>and</strong> where the equilibrium between dem<strong>and</strong> by consumers <strong>and</strong> supply<br />
by all firms determines the price. While the analysis <strong>of</strong> Cournot goes back<br />
to the first half <strong>of</strong> the XIX century, his equilibrium concept corresponds to<br />
the one that today we associate with Nash (1950): 1 each firm independently<br />
chooses its strategy to maximize pr<strong>of</strong>its taking as given the strategy <strong>of</strong> each<br />
1 Nash (1950) introduced mixed strategy equilibria <strong>and</strong> provided a general pro<strong>of</strong><br />
<strong>of</strong> the existence <strong>of</strong> these equilibria.
2 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
other firm. This idea can be applied to more general market structures <strong>and</strong><br />
also when firms choose strategies different from their output, for instance<br />
when they choose their prices, or their investments in R&D. Therefore, we<br />
will generally refer to a Nash equilibrium when an exogenous number <strong>of</strong> firms<br />
compete choosing their strategies simultaneously. This equilibrium concept<br />
is at the basis <strong>of</strong> any analysis <strong>of</strong> strategic interactions between independent<br />
agents, <strong>and</strong> in particular at the basis <strong>of</strong> the theory <strong>of</strong> industrial organization.<br />
The second typology <strong>of</strong> competition extends these models <strong>of</strong> imperfect<br />
competition to endogenous entry <strong>of</strong> firms. A market is in equilibrium only<br />
when there are not further incentives for other firms to enter into it <strong>and</strong><br />
conquer positive extra-pr<strong>of</strong>its. This idea is <strong>of</strong>ten associated with the studies<br />
on competitive markets in partial equilibrium <strong>of</strong> the second half <strong>of</strong> the XIX<br />
century, in particular with Marshall (1890). Therefore, we will refer to this<br />
equilibrium as the Marshall equilibrium. In modern terms, the concept <strong>of</strong><br />
Nash equilibrium with free entry characterizes this situation. Formal treatments<br />
have been provided by von Weizsäcker (1980) <strong>and</strong> Novshek (1980) for<br />
competition in quantities <strong>and</strong> (neglecting the strategic interactions) by Dixit<br />
<strong>and</strong> Stiglitz (1977) for competition in prices. In general, equilibria with endogenous<br />
entry are the natural way to think <strong>of</strong> medium <strong>and</strong> long run equilibria<br />
both in partial <strong>and</strong> general equilibrium. Nevertheless, they have been rarely<br />
used in industrial organization, where the number <strong>of</strong> competitors is <strong>of</strong>ten<br />
assumed exogenous to focus on the strategic interactions between predetermined<br />
competitors, <strong>and</strong> also in general equilibrium macroeconomic analysis<br />
with imperfect competition (which <strong>of</strong>ten abstracts from entry processes to<br />
focus on price rigidities).<br />
The third typology <strong>of</strong> competition was introduced by Stackelberg (1934)<br />
who studied markets where a firm has a leadership over the others. While in<br />
every day talks a market leadership refers to a vague concept <strong>of</strong> competitive<br />
advantage, in economic jargon a leadership is associated with a first mover<br />
advantage, that is the ability to choose strategies <strong>and</strong> commit to them before<br />
the other firms. Under Stackelberg competition, the leader can exploit its<br />
first mover advantage taking into account the reactions <strong>of</strong> the followers. 2<br />
Notice that the behavior <strong>of</strong> a leader in a Stackelberg equilibrium requires<br />
a commitment power whose credibility is crucial but sometimes not realistic<br />
(see Schelling, 1960). However, Dixit (1980) <strong>and</strong> Fudenberg <strong>and</strong> Tirole (1984)<br />
have shown that proper preliminary investments can be a valid substitute for<br />
this commitment: a firm can invest in cost reductions, in advertising, in R&D<br />
or in other strategic investments to obtain a competitive advantage over the<br />
other firms. We will return to this possibility in the next chapter, while in this<br />
one we will analyze the simpler case in which a leader has indeed the ability<br />
2 Only later on, Selten (1965) introduced the concept <strong>of</strong> subgame perfect equilibrium<br />
for dynamic games (<strong>and</strong> the Stackelberg equilibrium belongs to this class),<br />
while Harsanyi (1967-68) introduced Bayesian equilibria with uncertainty. For an<br />
introduction to game theory see Fudenberg <strong>and</strong> Tirole (1991) or Myerson (1991).
1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry 3<br />
to commit to strategies before the other firms. For example, the leader can<br />
choose how much to produce before them. Since the equilibrium price depends<br />
on the production <strong>of</strong> all the firms, the followers must take in consideration the<br />
production <strong>of</strong> the leader when they decide their own production: for instance,<br />
they may want to produce less if the leader has decided to produce more.<br />
But the leader is aware <strong>of</strong> these reactions, <strong>and</strong> decides its own production<br />
level taking into account the expected behavior <strong>of</strong> the followers: for example,<br />
the leader may want to produce a lot to induce the followers to reduce their<br />
production. Similarly, a price leader chooses its own price taking into account<br />
the impact <strong>of</strong> this choice on the prices adopted by the followers. Imagine that<br />
the followers are going to increase their prices when they face a price increase<br />
by the leader: then, the leader may want to choose a high price to start with,<br />
so that all firmswillendupwithhighprices.<br />
The last typology <strong>of</strong> competition completes our taxonomy <strong>of</strong> the basic<br />
forms <strong>of</strong> market interaction combining the analysis <strong>of</strong> leadership <strong>and</strong> entry.<br />
In the second half the XX century there have been some attempts to model<br />
both these elements. One is the literature on entry deterrence associated<br />
with Bain (1956), Sylos Labini (1956) <strong>and</strong> Modigliani (1958), who took into<br />
consideration the effects <strong>of</strong> entry on the predatory behavior <strong>of</strong> market leaders<br />
mainly in the case <strong>of</strong> perfectly substitute goods <strong>and</strong> constant or decreasing<br />
marginal costs. Another important attempt is associated with the theory<br />
<strong>of</strong> contestable markets by Baumol et al. (1982), which shows that, in the<br />
absence <strong>of</strong> sunk costs <strong>of</strong> entry, the possibility <strong>of</strong> “hit <strong>and</strong> run” strategies<br />
by potential entrants is compatible only with an equilibrium price equal to<br />
the average cost. One <strong>of</strong> the main implications <strong>of</strong> this result is that “one<br />
firm can be enough” for competition when there is at least one aggressive<br />
potential entrant. This theory <strong>and</strong> its implications do not apply when goods<br />
are imperfect substitute or firm compete in quantities rather than in prices,<br />
which represents a crucial theoretical gap.<br />
These <strong>and</strong> other attempts were not developed in a coherent <strong>and</strong> general<br />
game theoretic framework. The development <strong>of</strong> such a framework is the focus<br />
<strong>of</strong> this book, whose theoretical contribution is the characterization <strong>of</strong> the<br />
Stackelberg equilibrium with endogenous entry <strong>and</strong> <strong>of</strong> its applications. This<br />
equilibrium is characterized by rational strategies adopted in different stages.<br />
In a first stage, the leader chooses its strategy under rational expectations on<br />
the strategies that will be adopted by the followers <strong>and</strong> on the entry decisions<br />
<strong>of</strong> these followers. In a second stage the followers decide whether to enter in<br />
the market or not according to their expectations on pr<strong>of</strong>itability. In the<br />
last stage, the followers simultaneously choose their strategies to maximize<br />
pr<strong>of</strong>its, knowing the strategy <strong>of</strong> the leader <strong>and</strong> taking as given the strategies<br />
<strong>of</strong> the other followers.<br />
This introductory chapter presents, in the simplest possible way, some<br />
examples <strong>of</strong> these four different forms <strong>of</strong> competition <strong>and</strong> equilibria. Our<br />
initial focus is on models <strong>of</strong> competition in quantities. After presenting the
4 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
basic linear model which assumes constant marginal costs <strong>and</strong> homogenous<br />
goods in Section 1.1, we extend it to U-shaped cost functions <strong>and</strong> to product<br />
differentiation in Section 1.2. In Section 1.3, we present a simple model <strong>of</strong><br />
competition in prices with a Logit dem<strong>and</strong> function. Finally, in Section 1.4,<br />
we discuss a simple model <strong>of</strong> competition for the market (a contest where<br />
firms compete investing with the purpose <strong>of</strong> conquering a new market), <strong>and</strong><br />
we analyze the role <strong>of</strong> incumbent monopolists (with or without a leadership<br />
in the competition for the market). Section 1.5 concludes.<br />
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities<br />
Our initial example will be about the simplest situation one can think <strong>of</strong>: a<br />
market for a single homogenous good whose supply requires a positive fixed<br />
cost <strong>of</strong> production <strong>and</strong> a constant additional cost for each unit produced,<br />
which means that the marginal cost <strong>of</strong> production is constant. To be more<br />
formal, imagine a good whose dem<strong>and</strong> is linearly decreasing in the price, say<br />
D(p) =a−p where a>0 is a parameter representing the size <strong>of</strong> the market. If<br />
total production by all the firms is Q = P n<br />
i=1 q i,whereq i is the production<br />
<strong>of</strong> each firm i =1, 2, ..., n, in equilibrium between supply <strong>and</strong> dem<strong>and</strong> we<br />
must have Q = D(p) =a − p, which provides the so called inverse dem<strong>and</strong><br />
function in equilibrium:<br />
p = a − Q = a −<br />
nX<br />
q i (1.1)<br />
i=1<br />
Basically, the larger production is, the smaller the equilibrium price must<br />
be.<br />
Imagine now that each firm can produce the good with the same st<strong>and</strong>ard<br />
technology. Producing q unitsrequiresafixed cost <strong>of</strong> production F ≥ 0 <strong>and</strong><br />
a variable cost cq where c ∈ [0,a) is a constant marginal cost <strong>of</strong> production.<br />
Notice that, while the average variable cost is constant (equal to c), the<br />
average total cost (equal to c+F/q) is decreasing in the output. In conclusion,<br />
the pr<strong>of</strong>it function <strong>of</strong> a firm i is the difference between revenues <strong>and</strong> costs:<br />
π i = pq i − cq i − F = (1.2)<br />
à !<br />
nX<br />
= a − q i q i − cq i − F<br />
i=1<br />
Before analyzing different forms <strong>of</strong> competition between many firms in<br />
this set up, we will investigate a few simple <strong>and</strong> extreme situations where<br />
one or two firms only are active in this market <strong>and</strong> derive some preliminary<br />
implications for antitrust analysis.
1.1.1 Monopoly <strong>and</strong> <strong>Antitrust</strong> Issues<br />
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities 5<br />
Our first investigation <strong>of</strong> the market described above focuses on a monopoly.<br />
Consider a single firm producing q. Itspr<strong>of</strong>it mustbegivenbyπ =(a −<br />
q)q − cq − F . Its maximization requires an output satisfying the optimality<br />
condition ∂π/∂q = a − 2q − c =0, 3 which can be solved for the monopolistic<br />
output:<br />
q M = a − c<br />
2<br />
The monopolistic price can be derived from the inverse dem<strong>and</strong> function as<br />
p M = a − q M =(a + c)/2, <strong>and</strong> the associated pr<strong>of</strong>its are: 4<br />
(a − c)2<br />
π M = − F<br />
4<br />
Imagine now that another firm enters in the market. When the two firms<br />
compete at the same level, it is natural to imagine that their strategic choices<br />
are taken simultaneously. In the equilibrium <strong>of</strong> this duopoly, both firms must<br />
choose their output levels independently, <strong>and</strong> these output levels must be<br />
consistent with each other. The result is a Cournot equilibrium.<br />
Consider firms i <strong>and</strong> j. If they compete choosing independently their<br />
outputs, firm i has the following pr<strong>of</strong>it functionπ i =(a − q i − q j )q i − cq i − F ,<br />
<strong>and</strong> total production is now Q = q i + q j ; <strong>of</strong> course the pr<strong>of</strong>it <strong>of</strong>firm j is<br />
the same after changing all indexes. Pr<strong>of</strong>it maximization by firm i requires<br />
∂π i /∂q i =0or a − 2q i − q j = c, from which we obtain the so called reaction<br />
function:<br />
q i (q j )= a − c − q j<br />
2<br />
This is a rule <strong>of</strong> behavior for firm i whichcanbeinterpretedinterms<strong>of</strong><br />
expectations: the larger is the expected production <strong>of</strong> firm j, the smaller<br />
should be the optimal production <strong>of</strong> firm i. Firmj will follow a similar rule:<br />
q j (q i )= a − c − q i<br />
2<br />
The geniality <strong>of</strong> Cournot’s idea is that in equilibrium the two rules must be<br />
consistent with each other. In terms <strong>of</strong> expectations, the equilibrium production<br />
<strong>of</strong> each firm must be the optimal one given the expectation that the other<br />
firm adopts its equilibrium production. Mathematically, we can solve the system<br />
<strong>of</strong> the two reaction functions to find out the production <strong>of</strong> each firm in<br />
3 The second order condition ∂ 2 π/∂q∂q = −2 < 0 guarantees that the pr<strong>of</strong>it<br />
function is concave, so that the solution corresponds to a maximum.<br />
4 We will assume that F is small enough to allow pr<strong>of</strong>itable entry by one firm in<br />
the market.
6 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
equilibrium. It is easy to verify that there is only one consistent equilibrium,<br />
<strong>and</strong> it implies that each firm produces the same amount:<br />
q = a − c<br />
3<br />
Accordingly, the equilibrium price is p =(a +2c) /3, <strong>and</strong>thepr<strong>of</strong>it <strong>of</strong>each<br />
firm is:<br />
(a − c)2<br />
π C = − F<br />
9<br />
<strong>Competition</strong> increases total production <strong>and</strong> reduces the price <strong>and</strong> the pr<strong>of</strong>its<br />
compared to the monopolistic case. For this reason, the firms may engage<br />
in alternative agreements or strategies that can increase the price <strong>and</strong> their<br />
pr<strong>of</strong>its. Any practice that leads to higher prices ends up hurting consumers. 5<br />
The scope <strong>of</strong> antitrust policy is precisely to avoid this kind <strong>of</strong> anti-competitive<br />
behavior. Here, we will sketch the main anti-competitive practices that can<br />
emerge in such a simple context.<br />
Mergers. As we noticed, the Cournot duopoly generates lower pr<strong>of</strong>its for<br />
each firm compared to a monopoly. Moreover, also the sum <strong>of</strong> the pr<strong>of</strong>its<br />
<strong>of</strong> both firmsislowerthanthepr<strong>of</strong>its <strong>of</strong> the monopolist. This implies that<br />
there is an incentive for one firm to merge with the other one, monopolize<br />
the market <strong>and</strong> increase total pr<strong>of</strong>its. Since this induces a higher final price,<br />
antitrust authorities should prevent a similar horizontal merger as an attempt<br />
to monopolize the market. 6<br />
Abuse <strong>of</strong> Dominance. There is another possibility for one <strong>of</strong> the two firms<br />
to increase its pr<strong>of</strong>its. This possibility emerges when this firm can act as a<br />
leader <strong>and</strong> choose its output before the second firm. In this case, the leader i<br />
could chose a output level ¯q which is high enough to convince the second firm<br />
j to avoid entry. This entry deterring output level can be calculated as follows.<br />
Consider the reaction function <strong>of</strong> firm j derived above: this tells us that when<br />
firm i produces q, firm j finds it optimal to produce q j (q) =(a − c − q)/2 so<br />
as to obtain pr<strong>of</strong>its π j (q) =(a−c−q) 2 /4−F .Now,theleaderi is aware that<br />
producing ¯q = a − c − 2 √ F will reduce the pr<strong>of</strong>its <strong>of</strong> the other firm j to zero<br />
(π j (¯q) =0). This is the entry deterring strategy, <strong>and</strong> it allows the leader to<br />
remain alone in the market. If this firm has the market power to choose its<br />
5 In this model with linear dem<strong>and</strong>, consumer surplus is simply the area below the<br />
dem<strong>and</strong> curve <strong>and</strong> above the market price, which corresponds to Q 2 /2. Welfare<br />
is traditionally defined as the sum <strong>of</strong> consumer surplus <strong>and</strong> firms’ pr<strong>of</strong>its, W =<br />
Q 2 /2+ n<br />
i=1 π i.<br />
6 Notice, however, that in case the merger between the two firms allows to save<br />
one <strong>of</strong> the two fixedcosts,thegaininefficiency may overcompensate the loss<br />
in consumer surplus after the merger (see Williamson, 1968, <strong>and</strong> Farrell <strong>and</strong><br />
Shapiro, 1990, on a more general analysis <strong>of</strong> efficiencies in horizontal mergers).
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities 7<br />
strategy before the rival, it can use this power to increase pr<strong>of</strong>its excluding<br />
entry. 7 Moreover, if this firm remains alone in the market, it could be able<br />
to restore the monopolistic price in the future. When this is the case, the<br />
exclusionary strategy ends up increasing the final price, therefore antitrust<br />
authorities should punish it as a predatory strategy.<br />
Collusion. A third way to increase pr<strong>of</strong>its requires collusion. To see how<br />
it works in our simple setup, let us go back to the symmetric duopoly. The<br />
reduction in total pr<strong>of</strong>its associated with Cournot competition (compared<br />
to the monopolistic outcome) was due to the fact that each firm did not<br />
take into consideration the impact <strong>of</strong> its own production on the pr<strong>of</strong>its <strong>of</strong><br />
the other firm, <strong>and</strong> hence tended to produce too much from the point <strong>of</strong><br />
view <strong>of</strong> joint pr<strong>of</strong>it maximization. This externality leads to a price reduction<br />
<strong>and</strong> to a decline in total pr<strong>of</strong>its. For this reason the two firms may try to<br />
collude <strong>and</strong> agree on limiting their production at a lower level, possibly at<br />
the monopolistic level. Under perfect collusion, each one <strong>of</strong> the two firms<br />
produces half <strong>of</strong> the monopolistic output, ˜q =(a − c)/4, <strong>and</strong>obtainspr<strong>of</strong>its<br />
˜π =(a − c) 2 /8 − F .<br />
However, only a strong <strong>and</strong> reciprocal commitment could guarantee that<br />
such a collusive behavior is sustainable, because in the absence <strong>of</strong> a commitment<br />
each firm would have incentives to deviate <strong>and</strong> produce more than<br />
that. For instance, if a firm is sure that the other one produces at the collusive<br />
level, this firm can deviate from the collusive strategy <strong>and</strong> choose an<br />
output q D that maximizes π =(a − q D − ˜q)q D − cq D − F .Theoptimaldeviation<br />
is exactly q D =3(a − c)/8. After deviating from the collusive strategy,<br />
this firm increases its pr<strong>of</strong>its to π D =9(a − c) 2 /64 − F , which is above the<br />
collusive pr<strong>of</strong>its, while the pr<strong>of</strong>its <strong>of</strong> the other firm are reduced below them.<br />
This pr<strong>of</strong>itable deviation should not surprise, because there must be always<br />
apr<strong>of</strong>itable deviation for each firm when we are not in the Cournot equilibrium.<br />
Not by chance, we can also provide an alternative definition <strong>of</strong> the<br />
Cournot equilibrium as one in which there are not pr<strong>of</strong>itable deviations for<br />
any firm.<br />
It is important to notice that collusive outcomes can be reached more<br />
easily when interactions are repeated over time, because deviations can be<br />
punished in the future, <strong>and</strong> the threat <strong>of</strong> punishments can reduce the incentives<br />
to deviate. The theory <strong>of</strong> collusion has studied the conditions under<br />
which monopolistic pr<strong>of</strong>its can be sustained in dynamic games. For instance,<br />
if the same competition is repeated infinite times, each firm discounts the<br />
future, <strong>and</strong> each deviation is punished with reversion to the Cournot equilibrium<br />
forever, collusion is sustainable if <strong>and</strong> only if firms are patient enough.<br />
7 Notice, however, that the exclusionary strategy does not necessarily increase the<br />
price <strong>and</strong>, even if it increases the price, it does not necessarily reduce welfare<br />
(measured as consumer surplus plus pr<strong>of</strong>its). If the fixed cost <strong>of</strong> production is<br />
high enough, entry deterrence may require a higher price but it may be more<br />
efficient from a welfare point <strong>of</strong> view.
8 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
Of course, collusion could be sustained more easily if harder punishments<br />
were available (for instance with a reversion to zero pr<strong>of</strong>its forever). 8 Since<br />
collusive cartels allow firms to set higher equilibrium prices, antitrust authorities<br />
should prevent similar agreements.<br />
As we have seen, simple games can be useful to underst<strong>and</strong> basic strategic<br />
interactions <strong>and</strong> to approach some <strong>of</strong> the fundamental antitrust issues.<br />
However, a more complete analysis needs to take into account the presence<br />
<strong>of</strong> more than just one or two firms, <strong>and</strong> possibly also to endogenize entry in<br />
the market. To these tasks we now turn.<br />
1.1.2 Nash Equilibrium<br />
We now move to the study <strong>of</strong> a generalized Nash competition between many<br />
firms. In particular, imagine that there are n firms in the same market described<br />
above. Each firm i will have pr<strong>of</strong>its:<br />
⎛<br />
⎞<br />
nX<br />
π i = ⎝a − q i − q j<br />
⎠ q i − cq i − F (1.3)<br />
j=1,j6=i<br />
<strong>and</strong> will choose its production q i to satisfy the first order condition a − 2q i −<br />
P n<br />
j=1,j6=i q j = c, which generates the reaction function:<br />
q i = a − P n<br />
j=1,j6=i q j − c<br />
2<br />
Notice that this is decreasing in the output <strong>of</strong> each other firm, ∂q i /∂q j < 0.<br />
Therefore, when a firm is expected to increase its own production, any other<br />
firm has an incentive to choose a lower production level. This is a typical<br />
property <strong>of</strong> models where firms compete in quantities.<br />
Thesystem<strong>of</strong>n conditions provides equilibrium outputs as in the duopoly<br />
case. However, its solution is immediate if we notice that all firms will produce<br />
8 Assume that the punishment is reversion to the Cournot equilibrium <strong>and</strong> that<br />
the discount factor is δ ∈ (0, 1). Collusion is sustainable if the discounted pay<strong>of</strong>f<br />
from collusion forever, ˜π +δ˜π +δ 2˜π +... =˜π/(1−δ), is higher than the deviation<br />
pay<strong>of</strong>f in one period plus the Cournot pay<strong>of</strong>f forever after that, π D +δπ C /(1−δ).<br />
This requires δ>(π D − ˜π) / (π D − π C). Substituting for the pay<strong>of</strong>fs, one can<br />
find that collusion is sustainable when δ > 9/17. Thefirst generalizations <strong>of</strong><br />
this result, known as Folk Theorem, are in Friedman (1971) <strong>and</strong> Aumann <strong>and</strong><br />
Shapley (1976). Of course, collusion could be sustained more easily if punishment<br />
was harder. Considering the maximum punishment which delivers zero expected<br />
pay<strong>of</strong>f for the deviator, Abreu (1986) has verified under which conditions such a<br />
punishment is itself sustainable, relaxing the condition above (see also Fudenberg<br />
<strong>and</strong> Maskin, 1986). For a wide treatment on supergames <strong>and</strong> dynamic games see<br />
Mailath <strong>and</strong> Samuelson (2006).
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities 9<br />
the same output satisfying a − 2q − (n − 1)q = c. This implies the following<br />
output per firm as a function <strong>of</strong> n: 9<br />
q(n) = a − c<br />
(1.4)<br />
n +1<br />
with total production Q(n) =n(a − c)/(n +1), which is increasing in the<br />
number <strong>of</strong> firms. The equilibrium price can be derived as:<br />
p(n) = a + nc<br />
(1.5)<br />
n +1<br />
which is decreasing in the number <strong>of</strong> firms <strong>and</strong> approaching the marginal cost<br />
<strong>of</strong> production when the number <strong>of</strong> firms increases. Nevertheless, the pr<strong>of</strong>its<br />
<strong>of</strong> each firm are constrained by the fixed costs <strong>of</strong> production:<br />
π(n) =<br />
µ 2 a − c<br />
− F<br />
n +1<br />
The pr<strong>of</strong>its <strong>of</strong> each single firm are clearly decreasing when the number <strong>of</strong><br />
competitors is increasing. This suggests that in the medium <strong>and</strong> long run,<br />
new firms will enter in the market as long as there are positive pr<strong>of</strong>its to<br />
be made, <strong>and</strong> they will stop entering when the number <strong>of</strong> firmsachievesan<br />
upper bound. This leads us to the next equilibrium concept.<br />
1.1.3 Marshall Equilibrium<br />
It is now extremely simple to extend the model to endogenize entry. Formally,<br />
consider the following sequence <strong>of</strong> moves:<br />
1) in the first stage all potential entrants simultaneously decide “in” or<br />
“out”;<br />
2) in the second stage all the firms that have entered choose their own<br />
strategy q i .<br />
In what follows we will mainly refer to F as to a technological cost <strong>of</strong><br />
production, but one could think <strong>of</strong> it as including other concrete fixed costs<br />
<strong>of</strong> entry or opportunity costs <strong>of</strong> participation to the market, as the pr<strong>of</strong>its<br />
that an entrepreneur can obtain in another sector. Beyond the particular<br />
interpretation, the role in constraining entry is the same.<br />
As we have seen, in the case <strong>of</strong> a Nash equilibrium the entry <strong>of</strong> a new<br />
firm enhances competition leading to a reduction in the pr<strong>of</strong>it <strong>of</strong> each single<br />
firm in the market. If we assume that entry takes place as long as positive<br />
pr<strong>of</strong>its can be obtained, a Marshall equilibrium should be characterized by a<br />
number <strong>of</strong> firms n satisfying a no entry condition π(n +1)< 0 <strong>and</strong> a no exit<br />
condition π(n) ≥ 0. When the fixed cost <strong>of</strong> production is small enough, this<br />
9 One can verify that both the cases <strong>of</strong> a monopoly <strong>and</strong> <strong>of</strong> the Cournot duopoly<br />
are particular cases for n =1<strong>and</strong> n =2.
10 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
equilibrium number is quite large. In these cases it is natural to take a short<br />
cut <strong>and</strong> approximate the endogenous number <strong>of</strong> firms with the real number<br />
satisfying the zero pr<strong>of</strong>it condition π(n) =0,thatis:<br />
n = a √ − c − 1<br />
F<br />
This allows one to derive the equilibrium output per firm under Marshall<br />
competition:<br />
q = √ F (1.6)<br />
the total production Q = a − c − √ F , <strong>and</strong> the equilibrium price:<br />
p = c + √ F (1.7)<br />
which implies a mark up on the marginal cost to cover the fixed costs <strong>of</strong><br />
production. When the fixed costs are zero, the outcome corresponds to the<br />
classic equilibrium with perfect competition in which the price is equal to the<br />
marginal cost <strong>and</strong> the number <strong>of</strong> firmsisindeterminate.Inthemorerealistic<br />
case in which start up costs for each firm are positive, the equilibrium is<br />
inefficient <strong>and</strong> there are too many firms pricing above their marginal cost. 10<br />
1.1.4 Stackelberg Equilibrium<br />
Let us now consider the case in which one <strong>of</strong> the firms has a first mover<br />
advantage <strong>and</strong> can choose its output in a first stage before the followers,<br />
while these choose their own output in a second stage <strong>and</strong> independently<br />
from each other. Let us define the production <strong>of</strong> the leader as q L .Inthe<br />
second stage each follower decides how much to produce according to the<br />
first order condition a − q L − q i − P n<br />
j=1,j6=L q j = c, wheren is the number <strong>of</strong><br />
firms (including the leader). Assuming that all the followers find it convenient<br />
to be active, in a symmetric equilibrium each follower produces:<br />
q(q L ,n)= a − q L − c<br />
n<br />
10 Adopting the st<strong>and</strong>ard definition <strong>of</strong> welfare (which here corresponds to the consumer<br />
surplus because all firms earn no pr<strong>of</strong>its under free entry), we have:<br />
W FE = Q2<br />
2 = (a − c − √ F ) 2<br />
2<br />
Notice that in this case the firstbestwouldrequireonesinglefirm producing<br />
Q = a − c with welfare:<br />
W FB =<br />
(a − c)2<br />
2<br />
− F
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities 11<br />
As we noticed before, ∂q(q L ,n)/∂q L < 0: the production <strong>of</strong> the leader partially<br />
crowds out the production <strong>of</strong> the other firms. Accordingly, in the first<br />
stage the leader perceives its pr<strong>of</strong>its as:<br />
π L =[a − q L − (n − 1)q(q L ,n)] q L − cq L − F<br />
We can already see what will be the impact <strong>of</strong> the behavior <strong>of</strong> the followers<br />
on the leader: since a higher production <strong>of</strong> the leader reduces the production<br />
<strong>of</strong> the followers, the leader has an indirect (or strategic) incentive to increase<br />
its production. Such an aggressive strategy reduces the production <strong>of</strong> the<br />
followers <strong>and</strong> shifts pr<strong>of</strong>its toward the same leader. Formally, we can rewrite<br />
the pr<strong>of</strong>its <strong>of</strong> the leader as:<br />
∙<br />
π L = a − q L − (n − 1) (a − q ¸<br />
L − c)<br />
q L − cq L − F =<br />
n<br />
µ <br />
a − c − qL<br />
=<br />
q L − F<br />
n<br />
which leads to the optimal strategy:<br />
q L = a − c<br />
(1.8)<br />
2<br />
In this particular example the leader finds it optimal to commit to produce<br />
at the monopolistic level. As a consequence, each one <strong>of</strong> the followers will end<br />
up producing:<br />
µ a − c<br />
q<br />
2 ,n = a − c<br />
(1.9)<br />
2n<br />
The total output becomes:<br />
µ<br />
Q =(a − c) 1 − 1 <br />
2n<br />
<strong>and</strong> the equilibrium price is:<br />
p(n) = a µ<br />
2n + c 1 − 1 <br />
2n<br />
(1.10)<br />
which again tends toward the marginal cost when the number <strong>of</strong> firms increases.<br />
The pr<strong>of</strong>its for the leader <strong>and</strong> for each follower are respectively:<br />
π L (n) =<br />
(a − c)2<br />
4n<br />
− F , π(n) =<br />
(a − c)2<br />
4n 2<br />
− F
12 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
Of course, entry <strong>of</strong> followers occurs if positive pr<strong>of</strong>its can be obtained. 11 When<br />
this is the case, we expect that, at least in the medium or long run, followers<br />
will keep entering in the market until positive pr<strong>of</strong>its can be made. Since<br />
the pr<strong>of</strong>its <strong>of</strong> the followers are decreasing in the number <strong>of</strong> firms active in<br />
themarket,theentryprocesswillhaveanaturallimit.Wenowmovetothe<br />
equilibrium in which entry occurs until all the pr<strong>of</strong>itable opportunities are<br />
exploited by the followers. As we will see, this equilibrium with endogenous<br />
entry is quite different from the one analyzed here.<br />
1.1.5 Stackelberg Equilibrium with Endogenous Entry<br />
Let us finally move to the last case, in which there is still a leader in the<br />
market, but this is facing endogenous entry <strong>of</strong> followers. Formally, following<br />
Etro (2006,a, pp. 147-8) consider the following sequence <strong>of</strong> moves:<br />
1) in the first stage, the leader chooses its own output q L ;<br />
2) in the second stage, after knowing the output <strong>of</strong> the leader, all potential<br />
entrants simultaneously decide “in” or“out”;<br />
3) in the third stage, all the followers that have entered choose their own<br />
output q i (hence, the followers play Nash between themselves).<br />
In this case, the leader has to take into account how its own commitment<br />
affects not only the strategy <strong>of</strong> the followers but also their entry decision. As<br />
we have already seen, in the last stage, if there are n ≥ 2 firms in the market<br />
<strong>and</strong> the leader produces q L , each follower produces:<br />
q(q L ,n)= a − q L − c<br />
n<br />
This implies that the pr<strong>of</strong>its <strong>of</strong> each follower are:<br />
µ 2 a − c − qL<br />
π(q L ,n)=<br />
− F (1.11)<br />
n<br />
which are clearly decreasing in the number <strong>of</strong> firms. This would imply that<br />
further entry or exit does not take place when π(q L ,n+1) ≤ 0 <strong>and</strong> π(q L ,n) ≥<br />
0. Moreover, no follower will find it optimal to enter in the market if π(q L , 2) ≤<br />
0, that is if not even a single follower can obtain positive pr<strong>of</strong>its given the<br />
output <strong>of</strong> the leader. This condition is equivalent to:<br />
q L ≥ a − c − 2 √ F<br />
Therefore when the leader adopts an aggressive strategy producing more<br />
than this cut-<strong>of</strong>f level entry will be deterred, but when the leader produces<br />
11 At least one follower has incentives to enter in the market if π(2) > 0 or F <<br />
(a−c) 2 /16, otherwise the leader supplies its monopolistic production <strong>and</strong> no one<br />
else enters. In what follows we assume away this possibility (which corresponds<br />
to the case <strong>of</strong> a “natural monopoly”).
1.1 A Simple Model <strong>of</strong> <strong>Competition</strong> in Quantities 13<br />
less than the above cut-<strong>of</strong>f the number <strong>of</strong> entrants will be determined by a<br />
free entry condition. In this last case, ignoring the integer constraint on the<br />
number <strong>of</strong> firms, 12 we can approximate the number <strong>of</strong> firmsasarealnumber<br />
that satisfies π(q L ,n)=0. This implies:<br />
n = a − c − q L<br />
√<br />
F<br />
(1.12)<br />
When this is the endogenous number <strong>of</strong> firms, each one <strong>of</strong> the followers is<br />
producing:<br />
µ<br />
q q L , a − c − q <br />
L<br />
√ = √ F<br />
F<br />
which is independent from the strategy <strong>of</strong> the leader. Hence, the higher the<br />
production <strong>of</strong> the leader, the lower the number <strong>of</strong> entrants, while the production<br />
<strong>of</strong> each one <strong>of</strong> them will be the same. This would imply a constant level<br />
<strong>of</strong> total production Q = q L +(n − 1)q(q L ,n)=a − c − √ F , <strong>and</strong> a constant<br />
price p = a − Q = c + √ F , which would be equivalent to the equilibrium<br />
price emerging under Marshall competition.<br />
After having derived the behavior <strong>of</strong> the followers, it is now time to move<br />
to the first stage <strong>and</strong> examine the behavior <strong>of</strong> the leader. First <strong>of</strong> all, let us<br />
remind ourselves that entry takes place only for a production level which is<br />
not too high. If this is the case, the pr<strong>of</strong>its <strong>of</strong> the leader must be:<br />
π L = pq L − cq L − F = q L<br />
√<br />
F − F if qL
14 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
The pr<strong>of</strong>its <strong>of</strong> the leader are then:<br />
π L =2 √ ³<br />
F a − c − 2 √ ´<br />
F − F (1.17)<br />
One way to look at this result is by considering the role <strong>of</strong> the fixed cost <strong>of</strong><br />
production. When this is zero, we are in the st<strong>and</strong>ard neoclassical situation<br />
where perfect competition takes place: the number <strong>of</strong> firmsisindeterminate<br />
<strong>and</strong> the price must be equal to the marginal cost. However, whenever there<br />
is a small but positive fixed cost <strong>of</strong> production, the leader finds it optimal<br />
to produce enough to deter entry. 13 Constant returns to scale (holding for<br />
F =0) are not an minor approximation: a small departure from them leads<br />
to a radical change in the market structure. And when the fixed costs <strong>of</strong><br />
production are high, the leader is able to obtain substantial pr<strong>of</strong>its. 14<br />
Another way to look at the result is to imagine that there are some potential<br />
entrants <strong>and</strong> we can establish a relation between their number <strong>and</strong><br />
the market equilibrium: when the number <strong>of</strong> potential entrants is low enough<br />
(<strong>and</strong> the free entry condition is not binding) the market is characterized by<br />
all these firms being active. When there are many potential entrants (<strong>and</strong><br />
entry is endogenized) there is just one firm in equilibrium, the leader. Furthermore,<br />
it is interesting to compare the free entry equilibrium with <strong>and</strong><br />
without a leader. In the Stackelberg equilibrium with endogenous entry the<br />
limit price is higher than the equilibrium price in the Marshall equilibrium<br />
(the mark up p − c is doubled from √ F to 2 √ F ), consequently the consumer<br />
surplus is reduced. However, welfare as the sum <strong>of</strong> consumer surplus <strong>and</strong><br />
pr<strong>of</strong>its is higher in the Stackelberg equilibrium with endogenous entry than<br />
in the Marshall equilibrium. 15<br />
13 This form <strong>of</strong> entry deterrence is radically different from that emerging in the<br />
contestable markets theory <strong>of</strong> Baumol et al. (1982). First, they focused on price<br />
competition, which led to a limit price assigning zero pr<strong>of</strong>its to the leader, while<br />
our model <strong>of</strong> quantity competition leads to a limit price assigning positive pr<strong>of</strong>its<br />
to the leader. Second, their equilibrium was the same with exogenous or endogenous<br />
entry, while the role <strong>of</strong> the costs <strong>of</strong> production in endogenizing entry is<br />
crucial in our model. In the Appendix we will discuss how to endogenize the<br />
fixed costs.<br />
14 For instance, imagine that fixed costs are F =(a − c) 2 /25. Then the pr<strong>of</strong>its <strong>of</strong> a<br />
leader facing endogenous entry can be calculated as π L =(a − c) 2 /5. Compare<br />
these to the pr<strong>of</strong>its <strong>of</strong> a monopolist in the same market: its pr<strong>of</strong>its would be<br />
π M =(a−c) 2 /4−F = 21(a−c) 2 /100. It can be easily verified that the difference<br />
betweenthetwoislessthan5%.<br />
15 Welfare can be now calculated as:<br />
W S = Q2<br />
2 + π (a − c)2<br />
L = − 3F<br />
2<br />
It can be verified that welfare is higher under Stackelberg competition with endogenous<br />
entry for any F
1.2 Increasing Marginal Costs <strong>and</strong> Product Differentiation 15<br />
The extreme result on entry deterrence that we have just found holds<br />
under more general conditions. For instance, as we will see in Chapter 3,<br />
as long as goods are perfect substitutes, any kind <strong>of</strong> dem<strong>and</strong> function will<br />
generate entry deterrence by the leader when entry <strong>of</strong> followers is endogenous.<br />
However, when the cost function departs from the linear version (that we used<br />
until now) <strong>and</strong> when imperfect substitutability between goods is introduced,<br />
entry deterrence may not be the optimal strategy anymore. Nevertheless, the<br />
leader will still play in a very aggressive way, producing always more than<br />
the followers when their entry is endogenous. To show this we will now turn<br />
to two related extensions <strong>of</strong> the basic model.<br />
1.2 Increasing Marginal Costs <strong>and</strong> Product<br />
Differentiation<br />
The example adopted until now was extremely simple <strong>and</strong> stylized. Perfectly<br />
homogenous goods <strong>and</strong> marginal costs <strong>of</strong> production that are always constant<br />
are quite unrealistic features for many sectors. Most traditional markets are<br />
characterized by more complex shapes <strong>of</strong> the cost function <strong>and</strong> by substantial<br />
differentiation between products. Consider the market for cars. Companies<br />
like GM, Ford, Toyota, Nissan, VW, Porsche, Renault or FIAT <strong>of</strong>fer many<br />
different models, sometimes under different br<strong>and</strong>s (for instance Alfa Romeo,<br />
Lancia, Maserati <strong>and</strong> Ferrari for FIAT), <strong>and</strong> always in multiple versions by<br />
engine size, color, varieties <strong>of</strong> optional tools, <strong>and</strong> so on: each product appeals<br />
toadifferent class <strong>of</strong> customers <strong>and</strong> is sold at a different price. Moreover,<br />
the production <strong>of</strong> each model has not a constant unitary cost: on one side,<br />
economies <strong>of</strong> scale can be reached at the plant level through large production,<br />
on the other, larger output levels may require additional investments<br />
in plants, employees, <strong>and</strong> other inputs. Generally speaking, for each model<br />
there is a level <strong>of</strong> production that minimizes average costs, <strong>and</strong> average costs<br />
have a U shape around this efficient level.<br />
The simple model <strong>of</strong> competition in quantities studied in the previous<br />
section can be easily extended to take these realistic dimensions into account.<br />
For simplicity, we will consider the two issues separately. First, we will depart<br />
from the assumption <strong>of</strong> constant marginal costs assuming a U-shaped cost<br />
function, <strong>and</strong> then we will depart from the assumption <strong>of</strong> homogenous goods<br />
introducing imperfect substitutability between goods.<br />
assumption F
16 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
1.2.1 U-shaped Cost Functions<br />
In many markets, marginal costs <strong>of</strong> production are increasing at least beyond<br />
a certain level <strong>of</strong> output. Jointly with the presence <strong>of</strong> fixed costs <strong>of</strong> production,<br />
this leads to U-shaped average cost functions. Since technology <strong>of</strong>ten exhibits<br />
this pattern, it is important to analyze this case, <strong>and</strong> we will do it assuming<br />
a simple quadratic cost function.<br />
In particular, the general pr<strong>of</strong>it forfirm i becomes:<br />
π i = q i<br />
⎛<br />
⎝a − q i −<br />
nX<br />
j=1,j6=i<br />
q j<br />
⎞<br />
⎠ − dq2 i<br />
2 − F (1.18)<br />
where d>0 represents the degree <strong>of</strong> convexity <strong>of</strong> the cost function. When<br />
d =0we are back to the case <strong>of</strong> a constant marginal cost (zero in such a<br />
case). When d>0 the average cost function is U-shaped. One can easily<br />
verify that the marginal cost is increasing <strong>and</strong> convex, <strong>and</strong> it crosses the<br />
average total cost at its bottom, that is at the efficient scale <strong>of</strong> production:<br />
the one that minimizes average costs. This efficient scale <strong>of</strong> production can<br />
be derived formally as:<br />
µ dq<br />
ˆq =argmin<br />
2 + F r<br />
2F<br />
=<br />
q d<br />
Let us look now at the different forms <strong>of</strong> competition. Our four main<br />
equilibria can be derived as before. In particular, Nash competition would<br />
generate the individual output:<br />
a<br />
q(n) =<br />
(1.19)<br />
n + d +1<br />
for each firm. 16 Under Marshall competition each firm would produce:<br />
r<br />
2F<br />
q = < ˆq (1.20)<br />
2+d<br />
with a number <strong>of</strong> firms approximated by:<br />
r<br />
2+d<br />
n = a<br />
2F − d − 1<br />
Notice that the equilibrium production level is below the cost minimizing<br />
level. This is not surprising since imperfect competition requires a price above<br />
16 Amir (2005) shows that in this case industry pr<strong>of</strong>its have an inverse U shape with<br />
amaximumforn =1+2d, while welfare always decreases with n. He generalizes<br />
this dimension in a number <strong>of</strong> ways <strong>and</strong> shows that with strong scale economies<br />
(d
1.2 Increasing Marginal Costs <strong>and</strong> Product Differentiation 17<br />
marginal cost <strong>and</strong> free entry requires a price equal to the average cost <strong>and</strong><br />
above the marginal cost. Since the average cost is always decreasing when it<br />
is higher than the marginal cost, it must be that individual output is smaller<br />
than the efficient scale (von Weizsäcker, 1980).<br />
Under Stackelberg competition, the leader produces:<br />
a(1 + d)<br />
q L (n) =<br />
[2(1 + d)+d(n + d)]<br />
<strong>and</strong> each follower produces:<br />
(1.21)<br />
a [1 + d + d(n + d)]<br />
q(n) =<br />
(1.22)<br />
[2(1 + d)+d(n + d)] (n + d)<br />
Notice that, contrary to the basic linear case, here the leader produces less<br />
than a pure monopolist <strong>and</strong> its production diminishes with the number <strong>of</strong><br />
entrants.<br />
Finally, consider Stackelberg competition with endogenous entry (Etro,<br />
2008). In the last stage an entrant chooses q(q L ,n)=(a − q L )/(n + d), but<br />
the zero pr<strong>of</strong>it condition for the followers delivers a number <strong>of</strong> firms:<br />
Ãr !<br />
2+d<br />
n =(a − q L )<br />
− d<br />
2F<br />
<strong>and</strong> each entrant produces:<br />
r<br />
2F<br />
q =<br />
(1.23)<br />
2+d<br />
which is the same output as with Marshall competition. Of course this<br />
happens when there is effective entry, that is when n ≥ 2 or q L < a −<br />
(2 + d) p 2F/(2 + d). In such a case, total production is Q = a − (1 +<br />
d) p 2F/(2 + d), <strong>and</strong> the price becomes:<br />
r<br />
2F<br />
p =(1+d)<br />
2+d<br />
Both total production <strong>and</strong> the equilibrium price are independent from the<br />
leader’s production. The gross pr<strong>of</strong>it function <strong>of</strong> the leader in the first stage<br />
can be derived as:<br />
π L = pq L − d 2 q2 L − F =<br />
r<br />
2F<br />
=(1+d)<br />
2+d q L − d 2 q2 L − F<br />
which is concave in q L .Aslongasd is large enough, we have an interior<br />
optimum <strong>and</strong> in equilibrium the leader allows other firms to enter in the<br />
market <strong>and</strong> produces:
18 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
q L = 1+d<br />
d<br />
r<br />
2F<br />
> ˆq (1.24)<br />
2+d<br />
Notice that the leader is applying a simple pricing rule which equates the<br />
price derived above p =(1+d) p 2F/(2 + d) to the marginal cost, which is<br />
dq L in this model. Of course, the leader can price at the marginal cost <strong>and</strong><br />
obtain positive pr<strong>of</strong>its because its marginal cost <strong>of</strong> production is above its<br />
average cost. This can only happen in the region where the average total costs<br />
are increasing, which implies a production for the leader above the efficient<br />
scale.<br />
Finally, the equilibrium number <strong>of</strong> firms is:<br />
r µ 2+d 1+d<br />
n = a<br />
2F − d<br />
<br />
+ d<br />
Total output <strong>and</strong> price are the same as in the Marshall equilibrium, therefore<br />
the consumer surplus is unchanged, but welfare must be higher since the<br />
leader makes positive pr<strong>of</strong>its. 17<br />
Notice that the leader is producing always more than each follower. While<br />
the followers produce below the efficient scale, the leader produces above<br />
the efficient scale. The intuition is as follows. Followers have to produce at<br />
a price where their marginal revenue equates their marginal cost, <strong>and</strong> free<br />
entry implies that the price has to be equal to the average cost. But marginal<br />
<strong>and</strong> average costs are the same at the efficient scale, therefore the followers<br />
must be producing below this efficient scale. Now, since the equilibrium price<br />
is determined by the endogenous entry condition, it represents the perceived<br />
marginal revenue for the leader, <strong>and</strong> the leader must produce where this<br />
perceived marginal revenue equates the marginal cost, which in this case<br />
must be above the efficient scale for pr<strong>of</strong>its to be positive.<br />
1.2.2 Product Differentiation<br />
We now move to another simple extension <strong>of</strong> the basic linear model introducing<br />
product differentiation <strong>and</strong> imperfect substitutability between the goods<br />
supplied by the firms. We retain the initial assumptions <strong>of</strong> constant marginal<br />
costs <strong>and</strong> competition in quantities.<br />
For simplicity, consider the inverse dem<strong>and</strong> function for firm i:<br />
17 In general, the pr<strong>of</strong>it <strong>of</strong> the leader in case <strong>of</strong> an interior solution is:<br />
π L =<br />
F<br />
d(2 + d) > 0<br />
In the alternative case <strong>of</strong> entry deterrence, the leader produces q L = a − (2 +<br />
d) 2F/(2 + d). Thepr<strong>of</strong>its <strong>of</strong> the leader are larger under entry deterrence when<br />
d is low enough or F is high enough.
p i = a − q i − b X j6=i<br />
1.2 Increasing Marginal Costs <strong>and</strong> Product Differentiation 19<br />
q j (1.25)<br />
where b ∈ (0, 1] is an index <strong>of</strong> substitutability between goods. Of course, for<br />
b =0goods are perfectly independent <strong>and</strong> each firmsellsitsowngoodas<br />
a pure monopolist, while for b =1we are back to the case <strong>of</strong> homogeneous<br />
goods. In this more general framework the pr<strong>of</strong>it functionforfirm i is:<br />
π i = q i<br />
⎛<br />
⎝a − q i − b<br />
nX<br />
j=1,j6=i<br />
q j<br />
⎞<br />
⎠ − cq i − F (1.26)<br />
The four main equilibria can be derived as usual. In particular a Nash equilibrium<br />
would generate the individual output:<br />
a − c<br />
q(n) =<br />
2+b(n − 1)<br />
for each firm. In the Marshall equilibrium each firm would produce:<br />
(1.27)<br />
q = √ F (1.28)<br />
with a number <strong>of</strong> firms:<br />
n =1+ a − c<br />
b √ F − 2 b<br />
Under Stackelberg competition, the leader produces:<br />
(a − c)(2 − b)<br />
q L =<br />
2<br />
<strong>and</strong> each follower produces:<br />
(1.29)<br />
(a − c)[2 − b(2 − b)]<br />
q(n) = (1.30)<br />
2[2 + b(n − 2)]<br />
Finally, consider Stackelberg competition with endogenous entry. As long as<br />
substitutability between goods is limited enough (b is small) there are entrants<br />
producing q(q L ,n)=(a − bq L − c)/[2 + b(n − 2)]. Setting their pr<strong>of</strong>its equal<br />
to zero, the endogenous number <strong>of</strong> firms results in:<br />
n =2+ a − bq L − c<br />
b √ − 2<br />
F b<br />
implying once again a constant production:<br />
q = √ F (1.31)<br />
for each follower. Plugging everything into the pr<strong>of</strong>it function <strong>of</strong> the leader,<br />
we have:
20 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
π L = q L [a − q L − b(n − 1)q] − cq L − F =<br />
= q L<br />
h<br />
(2 − b) √ F − (1 − b)q L<br />
i<br />
− F<br />
that is maximized when the leader produces:<br />
q L =<br />
2 − b √<br />
F (1.32)<br />
2(1 − b)<br />
which is always higher than the production <strong>of</strong> the followers. This strategy<br />
leaves space to the endogenous entry <strong>of</strong> firms so that the total number <strong>of</strong><br />
firms in the market is:<br />
n =2+ a − c<br />
b √ F − 2 b − 2 − b<br />
2(1 − b)<br />
Notice that the leader will <strong>of</strong>fer its good at a lower price than the followers,<br />
namely:<br />
µ<br />
p L = c + 1 −<br />
2 b √F 0<br />
Again, this outcome emerges only if the degree <strong>of</strong> product differentiation is high<br />
enough. In the alternative case <strong>of</strong> entry deterrence, the production <strong>of</strong> the leader<br />
is q L =(a − c − 2 √ F )/b <strong>and</strong> the limit price is p L =[c − (1 − b)a +2 √ F ]/b. Entry<br />
deterrence is optimal for b or F large enough.
1.3 A Simple Model <strong>of</strong> <strong>Competition</strong> in Prices 21<br />
<strong>of</strong> contestable markets associated with Baumol et al. (1982) shows that a<br />
single firm sets the price at a market clearing level which equates the average<br />
total costs <strong>and</strong> obtains zero pr<strong>of</strong>its again. With U-shaped cost functions, the<br />
Bertr<strong>and</strong> equilibrium boils down to a price equal to the minimum average<br />
cost for each firm, since any different strategy either would leave space for<br />
pr<strong>of</strong>itable deviations, or would lead to losses. 19 Things are not that simple<br />
when products are differentiated, the case to which we now turn.<br />
<strong>Competition</strong> in prices is crucial in markets where the products are highly<br />
differentiated. In this case, as we have already seen in the last section, each<br />
firm has a limited market power because it supplies a unique product which<br />
is only partially substitutable with the products <strong>of</strong> the other firms. Think <strong>of</strong><br />
the fashion market, which is characterized by strong product differentiation,<br />
segmentation depending on the target customers, <strong>and</strong> competition in prices.<br />
Established luxury br<strong>and</strong>s as Armani, Versace, D&G, Gucci, Etro, Yves Saint<br />
Laurent, Louis Vuitton <strong>and</strong> others <strong>of</strong>fer different sophisticated clothes at predetermined<br />
prices in every season. Other companies which target wider markets,<br />
as Gap, Abercrombie, Benetton, Zara, H&M <strong>and</strong> so on, provide largely<br />
differentiated products <strong>and</strong> engage in analogous or even stronger forms <strong>of</strong><br />
price competition. 20 In this section, we will focus on the peculiarities <strong>of</strong> similar<br />
markets where goods are imperfect substitutes <strong>and</strong> firms choose their<br />
prices.<br />
In this introductory analysis <strong>of</strong> price competition, we will employ a model<br />
based on a simple form <strong>of</strong> the dem<strong>and</strong> function, the so-called Logit dem<strong>and</strong>.<br />
This is particularly interesting because it is simple but flexible enough to<br />
depict real world dem<strong>and</strong> functions: not by chance it is widely used in econometric<br />
studies 21 to estimate dem<strong>and</strong> in various industries <strong>and</strong> in marketing<br />
analysis. 22<br />
The simplest form <strong>of</strong> the Logit dem<strong>and</strong> is:<br />
D i =<br />
Ne −λp i<br />
h Pn<br />
j=1 e−λpj i (1.33)<br />
where <strong>of</strong> course p i is the price <strong>of</strong> firm i, λ>0 is a parameter governing the<br />
slope <strong>of</strong> the dem<strong>and</strong> function, <strong>and</strong> N is a scale factor that can be thought <strong>of</strong><br />
19 It is immediate to verify that these equilibria correspond to a Stackelberg equilibrium<br />
in prices with endogenous entry in the case <strong>of</strong> homogenous goods. Therefore,<br />
the theory <strong>of</strong> Stackelberg competition with endogenous entry can be seen as a<br />
generalization <strong>of</strong> the theory <strong>of</strong> contestable markets to product differentiation,<br />
<strong>and</strong>tootherforms<strong>of</strong>competition.<br />
20 For a recent analysis <strong>of</strong> the fashion industry see Dallocchio et al. (2006).<br />
21 See McFadden (1974).<br />
22 The classic reference on product differentiation <strong>and</strong> price competition is Anderson<br />
et al. (1992). See also Anderson <strong>and</strong> de Palma (1992) for the first analysis<br />
<strong>of</strong> Nash <strong>and</strong> Marshall equilibria within the Logit model <strong>of</strong> price competition.
22 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
as the total income or the total number <strong>of</strong> agents expressing this aggregate<br />
dem<strong>and</strong>. Since we focus on substitute goods, such a dem<strong>and</strong> for firm i is<br />
decreasing in the price <strong>of</strong> the same firm i <strong>and</strong> increasing in the price <strong>of</strong> any<br />
other firm j. The general pr<strong>of</strong>it functionforafirm facing this dem<strong>and</strong> <strong>and</strong><br />
producing with a constant marginal cost c <strong>and</strong> a fixed cost F 0. This important property,<br />
which holds virtually in all realistic models <strong>of</strong> competition in prices, suggests<br />
that a higher price by one firm induces other firms to increase their prices<br />
as well. In other words, an accommodating behavior <strong>of</strong> one firm leads other<br />
firms to be accommodating too.<br />
To conclude our analysis <strong>of</strong> the Nash equilibrium, notice that in a symmetric<br />
situation with a price p for each firm, dem<strong>and</strong> boils down to D = N/n<br />
<strong>and</strong> the equilibrium price is decreasing in the number <strong>of</strong> firms:<br />
p(n) =c +<br />
1<br />
λ (1 − 1/n)<br />
(1.35)<br />
In a Marshall equilibrium one can easily derive that the number <strong>of</strong> active<br />
firms is:<br />
n =1+ N λF<br />
<strong>and</strong> each one <strong>of</strong> these sells its product at the price:<br />
p = c + 1 λ + F N<br />
(1.36)<br />
Let us now move to models <strong>of</strong> price leadership. Of course it can be even<br />
harder for a firm to commit to a price rather than to a different strategy (as
1.3 A Simple Model <strong>of</strong> <strong>Competition</strong> in Prices 23<br />
the quantity <strong>of</strong> production). However, price commitments can be reasonable<br />
in the short run (for instance in seasonal markets), or when there are small<br />
menu costs <strong>of</strong> changing prices or it is costly to acquire the information needed<br />
to reoptimize on the price choice. In the next chapter we will deal with the<br />
commitment problem in a deeper way <strong>and</strong> we will suggest that there are<br />
realistic ways in which a strategic investment can be a good substitute for a<br />
commitment to a strategy. For now we will assume that a firm can simply<br />
commit to a pricing strategy <strong>and</strong> analyze the consequence <strong>of</strong> this.<br />
Concerning the Stackelberg equilibrium we do not have analytical solutions.<br />
However, it is important to underst<strong>and</strong> the nature <strong>of</strong> the incentives<br />
<strong>of</strong> the firms, which is now rather different from the model with competition<br />
in quantities. Here the leader is aware that an increase in its own price will<br />
lead the followers to increase their prices, which exerts a positive effect on<br />
the pr<strong>of</strong>its <strong>of</strong> the leader. Accordingly, the commitment possibility is generally<br />
used adopting an accommodating strategy: the leader chooses a high price to<br />
induce its followers to choose high prices as well. 23 The only case in which<br />
this does not happen is when the fixed costs <strong>of</strong> production are high enough<br />
<strong>and</strong> the leader finds it better to deter entry. This can only be done adopting a<br />
low enough price: therefore the leader can be aggressive only for exclusionary<br />
purposes.<br />
This st<strong>and</strong>ard result emphasizes a possible inconsistency within the model<br />
<strong>of</strong> price leadership, at least when applied to describe real markets. We have<br />
suggested that leaders are accommodating when the fixed costs <strong>of</strong> production<br />
(or entry) are small, because in such a case an exclusionary strategy would<br />
require to set a very low price <strong>and</strong> would be too costly. But these are exactly<br />
the conditions under which other firms may want to enter in the market: fixed<br />
costs are low <strong>and</strong> exclusionary strategies by incumbents are costly. Therefore,<br />
the assumption that the number <strong>of</strong> firms (<strong>and</strong> in particular <strong>of</strong> the number <strong>of</strong><br />
followers) is fixed becomes quite unrealistic.<br />
Let us look at the Stackelberg equilibrium with endogenous entry. The<br />
solution in this case is slightly more complex, but it can be fully characterized.<br />
First <strong>of</strong> all, as usual, let us look at the stage in which the leader has already<br />
chosen its price p L <strong>and</strong> the followers enter <strong>and</strong> choose their prices. As before,<br />
their choice will follow the rule:<br />
1<br />
p i = c +<br />
λ(1 − D i /N )<br />
where the dem<strong>and</strong> on the right h<strong>and</strong> side depends on the price <strong>of</strong> the leader<br />
<strong>and</strong> all the other prices as well. However, under free entry we must have also<br />
that the markup <strong>of</strong> the followers exactly covers the fixed cost <strong>of</strong> production:<br />
D i (p i − c) =F<br />
23 Nevertheless, the followers will have incentives to choose a lower price than the<br />
leader, <strong>and</strong> each one <strong>of</strong> them will then have a larger dem<strong>and</strong> <strong>and</strong> pr<strong>of</strong>its than the<br />
leader: there is a second-mover advantage rather than a first-mover advantage.
24 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
If the price <strong>of</strong> the leader is not too low or the fixed cost not to high, there<br />
is indeed entry in equilibrium <strong>and</strong> we can solve these two equations for the<br />
dem<strong>and</strong> <strong>of</strong> the followers <strong>and</strong> their prices in the symmetric equilibrium:<br />
p = c + 1 λ + F N , D = λF N<br />
N + λF<br />
(1.37)<br />
Notice that neither one or the other endogenous factors depend on the price<br />
chosen by the leader. Therefore, it must be that the strategy <strong>of</strong> the leader is<br />
going to affect only the number <strong>of</strong> followers entering in equilibrium, but not<br />
their prices or their equilibrium production.<br />
The leader is going to perceive this because its dem<strong>and</strong> can now be calculated<br />
as:<br />
D L =<br />
Ne−λp L<br />
P n<br />
i=1 e−λp j = Deλ(p−p L)<br />
Since neither p or D depend on the price <strong>of</strong> the leader as we have seen before,<br />
the perceived dem<strong>and</strong> by the leader is a simple function <strong>of</strong> its own price, <strong>and</strong><br />
its pr<strong>of</strong>its can be derived as:<br />
π L =(p L − c)D L − F =<br />
=(p L − c)De λ(p−pL) − F<br />
where we could use our previous results to substitute for p or D. Pr<strong>of</strong>it<br />
maximization by the leader provides its equilibrium price:<br />
p L = c + 1 λ 0<br />
which is positive under the assumption that F
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong> 25<br />
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong><br />
The last example we are going to consider introduces us to a topic that<br />
we will encounter later on in the book in Chapter 4, the competition to<br />
innovate <strong>and</strong> therefore conquer a market with new or better products. In<br />
many high-tech sectors, this is becoming a main form <strong>of</strong> competition, since<br />
the life <strong>of</strong> a product is quite short <strong>and</strong> R&D investment strategies to conquer<br />
future markets are much more important than price or production strategies.<br />
Consider the pharmaceutical sector: in this market companies like Pfizer,<br />
Bayer, Merck, H<strong>of</strong>fmann-La Roche, GlaxoSmithKline <strong>and</strong> many others invest<br />
a lot in R&D to develop, test <strong>and</strong> patent new drugs, while price competition<br />
over unpatented drugs plays a minor role. 25<br />
<strong>Competition</strong> for the market works as a sort <strong>of</strong> contest. Firms invest to<br />
innovate <strong>and</strong> to win the contest. It may be that the first innovator obtains a<br />
patent on the invention <strong>and</strong> exploits monopolistic pr<strong>of</strong>its for a while on its<br />
innovation. It may be that the same innovator just keeps it secret <strong>and</strong> exploits<br />
its leadership on the market until an imitator replaces it. In both ways the<br />
expectedgainfromaninnovationiswhatdrivesfirms to invest in R&D. In<br />
this framework we can also study alternative market structures depending on<br />
the timing <strong>of</strong> moves <strong>and</strong> on the entry conditions.<br />
Consider a simple contest between firms to obtain a drastic innovation<br />
which has an expected value V ∈ (0, 1) for the winner <strong>and</strong> generates no<br />
gains for the losers. Each contestant i invests resources z i ∈ [0, 1) to win<br />
the contest. This investment has a cost <strong>and</strong>, for simplicity, we will assume<br />
that the cost is quadratic, that is zi 2 /2. The investment provides the contestant<br />
with the probability z i to innovate. The innovator wins the contest if<br />
no other contestant innovates, for instance because in the case <strong>of</strong> multiple<br />
winners competition between them would drive pr<strong>of</strong>its<br />
Q<br />
away. Accordingly,<br />
n<br />
the probability to win the contest is Pr(i wins)=z i j=1,j6=i [1 − z j] ,that<br />
is its probability to innovate multiplied by the probability that no one else<br />
innovates. In conclusion, the general pr<strong>of</strong>it functionis: 26<br />
π i = z i<br />
n<br />
Y<br />
j=1,j6=i<br />
[1 − z j ] V − z2 i<br />
2 − F (1.39)<br />
Consider the Nash equilibrium. The first order condition for the optimal<br />
investment by a firm i is:<br />
25 See Sutton (1998, Ch. 8) for a description <strong>of</strong> competition for the market in the<br />
pharmaceutical industry.<br />
26 We adopt a more restrictive assumption, V ∈ ( √ 2F,1). This guarantees pr<strong>of</strong>itable<br />
entry for at least one firm. Indeed, a single firm would invest z = V < 1<br />
expecting π = V 2 /2 − F>0. Hence,investingz =1<strong>and</strong> innovating for sure can<br />
be pr<strong>of</strong>itable, but it is not optimal.
26 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
z i =<br />
nY<br />
j=1,j6=i<br />
[1 − z j ] V<br />
which shows that when the investment <strong>of</strong> a firm increases, the other firms<br />
have incentives to invest less: ∂z i /∂z j < 0. Eachfirm chooses its own investment<br />
without taking this externality into account, therefore competition<br />
for the market generates excessive investment from the firms point <strong>of</strong> view.<br />
Forinstance,inthecase<strong>of</strong>tw<strong>of</strong>irms, each one would invest z = V/(1 + V )<br />
in equilibrium, while collusion between them would reduce individual investment<br />
to the lower level ˜z = V/(1 + 2V ), which increases expected pr<strong>of</strong>its for<br />
each one <strong>of</strong> the two firms. 27 This suggests that a joint venture between firms<br />
competing for a market may end up reducing aggregate investment. However,<br />
notice that we cannot evaluate these outcomes from a welfare point <strong>of</strong> view<br />
without expliciting the social value <strong>of</strong> the innovation: if the social value <strong>of</strong><br />
the innovation is high enough, the investment is too low also in the Nash<br />
equilibrium, <strong>and</strong> R&D subsidies would be needed to restore social efficiency.<br />
Let us go back to the general case with n firms competing for the market.<br />
Now, the equilibrium investment is implicitly given by:<br />
z =(1− z) n−1 V<br />
In the Marshall equilibrium we must also take into account the endogenous<br />
entry condition:<br />
z(1 − z) n−1 V − z 2 /2=F<br />
<strong>and</strong> solving the system <strong>of</strong> the two conditions we have the number <strong>of</strong> agents:<br />
³<br />
log V/ √ ´<br />
2F<br />
n =1+ h<br />
log 1/(1 − √ i<br />
2F )<br />
<strong>and</strong> the investment:<br />
z = √ 2F (1.40)<br />
The investment <strong>of</strong> each firm increases with the size <strong>of</strong> the fixed cost <strong>of</strong> R&D,<br />
while entry decreases in the fixed cost <strong>and</strong> increases with the value <strong>of</strong> the<br />
innovation.<br />
27 Also in this case we can verify when collusion is sustainable by extending the<br />
model to an infinitely repeated game. Imagine that a deviation from collusion is<br />
punished with reversion to the Nash equilibrium. One can verify that the best<br />
deviation is z D =(1+V )V/(1 + 2V ), <strong>and</strong> collusion is sustainable for a discount<br />
factor δ>(1 + V ) 2 /(2 + 4V + V 2 ): more valuable innovations make it harder to<br />
sustain collusion.
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong> 27<br />
Consider now a Stackelberg equilibrium. As already noticed, when the<br />
investment by one firm is increased, the other firms have incentives to invest<br />
less: then in a Stackelberg equilibrium the leader exploits its first mover advantage<br />
by investing more than the followers, so as to reduce their investment<br />
<strong>and</strong> increase its relative probability <strong>of</strong> winning. For instance, in a Stackelberg<br />
duopoly the leader invests z L = V (1 − V )/(1 − 2V 2 ) <strong>and</strong> the follower invests<br />
z = V (1 − V − V 2 )/(1 − 2V 2 ).<br />
In a Stackelberg equilibrium with endogenous entry, as long as the investment<br />
<strong>of</strong> the leader z L is small enough to allow entry <strong>of</strong> at least one firm, the<br />
first order conditions <strong>and</strong> the free entry conditions are:<br />
(1 − z) n−2 (1 − z L )V = z, z(1 − z) n−2 (1 − z L )V = z 2 /2+F<br />
which deliver the same investment choice by each entrant as in the Marshall<br />
equilibrium, z = √ 2F , <strong>and</strong> the number or firms:<br />
h<br />
log (1 − z L )V/ √ i<br />
2F<br />
n(z L )=2+ h<br />
log 1/(1 − √ i<br />
2F )<br />
Putting these two equations together <strong>and</strong> substituting in the pr<strong>of</strong>it function<br />
<strong>of</strong> the leader, we would have:<br />
π L = z L (1 − z) n−1 V − z2 L<br />
2 − F =<br />
= z √ ³<br />
L 2F 1 − √ ´<br />
2F − z2 L<br />
1 − z L 2 − F (1.41)<br />
which has not an interior optimum: indeed, it is always optimal for the leader<br />
to deter entry investing enough. This requires a slightly higher investment<br />
than the one for which thehequilibrium number <strong>of</strong> firms would be n =2.<br />
Since n(z L )=2requires log (1 − z L )V/ √ i<br />
2F =0, we can conclude that the<br />
leader will invest:<br />
√<br />
2F<br />
¯z L =1−<br />
(1.42)<br />
V<br />
which is increasing in the value <strong>of</strong> innovations <strong>and</strong> decreasing in their fixed<br />
cost. Therefore, in a contest with a leader <strong>and</strong> free entry <strong>of</strong> participants, the<br />
leader invests enough to deter investment by the other firms <strong>and</strong> is the only<br />
possible winner <strong>of</strong> the contest.<br />
1.4.1 The Arrow’s Paradox<br />
Until now we investigated a form <strong>of</strong> competition for the market where all firms<br />
were at the same level. Often times, competition for the market is between an
28 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
incumbent leader that is already in the market with the leading edge technology<br />
(or with the best product) <strong>and</strong> outsiders trying to replace this leadership.<br />
In such a case the incentives to invest in innovation may be different <strong>and</strong> it is<br />
important to underst<strong>and</strong> how. Arrow (1962) was one <strong>of</strong> the first economists<br />
to examine this issue in a formal way. He found that incumbent monopolists<br />
have lower incentives than the outsiders to invest. His insight was simple but<br />
powerful: while the gains from an innovation for the incumbent monopolist<br />
are just the difference between pr<strong>of</strong>its obtained with the next innovation <strong>and</strong><br />
those obtained with the current one, the gains for any outsider are the full<br />
pr<strong>of</strong>its from the next innovation. Consequently, the incumbent has lower incentives<br />
to invest in R&D. The expected gains <strong>of</strong> the incumbent are even<br />
diminished when the number <strong>of</strong> outsiders increases. When the latter is high<br />
enough the incumbent has no more incentives to participate to the competition,<br />
<strong>and</strong>, in particular, when entry in the competition for the market is<br />
free, the incumbent does not invest at all. Such a strong theoretical result is<br />
<strong>of</strong> course too drastic to be realistic. Many technological leaders invest a lot<br />
in R&D, try to maintain their leadership, <strong>and</strong> they <strong>of</strong>ten manage: persistent<br />
leadership is not that unusual: for this reason the theoretical finding <strong>of</strong> Arrow<br />
is considered a paradoxical outcome, the “Arrow’s paradox” indeed. Before<br />
<strong>of</strong>fering a theoretical solution for this paradox, however, we will extend our<br />
model to include an asymmetry between an incumbent monopolist <strong>and</strong> the<br />
outsiders.<br />
Imagine a two period extension <strong>of</strong> the model. In the first period an incumbent<br />
monopolist can exploit its technology to obtain pr<strong>of</strong>its K ∈ (0,V].<br />
We can think <strong>of</strong> K as the rents associated with an initial leading technology<br />
or some other exogenous advantage. If these rents are constrained by a competitive<br />
fringe <strong>of</strong> firms, we can also think that an increase in the intensity <strong>of</strong><br />
competition reduces K. Inthefirst period any firm can invest to innovate <strong>and</strong><br />
conquer the gain V from the next innovation to be exploited in the second period.<br />
If no one innovates, the incumbent retains its pr<strong>of</strong>its K also in the second<br />
period. This happens with probability Pr(no innovation) = Q n<br />
j=1 [1 − z j].<br />
Then, assuming no discounting, the expected pr<strong>of</strong>its <strong>of</strong> the incumbent monopolist,<br />
that we now label with the index M, are:<br />
π M = K+z M<br />
n Y<br />
j=1,j6=M<br />
[1 − z j ] V +(1−z M )<br />
nY<br />
j=1,j6=M<br />
[1 − z j ] K− z2 M<br />
2<br />
−F (1.43)<br />
in case <strong>of</strong> positive investment in the contest, otherwise expected pr<strong>of</strong>its are<br />
givenonlybythecurrentpr<strong>of</strong>its plus the expected value <strong>of</strong> the current pr<strong>of</strong>its<br />
when no one innovates. The pr<strong>of</strong>its <strong>of</strong> the other firms are the same as previously.<br />
Before analyzing alternative forms <strong>of</strong> competition, notice that when the<br />
monopolist is assumed alone in the research activity, its optimal investment<br />
is z M = V − K. Hence, an incumbent monopolist (with K>0) haslower<br />
incentives to invest than a firm without current pr<strong>of</strong>its (with K =0): the<br />
so-called Arrow effect is in action. Moreover, if we think that the intensity
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong> 29<br />
<strong>of</strong> product market competition has a negative impact on the current pr<strong>of</strong>its<br />
K, while it has no impact on the value <strong>of</strong> the innovation (since this is drastic<br />
<strong>and</strong> the innovator will not be constrained by product market competitors), it<br />
clearly follows that an increase in the intensity <strong>of</strong> competition reduces K <strong>and</strong><br />
increases the investment <strong>of</strong> the monopolist <strong>and</strong> the probability <strong>of</strong> innovation<br />
z M .Aghion<strong>and</strong>Griffith (2005) put a lot <strong>of</strong> emphasis on this effect, which<br />
they label escape competition effect:<br />
“competition reduces pre-innovation rents...but not their post innovation<br />
rents since by innovating these firms have escaped the fringe.<br />
This,inturninducesthosefirms to innovate in order to escape competition<br />
with the fringe.” 28<br />
Now, consider a Nash equilibrium with a general number <strong>of</strong> firms. If the<br />
incumbent does not invest, the equilibrium is the same <strong>of</strong> the symmetric<br />
model, but the expected pr<strong>of</strong>it <strong>of</strong> the monopolist π M (z M ) must be:<br />
√ √<br />
2F (1 − 2F )K<br />
π M (0) = K +<br />
V<br />
which is increasing in K (decreasing in the intensity <strong>of</strong> competition) <strong>and</strong><br />
decreasing in the value <strong>of</strong> the innovation V (since this increases the incentives<br />
<strong>of</strong> other firms to innovate <strong>and</strong> replace the monopolist).<br />
If the monopolist is investing, however, the first order conditions for the<br />
monopolist <strong>and</strong> for the other firms in Nash equilibrium would be:<br />
z =(1− z) n−2 (1 − z M )V ,<br />
z M =(1− z) n−1 V − (1 − z) n−1 K<br />
which always imply a lower investment <strong>of</strong> the monopolist because <strong>of</strong> the<br />
Arrow effect. For instance, with two firms we have:<br />
z M =<br />
(1 − V )(V − K)<br />
1 − V (V − K)<br />
z =<br />
(1 − V )(V − K)+K<br />
1 − V (V − K)<br />
(1.44)<br />
It is interesting to verify what is the impact <strong>of</strong> an increase in the intensity <strong>of</strong><br />
product market competition, which lowers current pr<strong>of</strong>its K without affecting<br />
the value <strong>of</strong> the drastic innovation V : this increases the investment <strong>of</strong> the<br />
incumbent according to the escape competition effect, but it decreases the<br />
investment <strong>of</strong> the outsider. 29<br />
28 See Aghion <strong>and</strong> Griffith (2005, pp. 55-56). An increase <strong>of</strong> the intensity <strong>of</strong> competition<br />
is associated with a lower price <strong>of</strong> the competitive fringe or with a higher<br />
probability <strong>of</strong> entry <strong>of</strong> equally efficient firms.<br />
29 Of course, the Arrow effect could be counterbalanced if we introduced a technological<br />
advantage for the incumbent (Barro <strong>and</strong> Sala-i-Martin, 1995) or absorptive<br />
capacity <strong>of</strong> the incumbent (Wiethaus, 2006,a,b), that is the ability to<br />
imitate the innovation <strong>of</strong> an outsider.
30 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
When entry <strong>of</strong> firms is free, investors enter as long as the expected pr<strong>of</strong>its<br />
are positive, that is until the following zero pr<strong>of</strong>it condition holds:<br />
z(1 − z M )(1 − z) n−2 V = z 2 /2+F<br />
This implies that each one <strong>of</strong> the other firms invests again z = √ 2F ,<strong>and</strong><br />
we will now show that the incumbent monopolist prefers to withdraw from<br />
the contest <strong>and</strong> not invest in R&D. To see this, notice that the monopolist<br />
should invest less than the other firms, according to its optimality condition:<br />
z M (1 − z M )= √ 2F (1 − √ 2F )(V − K)/V<br />
This implies that the optimal investment <strong>of</strong> the monopolist should decrease<br />
with K: fromthesamelevelasfortheotherfirms z M = √ 2F when K =0<br />
toward zero investment z M =0when approaching K = V .Thepr<strong>of</strong>its <strong>of</strong> the<br />
monopolist in case <strong>of</strong> positive investment would be:<br />
√ √ ∙ ¸<br />
2F (1 − 2F ) (V − K)zM + K<br />
π M (z M )=K +<br />
− z2 M<br />
V<br />
1 − z M 2 − F<br />
where z M should be at its optimal level derived above. Notice that for K =0<br />
these expected pr<strong>of</strong>its are −F ,sothemonopolistprefersnottoinvestatall,<br />
<strong>and</strong> for K = V the expected pr<strong>of</strong>its tend to K + √ 2F (1− √ 2F )−F ,whichis<br />
again lower than the expected pr<strong>of</strong>its in case the monopolist does not invest<br />
at all. It can be verified that this is always the case for any K ∈ (0,V), 30<br />
hence the monopolist always prefers not to invest <strong>and</strong> decides to give up to<br />
any chances <strong>of</strong> innovation.<br />
Finally, notice that the escape competition effect disappears: an increase<br />
in the intensity <strong>of</strong> competition does not affect the investment <strong>of</strong> any firm or<br />
the aggregate probability <strong>of</strong> innovation. Perfect competition for the market<br />
eliminates any impact <strong>of</strong> competition in the market on the investment in<br />
innovation. 31<br />
In this simple example, the lack <strong>of</strong> incentives to invest for the monopolist<br />
emerges quite clearly. On the basis <strong>of</strong> this theoretical result, it is <strong>of</strong>ten<br />
claimed that monopolistic markets or markets with a clear leadership are<br />
less innovative. In a neat article on this topic appeared on the The Economist<br />
(2004, “Slackers or Pace-setters? Monopolies may have more incentives<br />
to innovate than economists have thought”, Economic Focus, May 22) this<br />
issue has been explained quite clearly:<br />
30 This immediate after comparing pr<strong>of</strong>its for the monopolist in case <strong>of</strong> zero <strong>and</strong><br />
positive investment in Nash equilibrium as functions <strong>of</strong> K.<br />
31 Not by chance, Aghion <strong>and</strong> Griffith (2005) obtained the escape competition effect<br />
in a model where the incumbent is exogenously the only investor. In the next<br />
section we present a model where the incumbent is endogenously the only investor<br />
to verify that both the Arrow effect <strong>and</strong> the escape competition effect disappear<br />
in that case.
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong> 31<br />
“By <strong>and</strong> large, <strong>of</strong>ficialdom these days continues to take a dim<br />
view <strong>of</strong> monopoly. <strong>Antitrust</strong> authorities in many countries do not<br />
shrink from picking fights with companies that they believe are too<br />
powerful. The biggest target in recent years, first in America <strong>and</strong><br />
now in Europe, has been Micros<strong>of</strong>t, creator <strong>of</strong> the operating system<br />
that runs on some 95% <strong>of</strong> the world’s personal computers. One <strong>of</strong><br />
the arguments against Micros<strong>of</strong>t is that its dominance <strong>of</strong> the desktop<br />
allows it to squeeze out smaller <strong>and</strong> (say the company’s critics) more<br />
innovative rivals.<br />
Despite this, compelling evidence that monopolists stifle innovation<br />
is harder to come by than simple theory suggests. Joseph<br />
Schumpeter, an Austrian economist, pointed out many years ago<br />
that established firms play a big role in innovation. In modern times,<br />
it appears that many product innovations, in industries from razor<br />
blades to s<strong>of</strong>tware, are made by companies that have a dominant<br />
share <strong>of</strong> the market. Most mainstream economists, however, have<br />
had difficulty explaining why this might be so. Kenneth Arrow, a<br />
Nobel prize-winner, once posed the issue as a paradox. Economic<br />
theory says that a monopolist should have far less incentive to invest<br />
in creating innovations than a firm in a competitive environment:<br />
experience suggests otherwise. How can this be so?<br />
One possibility might be that the empirical connection between<br />
market share <strong>and</strong> innovation is spurious: might big firms innovate<br />
more simply because they are big, not because they are dominant?<br />
Apaper 32 published a few years ago by Richard Blundell, Rachel<br />
Griffith <strong>and</strong> John Van Reenen, <strong>of</strong> Britain’s Institute for Fiscal Studies,<br />
did much to resolve this empirical question. In a detailed analysis<br />
<strong>of</strong> British manufacturing firms, it found that higher market shares<br />
do go with higher investment in research <strong>and</strong> development, which in<br />
turn is likely to lead to greater innovation. Still, the question remains:<br />
why does it happen?”<br />
We now turn to this theoretical issue.<br />
1.4.2 <strong>Innovation</strong> by Leaders<br />
In this section we will study innovation contests where a firm can act as a<br />
leader <strong>and</strong> commit to an investment level before the other firms (Etro, 2004).<br />
It is reasonable to imagine that the firm able to commit to an investment<br />
in R&D before the others is the same incumbent monopolist that has the<br />
leading edge technology. This will be our assumption.<br />
Consider Stackelberg competition where the incumbent monopolist is the<br />
first mover. The symmetric reaction <strong>of</strong> the other firms to the investment <strong>of</strong><br />
32 Blundell et al. (1999).
32 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
the leader is still governed by their equilibrium first order condition z =(1−<br />
z) n−2 (1−z L )V , where now z L is the investment <strong>of</strong> the leader, which is known<br />
at the time <strong>of</strong> the choice <strong>of</strong> the other firms. The above rule cannot be solved<br />
analytically, but it shows again that the investment <strong>of</strong> the outsider firms must<br />
be decreasing in the investment <strong>of</strong> the leader, ∂z/∂z L < 0: the higher the<br />
latter, the smaller the probability that no one innovates <strong>and</strong> therefore the<br />
expected gain from the investment <strong>of</strong> the followers is reduced. This implies<br />
that the leader has an incentive to choose a higher investment to strategically<br />
reduce the investment <strong>of</strong> the followers. However, the investment <strong>of</strong> the leader<br />
does not need to be higher than the investment <strong>of</strong> the other firms, because<br />
the Arrow effect is still pushing in the opposite direction. For instance, with<br />
two firms we have:<br />
z L =<br />
VK+(1− V )(V − K)<br />
1 − 2V (V − K)<br />
z =<br />
VK+(1− V )V − V<br />
3<br />
1 − 2V (V − K)<br />
(1.45)<br />
<strong>and</strong> the Arrow effect prevails on the Stackelberg effect whenever K>V 3 /(1−<br />
V ).<br />
When entry is endogenous, things are simpler. As long as the investment<br />
<strong>of</strong> the leader is small enough to allow entry <strong>of</strong> at least one outsider, the free<br />
entry condition is z(1 − z) n−2 (1 − z L )V = z 2 /2+F , which delivers again the<br />
investment z = √ 2F for each outsider. Putting together the two equilibrium<br />
conditions in the pr<strong>of</strong>it function <strong>of</strong> the leader, we would have:<br />
π L = K + z L (1 − z) n−1 (V − K) − z2 L<br />
2 − F =<br />
= K + z √ ³<br />
L 2F 1 − √ ´<br />
2F + K 1 − z L V<br />
√<br />
2F<br />
³<br />
1 − √ 2F<br />
´<br />
− z2 L<br />
2 − F<br />
whose third element, the one associated with the current pr<strong>of</strong>its obtained in<br />
case no one innovates, is independent from the choice <strong>of</strong> the leader. Consequently,<br />
the choice <strong>of</strong> the leader is taken exactly as in our earlier model (with<br />
K =0) <strong>and</strong> requires an investment:<br />
√<br />
2F<br />
¯z L =1−<br />
(1.46)<br />
V<br />
such that no other firm invests in innovation. Therefore, the pr<strong>of</strong>its <strong>of</strong> the<br />
leader can be calculated as a function <strong>of</strong> the value <strong>of</strong> the innovation π L =<br />
K +¯z L V +(1− ¯z L ) K − ¯z L 2 /2 − F .<br />
Welfare comparisons are ambiguous: on one side the aggregate probability<br />
<strong>of</strong> innovation is lower under Stackelberg competition with free entry rather<br />
than in the Marshall equilibrium, on the other side expenditure in fixed <strong>and</strong><br />
variable costs <strong>of</strong> research is lower in the first than in the second case. 33<br />
33 However, in a dynamic environment where the value <strong>of</strong> the innovation is endogenous,<br />
things would change. While without a leadership <strong>of</strong> the monopolist,
1.4 A Simple Model <strong>of</strong> <strong>Competition</strong> for the <strong>Market</strong> 33<br />
Moreover, notice that when the monopolist is the leader in the competition<br />
for the innovation, the Arrow effect disappears, because the choice <strong>of</strong><br />
the monopolist is independent from the current pr<strong>of</strong>its. The leadership in<br />
the competition for the market radically changes the behavior <strong>of</strong> a monopolist:<br />
from zero investment to maximum investment. The cited article <strong>of</strong> The<br />
Economist (2004) has discussed the relation between this theory <strong>and</strong> innovation<br />
by monopolists in real world markets. When entry is endogenous:<br />
“a market leader has a greater incentive than any other firm to<br />
keep innovating <strong>and</strong> thus stay on top. Blessed with scale <strong>and</strong> market<br />
knowledge, it is better placed than potential rivals to commit itself to<br />
financing innovations. Oddly–paradoxically, if you like–in fighting<br />
to maintain its monopoly it acts more competitively than firms in<br />
markets in which there is no obviously dominant player.<br />
The most important requirement for this result is a lack <strong>of</strong> barriers<br />
to entry: these might include, for example, big capital outlays to fund<br />
the building <strong>of</strong> new laboratories, or regulatory or licensing restrictions<br />
that make it hard for new firms to threaten an incumbent. If there<br />
are no such barriers, a monopolist will have an excellent reason to<br />
innovate before any potential competitor comes up with the next new<br />
thing. It st<strong>and</strong>s to lose its current, bloated pr<strong>of</strong>its if it does not; it<br />
st<strong>and</strong>s to gain plenty from continued market dominance if it does.<br />
If the world works in the way Mr Etro supposes, the fact that<br />
a dominant firm remains on top might actually be strong evidence<br />
<strong>of</strong> vigorous competition. However, observers (including antitrust authorities)<br />
may well find it difficult to work out whether a durable<br />
monopoly is the product <strong>of</strong> brilliant innovation or the deliberate<br />
strangulation <strong>of</strong> competitors. More confusing still, any half-awake<br />
monopolist will engage in some <strong>of</strong> the former in order to help bring<br />
about plenty <strong>of</strong> the latter. The very ease <strong>of</strong> entry, <strong>and</strong> the aggressiveness<br />
<strong>of</strong> the competitive environment, are what spur monopolists<br />
to innovate so fiercely.<br />
But what if there are barriers to entry? These tend to make the<br />
dominant firm less aggressive in investing in new technologies–in<br />
essence, because its monopoly with the existing technology is less<br />
likely to be challenged. Over time, however, other companies can innovate<br />
<strong>and</strong> gradually overcome the barriers... Meanwhile, the monopolist<br />
lives on marked time, burning <strong>of</strong>f the fat <strong>of</strong> its past innovations.<br />
the value <strong>of</strong> innovation would be just the value <strong>of</strong> expected pr<strong>of</strong>its from this<br />
innovation (the innovator will not invest further), with a leadership by the monopolist,<br />
the value <strong>of</strong> innovation should take into account the option value <strong>of</strong><br />
future leadership <strong>and</strong> future innovations: this would endogenously increase the<br />
value <strong>of</strong> being an innovator <strong>and</strong> would increase the aggregate incentives to invest.<br />
We will return on this important point in Chapter 4.
34 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
So much for theorizing. What might the practical implications be?<br />
One is that antitrust authorities should be especially careful when<br />
trying to stamp out monopoly power in markets that are marked by<br />
technical innovation. It could still be that firms like Micros<strong>of</strong>t are<br />
capable <strong>of</strong> using their girth to squish their rivals; the point is that<br />
continued monopoly is not cast-iron evidence <strong>of</strong> bad behavior.<br />
There might be a further implication for patent policy. Patents,<br />
after all, are government-endorsed monopolies for a given technology<br />
for a specified period. Mr Blundell <strong>and</strong> his colleagues found that<br />
the pharmaceutical industry provided the strongest evidence <strong>of</strong> correlation<br />
between market share <strong>and</strong> innovation. Thus strong patents,<br />
despite their recent bad press, can be a source <strong>of</strong> innovation. Generally,<br />
though, when one company dominates a market, people should<br />
be careful in assuming that it is guilty <strong>of</strong> sloth. It may be fighting<br />
for its life.”<br />
The idea behind this discussion can bedescribedinsimplertermsasa<br />
derivation <strong>of</strong> two sufficient conditions under which monopolists have incentivestoinvestinR&D<strong>and</strong>toinvestmorethanotherfirms:<br />
1) leadership<br />
for the monopolist <strong>and</strong> 2) endogenous entry for the outsiders in the race to<br />
innovate. We will return on these issues in Chapter 4, <strong>and</strong> discuss their policy<br />
implications in Chapters 5 <strong>and</strong> 6.<br />
Finally, we confirm that, also when the incumbent monopolist endogenously<br />
invests in R&D, the escape competition effect disappears: an increase<br />
in the intensity <strong>of</strong> product market competition as formalized by Aghion <strong>and</strong><br />
Griffith (2005) does not affect innovation when entry in the competition for<br />
the market is free. This may suggest that competition for the market could<br />
be a good substitute for competition in the market, another point on which<br />
we will return later in the book.<br />
1.5 Conclusions<br />
In this chapter we developed some toy models to compare different equilibria.<br />
Toy models can be quite suggestive <strong>and</strong> even provide many interesting<br />
insights, however they <strong>of</strong>ten hide very simplistic assumptions <strong>and</strong> it is hard<br />
to underst<strong>and</strong> whether certain results hold in general or just under specific<br />
assumptions. That is why it is now time to generalize our models at a deeper<br />
level. The objective <strong>of</strong> the next chapters will be an investigation <strong>of</strong> the general<br />
properties <strong>of</strong> our four alternative equilibria.<br />
Moreover, in this chapter we developed examples in which firms compete<br />
strategically in a symmetric way, or in which a firm is a leader <strong>and</strong> has a<br />
first mover advantage in the choice <strong>of</strong> its strategy. Since a commitment to<br />
a strategy (especially a price strategy) can lack credibility (especially in the<br />
long run), it is important to verify whether alternative credible commitments
1.5 Conclusions 35<br />
or strategic investments can sustain results similar to those derived here. In<br />
Chapter 2 we will approach this issue developing a general model <strong>of</strong> strategic<br />
commitments.<br />
Before moving to this task, however, it is important to summarize what<br />
we have learned with our toy models. First, we considered simple models<br />
<strong>of</strong> competition in quantities. We noticed that market leaders produce more<br />
output than each one <strong>of</strong> the other followers, both in the case <strong>of</strong> exogenous<br />
entry <strong>and</strong> in the case <strong>of</strong> endogenous entry. As we will see, this does not<br />
always hold with exogenous entry, but it always holds with endogenous entry.<br />
We also noticed that in certain situations (homogenous goods <strong>and</strong> constant<br />
marginal costs) leaders deter entry when entry is endogenous, while in other<br />
cases (U shaped average cost functions or imperfect substitutability between<br />
goods) they do not <strong>and</strong> allow entry. We also noticed that the behavior <strong>of</strong><br />
market leaders under price competition was radically different depending on<br />
the entry conditions. It is important to underst<strong>and</strong> what drives these results,<br />
<strong>and</strong> we will explore this issue in Chapter 3.<br />
Finally, we looked at a simple model <strong>of</strong> competition for the market <strong>and</strong><br />
obtained a surprising result. While incumbent monopolists do not have incentives<br />
to invest in R&D if the competition for innovating is free <strong>and</strong> symmetric<br />
between all firms, when these incumbents have a leadership in the competition<br />
for the market they also have strong incentives to invest <strong>and</strong> end up<br />
being the only investors. If this is the case, their leadership should be persistent<br />
over time <strong>and</strong> innovation <strong>and</strong> technological progress would be driven by<br />
market leaders. In Chapter 4 we will generalize the model <strong>of</strong> competition for<br />
the market in realistic ways <strong>and</strong> will try to evaluate these drastic results.
36 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
1.6 Appendix<br />
1. Taking Care <strong>of</strong> the Integer Constraint. In the derivation <strong>of</strong> the<br />
Stackelberg equilibrium with endogenous entry, homogenous goods <strong>and</strong> constant<br />
marginal costs <strong>of</strong> Section 1.1 we simplified things assuming that the<br />
number <strong>of</strong> firms was a real number. Here we verify that the equilibrium is<br />
exactly the same even if we consider, more realistically, that the number <strong>of</strong><br />
firms in the market must be an integer. We provide a constructive pro<strong>of</strong> since<br />
this is helpful to underst<strong>and</strong> the general behavior <strong>of</strong> the pr<strong>of</strong>its <strong>of</strong> the leader<br />
in a more general version where the integer constraint on the number <strong>of</strong> firms<br />
is taken in consideration.<br />
Given the production <strong>of</strong> the leader q L <strong>and</strong> the number <strong>of</strong> firms n, the<br />
reaction function <strong>and</strong> the pr<strong>of</strong>its <strong>of</strong> each follower are the same as before.<br />
However, the number <strong>of</strong> firms is a step function <strong>of</strong> the output <strong>of</strong> the leader.<br />
In particular, the number <strong>of</strong> firms is given by the integer number n ≥ 2 when<br />
the output <strong>of</strong> the leader is between s(n) <strong>and</strong> s(n − 1), where these cut-<strong>of</strong>fs<br />
are defined as:<br />
s(n) ≡ a − c − (n +1) √ F<br />
while only the leader can be pr<strong>of</strong>itably in the market (n =1)whenq L >s(1).<br />
Let us remember that for any exogenous number <strong>of</strong> firms the pr<strong>of</strong>its <strong>of</strong> the<br />
leader are maximized at the monopolistic output (a−c)/2, <strong>and</strong>thereforethis<br />
pr<strong>of</strong>its are increasing before <strong>and</strong> decreasing after this output level. Given this,<br />
we can determine the behavior <strong>of</strong> the pr<strong>of</strong>its <strong>of</strong> the leader in function <strong>of</strong> its<br />
output distinguishing three regions.<br />
The high output region, emerges for a small enough number <strong>of</strong> firms n<br />
such that s(n) > (a − c)/2 or n
1.6 Appendix 37<br />
It can be easily verified that π L (n) >π L (n +1) for any number <strong>of</strong> firms<br />
active in this region, therefore it is optimal to choose a production that<br />
maximizes pr<strong>of</strong>its with n =1, that is exactly the entry deterrence output<br />
s(1) = a − c − 2 √ F . This output delivers the pr<strong>of</strong>its:<br />
π L (1) = 2 √ ³<br />
F a − c − 2 √ ´<br />
F − F<br />
The low output region emerges for any high enough number <strong>of</strong> firms n<br />
such that s(n − 1) < (a − c)/2, orn>(a − c)/2 √ F .Thisimpliesthatthe<br />
pr<strong>of</strong>its <strong>of</strong> the leader are always increasing in the output. Trivially, it is never<br />
optimal to produce less than the monopolistic output.<br />
The third case emerges for a number <strong>of</strong> firms such that s(n) < (a−c)/2 <<br />
s(n − 1), or:<br />
µ a − c<br />
n ∈<br />
2 √ F − 1; a − c <br />
2 √ F<br />
In the interval <strong>of</strong> production x L ∈ [s(n),s(n − 1)] it is optimal for the leader<br />
to choose the monopolistic output level, because (only) in this interval pr<strong>of</strong>its<br />
have an inverted U shape. In this interval, the leader produces (a − c)/2 <strong>and</strong><br />
each one <strong>of</strong> the n−1 followers produces (a−c)/2n as in a st<strong>and</strong>ard Stackelberg<br />
model with an exogenous number <strong>of</strong> firms. The usual pr<strong>of</strong>its <strong>of</strong> the leader are<br />
then:<br />
(a − c)2<br />
π L (n) = − F<br />
4n<br />
<strong>and</strong> we need to verify that these are always smaller than what the leader can<br />
obtain with the entry deterrence strategy. Since:<br />
(a − c) 2<br />
π L (1) R π L (n) ⇔ n R<br />
8 √ F (a − c − 2 √ F )<br />
the pr<strong>of</strong>it maximizing choice <strong>of</strong> the leader could be in this region if there is<br />
anumber<strong>of</strong>firms n that belongs to the set derived above <strong>and</strong> that is lower<br />
than the cut-<strong>of</strong>f just obtained. However, this requires that this cut-<strong>of</strong>f is larger<br />
than (a − c)/2 √ F − 1 <strong>and</strong>:<br />
(a − c) 2<br />
8 √ F (a − c − 2 √ F ) > a − c<br />
2 √ − c)2<br />
− 1 iff F>(a<br />
F 16<br />
This is impossible because we assumed F
38 1. <strong>Competition</strong>, Leadership <strong>and</strong> Entry<br />
2. Endogenous Costs <strong>of</strong> Entry. The theory <strong>of</strong> Stackelberg competition<br />
with endogenous entry can also be seen as depicting the way a market<br />
leader can extract rents from a competitive market in the presence <strong>of</strong> fixed<br />
costs <strong>of</strong> entry. These costs can be interpreted as technological costs that are<br />
taken as given by the firms. However, they can also be endogenized imagining<br />
that they characterize the market <strong>and</strong> that the same leader can choose<br />
them in a preliminary stage. For instance, by investing in R&D or paying<br />
for an advertising campaign, or even by establishing certain barriers to entry<br />
associated with a cost <strong>of</strong> entry, the leader can set a sort <strong>of</strong> benchmark: all the<br />
other firms have to undertake the same investment, pay the same advertising<br />
campaign or face the same costs <strong>of</strong> entry to be able to compete in the market<br />
(Sutton, 1998).<br />
Imagine that the leader can choose the investment F . Consider for simplicity<br />
the linear example <strong>of</strong> competition in quantities <strong>of</strong> Section 1.1. The<br />
dem<strong>and</strong> <strong>and</strong> cost characteristics <strong>of</strong> this market depend on this investment so<br />
that the parameters a(F ) <strong>and</strong> c(F ) are now functions <strong>of</strong> the endogenous investment.<br />
This will be chosen to maximize the expected pr<strong>of</strong>its <strong>of</strong> the leader:<br />
π L (F )=2 √ F<br />
h<br />
a(F ) − c(F ) − 2 √ F<br />
i<br />
− F<br />
In general, the choice will imply a positive investment (otherwise the<br />
leader would expect zero pr<strong>of</strong>its). One can also show that from a welfare<br />
point <strong>of</strong> view, the leader will choose an excessive investment if this investment<br />
reduces its equilibrium production, but will choose a suboptimal investment<br />
in the opposite case. 34 In other words, leaders tend to do too little <strong>of</strong> good<br />
things <strong>and</strong> too much <strong>of</strong> bad things.<br />
For instance, imagine that F serves no real purpose other than raising<br />
the cost <strong>of</strong> entry (a 0 (F )=c 0 (F )=0). This is the case <strong>of</strong> what we usually call<br />
an artificial barrier to entry created by the dominant firm. The leader would<br />
maximize its expected pr<strong>of</strong>its choosing a positive barrier to entry:<br />
F ∗ (a − c)2<br />
=<br />
25<br />
which delivers the net pr<strong>of</strong>its:<br />
(a − c)2<br />
π L =<br />
5<br />
In other words the leader would create a completely useless barrier associated<br />
with a fixed cost (born by the leader as well) just to pr<strong>of</strong>it expostfrom<br />
its entry deterring strategy. Of course, in this case the welfare maximizing<br />
34 This is an immediate consequence <strong>of</strong> the definition <strong>of</strong> welfare as a sum <strong>of</strong> consumer<br />
surplus <strong>and</strong> pr<strong>of</strong>its. When pr<strong>of</strong>its <strong>of</strong> the leader are maximized the investment<br />
is excessive if the consumer surplus is decreasing in the investment, that is<br />
if output is decreasing. Of course this is still a second best comparison.
1.6 Appendix 39<br />
barrier would be F =0, which would lead to complete rent dissipation <strong>and</strong><br />
marginal cost pricing with zero pr<strong>of</strong>its for everybody. The moral <strong>of</strong> this story<br />
is that the priority in industrial policy should be to create the conditions for<br />
free entry <strong>and</strong> hence to fight against artificial barriers to entry, not to fight<br />
against leaders per se.
2. Strategic Commitments <strong>and</strong> Endogenous<br />
Entry<br />
In this chapter we will study a general model <strong>of</strong> market structure <strong>and</strong> characterize<br />
the incentives <strong>of</strong> a firm to adopt different strategic commitments<br />
to gain a competitive advantage over the rivals. We will develop a unified<br />
general framework in which st<strong>and</strong>ard models <strong>of</strong> competition in the market<br />
<strong>and</strong> for the market are nested, including those analyzed in Chapter 1 <strong>and</strong><br />
others analyzed <strong>and</strong> extended in Chapters 3 <strong>and</strong> 4. Virtually all models <strong>of</strong><br />
competition in quantities with homogenous or imperfectly substitute goods<br />
<strong>and</strong> with general shapes <strong>of</strong> the cost function are nested in our general model.<br />
Also encompassed are a wide class <strong>of</strong> models <strong>of</strong> competition in prices (as<br />
long as the dem<strong>and</strong> function satisfies some regularity conditions), including<br />
models with a constant expenditure dem<strong>and</strong> function or isoelastic dem<strong>and</strong><br />
functions (derived from quasilinear utilities or homotethic utilities àlaDixit<br />
<strong>and</strong> Stiglitz), <strong>and</strong> a wide class <strong>of</strong> models <strong>of</strong> competition for the market, whose<br />
detailed analysis will be postponed to Chapter 4.<br />
The initial focus <strong>of</strong> this chapter will be on Nash equilibria <strong>and</strong> on Marshall<br />
equilibria, that is on market structures characterized by symmetry between<br />
an exogenous number <strong>of</strong> firmsintheformercase<strong>and</strong>anendogenousnumber<br />
<strong>of</strong> firms in the latter case. Nash competition can be interpreted as a form <strong>of</strong><br />
competition with an exogenously limited number <strong>of</strong> firms, whose equilibrium<br />
can be seen as a short term equilibrium <strong>of</strong> a given market or even as a general<br />
equilibrium for a market where entry is exogenously constrained (for instance<br />
by legal or regulatory barriers to entry). A symmetric Nash equilibrium can be<br />
easily characterized through a single equilibrium pr<strong>of</strong>it maximizing condition<br />
that takes into account symmetry between the firms, for instance a mark up<br />
rule for the Cournot model <strong>of</strong> competition in quantities or for the Bertr<strong>and</strong><br />
model <strong>of</strong> competition in prices. Such a characterization allows one to study<br />
the comparative statics <strong>of</strong> the equilibrium variables <strong>and</strong> hence it is at the<br />
basis <strong>of</strong> the analysis <strong>of</strong> the interaction between exogenous variables (as costs,<br />
taxes, dem<strong>and</strong> parameters, or even the number <strong>of</strong> firms in the market) <strong>and</strong><br />
endogenous variables (output, prices, pr<strong>of</strong>its).<br />
Marshallian competition can be interpreted in terms <strong>of</strong> a medium or long<br />
run equilibrium in which there are not exogenous barriers to entry. In such a<br />
context entry is endogenously determined by the presence <strong>of</strong> pr<strong>of</strong>itable opportunities<br />
to be exploited. When these opportunities are exhausted, the entry
42 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
process stops. In the Marshall equilibrium, the strategies <strong>of</strong> the firms <strong>and</strong><br />
the number <strong>of</strong> firms are jointly determined by a pr<strong>of</strong>it maximizing condition<br />
<strong>and</strong> by an endogenous entry condition (typically a zero pr<strong>of</strong>it condition or a<br />
no arbitrage condition between entry in different sectors), both taking into<br />
account symmetry between firms. Also in such a case, we can easily verify<br />
the impact <strong>of</strong> changes in dem<strong>and</strong> <strong>and</strong> supply conditions <strong>and</strong> other exogenous<br />
policy parameters on the equilibrium variables, namely output, prices <strong>and</strong><br />
the number <strong>of</strong> firms.<br />
Building on this general framework <strong>and</strong> on this st<strong>and</strong>ard characterization<br />
<strong>of</strong> equilibria, we will introduce the analysis <strong>of</strong> market leaders verifying their<br />
incentives to adopt alternative strategic investments that can create a competitive<br />
advantage in the subsequent competition with the other firms. 1 As<br />
we will see, the behavior <strong>of</strong> the leaders changes when they face an exogenous<br />
number <strong>of</strong> competitors or an endogenous number. The first case has been<br />
characterized at least since the work <strong>of</strong> Fudenberg <strong>and</strong> Tirole (1984) <strong>and</strong><br />
Bulow et al. (1985) on duopolies. Suppose that firm i has the gross pr<strong>of</strong>its<br />
Π(x i ,X −i ,k), which depend on its strategy x i , the aggregate statistics X −i<br />
summarizing the strategies <strong>of</strong> the other firms, <strong>and</strong> the preliminary investment<br />
k. Then, the strategic incentives to invest for this firm depend on the impact<br />
<strong>of</strong> the investment on the marginal pr<strong>of</strong>itability (Π 13 ), 2 <strong>and</strong> on the nature<br />
<strong>of</strong> the strategic interaction between firms (Π 12 ). Therefore, they are typically<br />
different in models <strong>of</strong> competition in quantities, where output choices<br />
are <strong>of</strong>ten strategic substitutes (Π 12 < 0), <strong>and</strong> in models <strong>of</strong> competition in<br />
prices, where strategic complementarity usually holds (Π 12 > 0). On this<br />
basis, Tirole (1988) has built a taxonomy <strong>of</strong> business strategies that implies<br />
four different strategies for the leader: a firm may overinvest or underinvest<br />
initially to be more accommodating or aggressive subsequently. As shown in<br />
Etro (2006, a), things simplify drastically when entry <strong>of</strong> firmsisendogenous,<br />
because in such a case the strategic incentives to invest are independent<br />
from the strategic interaction between firms: the optimal investment <strong>of</strong> a<br />
firm depends only on whether the investment increases or not the marginal<br />
pr<strong>of</strong>itability, which leads to results that do not dependent on whether prices<br />
or quantities are the strategic variables. More precisely, when entry is endogenous,<br />
a firm invests always in the direction that leads to an aggressive<br />
behavior in the market. Of course, our interest in this outcome relies on the<br />
belief that in most situations entry in the markets is indeed endogenous, <strong>and</strong><br />
the proper way to analyze the behavior <strong>of</strong> firms should take this element into<br />
account.<br />
The abstract <strong>and</strong> general rules we just pointed out have a lot <strong>of</strong> applications<br />
to industrial organization <strong>and</strong> related fields, <strong>and</strong> this chapter will<br />
analyze a few <strong>of</strong> them. Fundamental strategic investments are those affect-<br />
1 See Singh et al. (1998) on the empirical relevance <strong>of</strong> strategic investments by<br />
leaders.<br />
2 Subscripts denote derivatives with respect to the arguments.
2. Strategic Commitments <strong>and</strong> Endogenous Entry 43<br />
ing supply, as cost reducing investments or overproduction in the presence<br />
<strong>of</strong> learning by doing, <strong>and</strong> those affecting dem<strong>and</strong>, as investments in quality<br />
improvements, in advertising, in product differentiation. We will show that<br />
when entry in the market is endogenous, a market leader has always a strategic<br />
incentive to overinvest in the first typology <strong>of</strong> investments because this<br />
leads to aggressive behavior, while the role <strong>of</strong> dem<strong>and</strong> enhancing investments<br />
is more complex.<br />
Another application concerns the theory <strong>of</strong> corporate finance: starting<br />
from the literature on the relation between the optimal financial structure<br />
<strong>and</strong> product market competition (Br<strong>and</strong>er <strong>and</strong> Lewis, 1986) we will examine<br />
the incentives to adopt strategic debt financing for markets with free entry.<br />
It turns out that under quantity competition there is always a strategic bias<br />
toward debt financing, while under price competition there is only when uncertainty<br />
affects costs, but not when it affects dem<strong>and</strong>. In general, departing<br />
from the st<strong>and</strong>ard Modigliani-Miller neutrality result, a financial tool like<br />
debt is useful when it constrains equity holders to adopt more aggressive<br />
strategies in the market, <strong>and</strong> this is the case when positive shocks increase<br />
marginal pr<strong>of</strong>its.<br />
Other new applications developed in detail here concern discrete commitments.<br />
We will examine the case <strong>of</strong> bundling strategies. In an influential<br />
paper, Whinston (1990) has studied bundling in a market for two goods. The<br />
primary good is monopolized by one firm, which competes with a single rival<br />
in the market for the secondary good. Under price competition in the secondary<br />
market, the monopolist becomes more aggressive in its price choice in<br />
the case <strong>of</strong> bundling <strong>of</strong> its two goods. Since a more aggressive strategy leads<br />
tolowerpricesforbothfirms as long as both are producing, the only reason<br />
why the monopolist may want to bundle its two goods is to deter entry <strong>of</strong> the<br />
rival in the secondary market. This conclusion can be highly misleading because<br />
it neglects the possibility <strong>of</strong> further entry in the market. We show that,<br />
if the secondary market is characterized by endogenous entry, the monopolist<br />
would always like to be aggressive in this market <strong>and</strong> bundling may be the<br />
right way to commit to an aggressive strategy: bundling would not necessarily<br />
exclude entry, but may increase competition in the secondary market <strong>and</strong><br />
reduce prices.<br />
Many other implications are relevant for antitrust policy. For instance,<br />
we will consider the theory <strong>of</strong> vertical restraints for interbr<strong>and</strong> competition<br />
(Rey <strong>and</strong> Stiglitz, 1988; Bonanno <strong>and</strong> Vickers, 1988), <strong>and</strong> show that a market<br />
leader facing endogenous entry would want to delegate distribution to a<br />
downstream retailer through wholesale prices below marginal cost: in such a<br />
case we have an example <strong>of</strong> a pro-competitive vertical restraint.<br />
Other results that are relevant for antitrust purposes concern the incentives<br />
to adopt limited interoperability, third degree price discrimination, <strong>and</strong><br />
aggressive pricing in the presence <strong>of</strong> network externalities or multi-sided markets.<br />
Finally, we will apply our result to horizontal mergers <strong>and</strong> show that
44 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
they create a strategic disadvantage for firms facing endogenous entry: therefore,<br />
in markets where entry is endogenous, mergers can only emerge when<br />
they create large efficiency gains, a point which is largely in line with the<br />
informal results <strong>of</strong> the Chicago school.<br />
The chapter is organized as follows. Sections 2.1 describes our general<br />
framework <strong>and</strong> Sections 2.2-2.3 characterize the Nash equilibrium <strong>and</strong> the<br />
Marshall equilibrium. Section 2.4 clarifies which models <strong>of</strong> competition in<br />
quantities, in prices <strong>and</strong> for the market are nested in the general framework,<br />
<strong>and</strong> derives some properties <strong>of</strong> these models. Section 2.5 analyzes general<br />
strategic investments in Nash <strong>and</strong> Marshall equilibria. Section 2.6-2.13 apply<br />
the results to a number <strong>of</strong> industrial organization issues. Section 2.14<br />
concludes.<br />
2.1 <strong>Market</strong> Structure<br />
A market structure is characterized by a number <strong>of</strong> firms, their strategies,<br />
a relationship that links all the strategies with the pr<strong>of</strong>it <strong>of</strong>eachfirm <strong>and</strong><br />
an equilibrium concept, which requires consistency between all the optimal<br />
strategies.<br />
In general, firms can choose many different strategies, for instance they<br />
can choose the price <strong>of</strong> their products, their quality, the investment in advertising<br />
<strong>and</strong> so on, <strong>and</strong> they can also choose these <strong>and</strong> other strategies for<br />
different products or different periods. However, in this chapter we will refer<br />
to the case <strong>of</strong> a single strategy. Imagine that in the market there are n firms<br />
<strong>and</strong> that the vector <strong>of</strong> their strategies is x =[x 1 ,x 2 , ..., x n ] where x i is the<br />
strategy <strong>of</strong> firm i. Wemaythinkthatdifferent firms have different features<br />
<strong>and</strong> different technological options, <strong>and</strong> there can be different pr<strong>of</strong>it functions,<br />
say π i (x) for each firm i. A market structure is a set <strong>of</strong> strategies x for<br />
n firms with pr<strong>of</strong>it functionsπ i (x), suchthatx i =argmaxπ i (x) for each i.<br />
A large portion <strong>of</strong> this book will deal with models <strong>of</strong> competition in the<br />
market where firms choose their output or their prices to maximize revenues<br />
net <strong>of</strong> the production cost c(·), which is increasing in the level <strong>of</strong> production,<br />
<strong>and</strong> net <strong>of</strong> a fixed cost F ≥ 0. In particular, we will deal with models <strong>of</strong><br />
quantity competition, as those studied in Chapter 1, where the strategy q i<br />
represents the level <strong>of</strong> production <strong>of</strong> firm i, <strong>and</strong> pr<strong>of</strong>its are given by:<br />
π i (q 1 , .., q i , .., q n )=q i p i (q 1 ,q 2 , .., q n ) − c(q i ) − F<br />
where p i (·) is the inverse dem<strong>and</strong>, decreasing in the output <strong>of</strong> every firm.<br />
Interesting applications that will be described in detail include models with<br />
linear dem<strong>and</strong>, as those adopted in Sections 1.1-2, models with isoelastic<br />
dem<strong>and</strong>s <strong>and</strong> homogenous goods, <strong>and</strong> models with imperfectly substitutable<br />
goods.
2.1 <strong>Market</strong> Structure 45<br />
Another wide class <strong>of</strong> models is based on price competition <strong>and</strong> imperfect<br />
substitution between goods, where the strategy p i represents the price <strong>of</strong> firm<br />
i <strong>and</strong> pr<strong>of</strong>its are given by:<br />
π i (p 1 , .., p i , .., p n )=p i D i (p 1 , .., p i ,..,p n ) − c [D i (p 1 , .., p i ,..,p n )] − F<br />
with a direct dem<strong>and</strong> function D i (p 1, .., p i , .., p n ) that decreases in the price <strong>of</strong><br />
firm i <strong>and</strong> increases in the price <strong>of</strong> the other firms. Applications that will be<br />
investigated later on include models with the Logit dem<strong>and</strong> (as those studied<br />
in the example <strong>of</strong> Section 1.3), models with isoelastic dem<strong>and</strong> functions,<br />
constant expenditure dem<strong>and</strong> functions <strong>and</strong> others.<br />
We will also study forms <strong>of</strong> competition for the market in which firms<br />
choose a strategy z i that allows to conquer a market whose value is V<br />
with a probability that depends also on the choices <strong>of</strong> the other firms,<br />
Pr i (z 1 , .., z i ,..,z n ). In such a case, expected pr<strong>of</strong>its are:<br />
E [π i (z 1 , .., z i , .., z n )] = Pr i (z 1 , .., z i ,..,z n )V − c (z 1 , .., z i , .., z n ) − F<br />
where the cost function c(·) can depend on the investment <strong>of</strong> each firm. Examples<br />
include the simple contest we studied in Section 1.4, more complicated<br />
patent races where firms invest over time <strong>and</strong> innovate according to complex<br />
stochastic processes, <strong>and</strong> also models <strong>of</strong> rent seeking where the probability<br />
that an agent obtains a generic rent is the ratio between the agent’s investment<br />
<strong>and</strong> total investment by all other agents.<br />
These market structures are general enough to include most <strong>of</strong> the realistic<br />
competitive frameworks analyzed in the theory <strong>of</strong> oligopoly. However, since<br />
a main topic <strong>of</strong> this book is the effect <strong>of</strong> entry on the strategic interaction<br />
between firms that have the same production technologies available <strong>and</strong> that<br />
face the same dem<strong>and</strong> structure, we need to impose some further restrictions<br />
on the functional forms to be used. In particular, in the main analysis we will<br />
focus on models in which all firms have the same cost technology <strong>and</strong> there<br />
are not exogenous differences or asymmetries between them.<br />
Accordingly, we will not deal with spatial models <strong>of</strong> horizontal or vertical<br />
differentiation like the Hotelling (1929) duopoly with spatial differentiation. 3<br />
3 Imagine two firms choosing their prices p 1 <strong>and</strong> p 2 with the pr<strong>of</strong>it functions<br />
π i (p i ,p j )=p i D(p i ,p j ) where dem<strong>and</strong>s are:<br />
D(p 1 ,p 2 )= (k 1 + k 2 )<br />
2<br />
+ (p 2 − p 1 )<br />
2(k 2 − k 1 ) , D(p 2,p 1 )= (2 − k 1 + k 2 )<br />
2<br />
− (p 2 − p 1 )<br />
2(k 2 − k 1 )<br />
Such an apparently complicated structure can be derived from a very simple<br />
situation. Imagine that consumers <strong>of</strong> a single unit <strong>of</strong> product are uniformly distributed<br />
along a market <strong>of</strong> unitary length, that is on [0, 1]. Onthismarkettwo<br />
firms are located at distances k 1 <strong>and</strong> k 2 >k 1 from the origin, produce homogenous<br />
goods at no cost <strong>and</strong> sell them at prices p 1 <strong>and</strong> p 2 . Each consumer at<br />
distance d from the origin will buy the good that minimizes the price plus a cost
46 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Clearly in such a model, the equilibrium prices <strong>and</strong> pr<strong>of</strong>its depend on the<br />
initial locations <strong>of</strong> the two firms/products. 4 In other words, pr<strong>of</strong>it functions<br />
<strong>and</strong> equilibrium outcomes depend on exogenous <strong>and</strong> firm specific parameters<br />
which introduce a substantial asymmetry between the firms. Moreover, if we<br />
were going to evaluate the entry opportunities in such a market, the result<br />
would be completely dependent on the location <strong>of</strong> the new entrants compared<br />
to the location <strong>of</strong> the incumbents. 5 The reason is that every new firm would<br />
compete just with its two closest rivals (for the consumers between them)<br />
<strong>and</strong> therefore each firm would have a different pr<strong>of</strong>it function depending on<br />
its particular competitors <strong>and</strong> their features. 6 Such a situation can depict<br />
markets where geographical location or, in a metaphorical sense, horizontal<br />
differentiation are a crucial element. However, it badly characterizes many<br />
other markets where each firm has to compete with all the other firms at<br />
once since all products in the market are potentially substitutes (which, in<br />
the applied analysis, is what defines a market). For this reason, our focus in<br />
this book, in line with the tradition associated with Chamberlin (1933), will<br />
be on models that allow for competition between symmetric firms.<br />
Finally, since we are interested in characterizing endogenous entry <strong>of</strong><br />
firms, we will limit our attention to markets where equilibrium pr<strong>of</strong>its decrease<br />
when entry occurs, a realistic feature that is not always verified in<br />
st<strong>and</strong>ard models. 7<br />
which is quadratic in the distance from the location <strong>of</strong> the corresponding firm,<br />
that is good i such that p i +(k i − d) 2 is smallest. This framework allows division<br />
<strong>of</strong> consumers between those buying good 1 <strong>and</strong> those buying good 2, delivering<br />
the dem<strong>and</strong>s above.<br />
4 Indeed, maximizing the two pr<strong>of</strong>it functions with respect to the prices <strong>and</strong> solving<br />
for them, one can find the equilibrium with p 1 =(2+k 1 + k 2 )(k 2 − k 1 )/3 <strong>and</strong><br />
p 2 =(4− k 1 − k 2 )(k 2 − k 1 )/3.<br />
5 Clearly one could endogenize the location decision (for instance, with two firms,<br />
they would choose maximum differentiation, placing themselves at the borders<br />
<strong>of</strong> the market with k 1 =0<strong>and</strong> k 2 =1). See the fundamental contribution <strong>of</strong><br />
D’Aspremont et al. (1979) for a formal <strong>and</strong> general treatment, <strong>and</strong> Anderson et<br />
al. (1992) for further discussion.<br />
6 We could easily extend the model to n firms symmetrically distributed along a<br />
circle where consmers are also distributed uniformly <strong>and</strong> choose between products<br />
as before (Vickrey, 1964). The Nash equilibrium would generate the price<br />
p =1/n 2 for each firm, <strong>and</strong> the Marshall equilibrium would imply n =1/ 3√ F<br />
firms selling at the price p = 3√ F 2 .<br />
7 For instance, we will exclude from our main analysis the basic model <strong>of</strong> price<br />
competition with linear dem<strong>and</strong> (associated with Bowley, 1924) as D i = a −p i +<br />
b j6=i p j. In the Nash-Bertr<strong>and</strong> equilibrium, this model implies that the pr<strong>of</strong>its<br />
<strong>of</strong> each firm increase in the number <strong>of</strong> firms. Something that makes no sense in<br />
real markets. See Section 3.4.5 on this point.
2.1 <strong>Market</strong> Structure 47<br />
More formally, in this book we will focus on a class <strong>of</strong> market structures<br />
with pr<strong>of</strong>it functions that are symmetric, additively separable <strong>and</strong> decreasing<br />
in the strategies <strong>of</strong> the other firms. For consistency, we will drop separate<br />
notations for different strategies <strong>and</strong> adopt a generic strategic variable x i ≥ 0<br />
for any firm i. Given the strategies x j for all j =1, 2, ..., n, eachfirm i has a<br />
net pr<strong>of</strong>it function:<br />
π i = Π (x i ,β i ) − F (2.1)<br />
which depends on two main factors: the strategy <strong>of</strong> the same firm x i <strong>and</strong> a<br />
factor which summarizes the strategies <strong>of</strong> the other firms β i .Weassumethat:<br />
Π 1 (x i ,β i ) R 0 for x i S x(β i )<br />
for some turning point x(β i ),<strong>and</strong>Π 11 (x, β) < 0, or more generally that<br />
Π (x, β) is quasiconcave in x. Therefore, it is an inverted U curve in x for<br />
any β.<br />
The effects (or spillovers) induced by the strategies <strong>of</strong> the other firms on<br />
firm i’s pr<strong>of</strong>its are summarized by:<br />
nX<br />
β i = h(x k ) (2.2)<br />
k=1,k6=i<br />
for some function <strong>of</strong> the strategies <strong>of</strong> each other firm h(x) that is assumed<br />
continuous, differentiable, non-negative <strong>and</strong> increasing in x. The gross pr<strong>of</strong>its<br />
are assumed to decrease in the strategies <strong>of</strong> the other firms <strong>and</strong> in their<br />
summary statistics β, thatisΠ 2 (x, β) < 0. 8<br />
In general, it could be that Π 12 is positive, so that we have strategic complementarity<br />
(since this implies ∂Π 1 (x i ,β i ) /∂x j > 0), from now on denoted<br />
with SC, or negative so that we have strategic substitutability (since this implies<br />
∂Π 1 (x i ,β i ) /∂x j < 0),denotedwithSSfromnowon.Intheformer<br />
case x 0 (β i ) > 0, which implies that the reaction functions are upward sloping<br />
(∂x(β i )/∂x j > 0 for all firms), in the latter x 0 (β i ) < 0, which implies<br />
that the reaction functions are downward sloping (∂x(β i )/∂x j < 0 for all<br />
firms). Of course, intermediate cases with non monotone reaction functions<br />
can emerge as well. An important outcome <strong>of</strong> the following analysis will concern<br />
the characterization <strong>of</strong> the firmsstrategiesunderdifferent conditions.<br />
For this purpose, let us introduce a behavioral definition: a strategy x is aggressive<br />
compared to another strategy x 0 if x>x 0 , <strong>and</strong> is accommodating in<br />
the opposite case; a firm adopting a strategy x>x 0 is more aggressive than<br />
a firm adopting a strategy x 0 .<br />
8 For models <strong>of</strong> competition in prices an axiomatic foundation for a similar pr<strong>of</strong>it<br />
function can be derived by a dem<strong>and</strong> system that satisfies the Independence<br />
from Irrelevant Alternatives property (the ratio <strong>of</strong> quantities dem<strong>and</strong>ed <strong>of</strong> any<br />
two goods is independent <strong>of</strong> the existence or price <strong>of</strong> a third good).
48 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
2.2 Nash Equilibrium<br />
Our first analysis is about competition between n firms. This number is kept<br />
exogenous <strong>and</strong> no other firms can enter in the market even if there are pr<strong>of</strong>itable<br />
opportunities to be exploited. This could happen because there are<br />
legal or institutional constraints on the number <strong>of</strong> actors in the market, or<br />
because the underlying technology is only available for a restricted number <strong>of</strong><br />
firms. In a Nash equilibrium every firm chooses its strategy to maximize its<br />
own pr<strong>of</strong>its given the strategies <strong>of</strong> the other firms <strong>and</strong> the equilibrium strategies<br />
must be consistent with each other. In this kind <strong>of</strong> game, a pure-strategy<br />
Nash equilibrium exists if the reaction functions are continuous or do not<br />
have downward jumps. While in general this may not hold, weak conditions<br />
for existence have been studied for many applications, 9 <strong>and</strong>inthisgeneral<br />
framework we will just assume the existence <strong>of</strong> a unique <strong>and</strong> symmetric equilibrium.Moreprecisely,wec<strong>and</strong>efine<br />
the following concept <strong>of</strong> symmetric<br />
equilibrium:<br />
Definition 2.1. A Nash Equilibrium between n firmsissuchthat:1)<br />
each firm chooses its strategy x to maximize its pr<strong>of</strong>its given the spillovers β<br />
from the other firms; 2) β =(n − 1)h(x).<br />
Notice that the last condition guarantees consistency between the fact<br />
that all firms choose the same strategy x <strong>and</strong> that the spillovers for each firm<br />
are at the same level β. We will assume that in equilibrium all firms make<br />
positive pr<strong>of</strong>its, or in other words, that the fixed cost is small enough to allow<br />
each firm to gain from being in the market.<br />
To characterize the equilibrium, notice that, given the strategy <strong>of</strong> each<br />
other firm, firm i chooses its own strategy to satisfy the first order condition<br />
Π 1 (x i ,β i )=0. Imposing symmetry in equilibrium between the followers we<br />
have:<br />
Π 1 [x, (n − 1)h(x)] = 0 (2.3)<br />
which completely defines the equilibrium strategy x. WerequireΠ 11 +(n −<br />
1)Π 12 h 0 (x) < 0 to assume local stability. 10<br />
To investigate the comparative properties <strong>of</strong> the Nash equilibrium with<br />
respect to the number <strong>of</strong> firms n, which is the only exogenous variable, let us<br />
totally differentiate the equilibrium condition to obtain:<br />
dx<br />
dn = Π 12 h(x)<br />
{−[Π 11 +(n − 1)Π 12 h 0 (x)]} T 0 if Π 12 T 0 (2.4)<br />
The related effects on pr<strong>of</strong>its are:<br />
9 See Vives (1999).<br />
10 In this book we will not deal with dynamic concepts <strong>of</strong> stability <strong>and</strong> evolutionary<br />
learning. On this issue see Fudenberg <strong>and</strong> Levine (1998).
2.3 Marshall Equilibrium 49<br />
dΠ<br />
dn = − Π 2 h(x)Π 11<br />
{−[Π 11 +(n − 1)Π 12 h 0 (x)]} < 0 (2.5)<br />
An increase in the number <strong>of</strong> firms increases the strategic choice <strong>of</strong> each<br />
firm if SC holds, <strong>and</strong> decreases it under SS, 11 while we always have a negative<br />
impact <strong>of</strong> entry on the pr<strong>of</strong>its <strong>of</strong> each firm as long as Π 2 < 0.<br />
2.3 Marshall Equilibrium<br />
Now we will drop the assumption that the number <strong>of</strong> firms is exogenous <strong>and</strong><br />
look at the more realistic situation in which firms can actually enter in the<br />
market if there are pr<strong>of</strong>itable opportunities to be exploited. If entry is free, it<br />
occurs until the gross pr<strong>of</strong>its are equal to the fixed costs <strong>of</strong> production. Nevertheless,<br />
one could also think <strong>of</strong> the pr<strong>of</strong>its in other sectors as constraining<br />
entry: according to this “general equilibrium” interpretation, a no arbitrage<br />
condition between sectors would make sure that net pr<strong>of</strong>its are equal in all<br />
sectors <strong>and</strong> it would endogenizes entry.<br />
As we noticed above, the pr<strong>of</strong>its for each firm in the Nash equilibrium are<br />
always decreasing in the number <strong>of</strong> firms. This implies that when there are<br />
positive pr<strong>of</strong>its in equilibrium with n firms, there is an incentive for outsiders<br />
to enter in the market. Then we can define a symmetric Nash equilibrium<br />
with endogenous entry as follows:<br />
Definition 2.2. A Marshall equilibrium is such that 1) each firm chooses<br />
its strategy x to maximize its pr<strong>of</strong>its given the spillovers β from the other<br />
firms; 2) the number <strong>of</strong> firms n is such that all firms make non negative<br />
pr<strong>of</strong>its <strong>and</strong> entry <strong>of</strong> one more firm would induce negative pr<strong>of</strong>its for all <strong>of</strong><br />
them; 3) β =(n − 1)h(x).<br />
To characterize the equilibrium, we still have the first order equilibrium<br />
condition:<br />
Π 1 [x, (n − 1)h(x)] = 0 (2.6)<br />
Moreover, we can impose the endogenous entry requirement as a zero pr<strong>of</strong>it<br />
condition. We will neglect the integer constraint on the number <strong>of</strong> firms: this<br />
is a good approximation when there are many firms - in general, the exact<br />
equilibrium number <strong>of</strong> firms would be the higher integer that is smaller than<br />
our equilibrium number.<br />
The endogenous entry condition becomes:<br />
11 It can be equivalently shown that the effect <strong>of</strong> any other exogenous parameter<br />
depends on its impact on the marginal effect <strong>of</strong> the strategic variable.
50 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Π [x, (n − 1)h(x)] = F (2.7)<br />
These two equations define the strategy <strong>of</strong> each firm <strong>and</strong> the number <strong>of</strong> firms<br />
as functions <strong>of</strong> the fixed cost. Local stability requires now Π 2 h(x) +Π 11 +<br />
(n − 1)Π 12 h 0 (x) < 0. To study the comparative statics <strong>of</strong> the system, we<br />
totally differentiate it with respect to F to obtain:<br />
dx<br />
dF = Π 12<br />
−Π 11 Π 2<br />
R 0 if Π 12 Q 0<br />
dn<br />
dF = Π 11 +(n − 1)Π 12 h 0 (x)<br />
< 0 (2.8)<br />
Π 11 Π 2 h(x)<br />
as long as Π 11 +(n − 1)Π 12 h 0 (x) < 0. Hence, a Marshall equilibrium implies<br />
strategies decreasing (increasing) in the fixed cost under SC (SS) <strong>and</strong><br />
anumber<strong>of</strong>firms decreasing in the fixed cost. An interesting interpretation<br />
<strong>of</strong> these comparative statics effects emerges if we think <strong>of</strong> a general equilibrium<br />
model where an increase in F corresponds to a positive shock on the<br />
pr<strong>of</strong>itability <strong>of</strong> the other sectors. Such a shock would make firms more aggressive<br />
in a market with SS <strong>and</strong> more accommodating in a market with SC,<br />
but it would always reduce the number <strong>of</strong> firms. For instance, if we think <strong>of</strong><br />
markets with competition in prices in general equilibrium, a positive shock<br />
in one sector has the effect <strong>of</strong> increasing prices <strong>and</strong> reducing entry in the<br />
other sectors, an implication rarely matched by macroeconomic models with<br />
imperfect competition (since these models typically neglect the endogeneity<br />
<strong>of</strong> entry). 12<br />
2.4 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the<br />
<strong>Market</strong><br />
In this section we will show that general models <strong>of</strong> competition in quantities<br />
<strong>and</strong> in prices <strong>and</strong> models <strong>of</strong> competition for the market are nested in our general<br />
framework, <strong>and</strong> we will analyze the Nash equilibrium <strong>and</strong> the Marshall<br />
equilibrium in these models.<br />
2.4.1 <strong>Competition</strong> in Quantities<br />
In Chapter 1 we examined some simple cases <strong>of</strong> competition in quantities.<br />
Here we will examine more general models <strong>of</strong> this kind. Consider a general<br />
dem<strong>and</strong> function:<br />
⎡<br />
⎤<br />
nX<br />
p i = p ⎣ x i , h(x j ) ⎦<br />
j6=i<br />
12 Introducing another exogenous parameter, say k, affecting each pr<strong>of</strong>it function<br />
Π(x i ,β i ,k) with Π 3 > 0, the strategies are decreasing in k whenever Π 13 Π 2 ><br />
Π 3Π 12, whiletheeffect on the number <strong>of</strong> firms is ambiguous.
2.4 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 51<br />
decreasing in both arguments, <strong>and</strong> a cost function c(x i ) for firm i,withc 0 (·) ><br />
0, wherex i is the quantity produced by firm i. Pr<strong>of</strong>its are then:<br />
⎡<br />
⎤<br />
nX<br />
π i = x i p ⎣ x i , h(x j ) ⎦ − c(x i ) − F (2.9)<br />
j6=i<br />
Using our definitions, the gross pr<strong>of</strong>it function can be written as:<br />
Π (x i ,β i )=x i p (x i ,β i ) − c(x i ) (2.10)<br />
<strong>and</strong> it can be easily verified that it is nested in our class <strong>of</strong> market structures<br />
(2.1) under weak conditions. This model is general enough to take into<br />
account different shapes <strong>of</strong> the cost function <strong>and</strong> imperfect substitutability<br />
between goods. In general, it can be characterized by SS or SC since we have<br />
Π 12 = p x + x i p xβ ,whosefirst element is negative <strong>and</strong> whose second element,<br />
proportional to the impact <strong>of</strong> a change <strong>of</strong> production <strong>of</strong> other firms on the<br />
slope <strong>of</strong> inverse dem<strong>and</strong>, has an ambiguous sign.<br />
Here, for simplicity, we will focus on the case where β i = P n<br />
k=1,k6=i x k,<br />
that is h(x i )=x i . For instance, assuming linear dem<strong>and</strong> functions as those<br />
studied in Chapter 1, we would have:<br />
p i = a − x i − bβ i , b ∈ (0, 1]<br />
where Π 12 (x i ,β i )=−b 0<br />
where Π 12 (x i ,β i )=−γ [a + β i − γx i ](x i + β i ) −γ−2 whose sign is positive<br />
for x i low enough <strong>and</strong> negative for x i high enough: consequently, the reaction<br />
functions have an inverse U shape. This dem<strong>and</strong> can be derived from a<br />
st<strong>and</strong>ard constant elasticity utility function:<br />
U = (a + P n<br />
J=1 C j) 1−γ<br />
+ C 0 (2.12)<br />
1 − γ<br />
for γ>0.
52 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
It is also possible to have situations in which SC holds always. For instance,<br />
Stackelberg (1934) presented an example with exponential dem<strong>and</strong><br />
p = exp [−(x i + β i ) υ ] which generates SC for υ ∈ (0, 1). Nevertheless, we<br />
should keep in mind that output strategies are complements only in the extreme<br />
cases in which dem<strong>and</strong> is highly convex.<br />
A general characterization <strong>of</strong> Cournot models is beyond our scope, therefore,<br />
in the rest <strong>of</strong> this section, we will focus on some particular cases.<br />
Homogenous goods. In the case <strong>of</strong> homogenous goods, the Nash-Cournot<br />
equilibrium condition under symmetry becomes: 13<br />
p(X)+xp 0 (X) =c 0 (x)<br />
where total output is X = nx (under the second order condition 2p 0 + xp 00 <<br />
0). This is the usual rule equating marginal revenue <strong>and</strong> marginal cost, <strong>and</strong><br />
can be rewritten as a mark-up rule (p − c 0 ) /p = −xp 0 /p, whose right h<strong>and</strong><br />
side is the inverse <strong>of</strong> the elasticity <strong>of</strong> direct dem<strong>and</strong> = −(dx/dp)(p/x).<br />
Therefore, we obtain the following expression for the equilibrium price:<br />
p(X) =<br />
c0 (x)<br />
1 − 1/<br />
(2.13)<br />
Focusing on the linear costs case with a constant marginal cost c, thecomparative<br />
statics with respect to the number <strong>of</strong> firms provide:<br />
dx x(E − 1)<br />
=<br />
dn 1+n − nE<br />
dp [n − E(n − 1)]xp0<br />
= < 0<br />
dn 1+n(1 − E)<br />
where E ≡ −xp 00 (nx)/p 0 (nx) is the elasticity <strong>of</strong> the slope <strong>of</strong> the inverse<br />
dem<strong>and</strong> with respect to the individual output <strong>of</strong> a firm, which is an index <strong>of</strong><br />
the convexity <strong>of</strong> the dem<strong>and</strong> function (E =0under linear dem<strong>and</strong>). Notice<br />
that the second order condition requires E
2.4 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 53<br />
The price increases less (more) than proportionally if E < (>)1/n, while<br />
pr<strong>of</strong>its decrease in the marginal cost unless E ∈ (2/n, 1+1/n). Noticethat<br />
this general Cournot model with n firms boils down to the monopoly model<br />
after imposing n =1, <strong>and</strong> we can verify that these comparative statics results<br />
match those emerging in the classic case <strong>of</strong> a monopoly for n =1.Forinstance,<br />
under linear dem<strong>and</strong> (E =0), a unitary increase <strong>of</strong> the marginal cost<br />
increases by half the monopolistic price, but by two thirds the duopolistic<br />
price, <strong>and</strong> so on until full shifting <strong>of</strong> the marginal cost on the price under<br />
perfect competition (for n →∞): a more convex dem<strong>and</strong> function leads to<br />
a larger shift <strong>of</strong> the cost change on the price. Generally, these results are<br />
quite useful since they can be used to evaluate the complex impact on the<br />
equilibrium prices <strong>and</strong> pr<strong>of</strong>its <strong>of</strong> an increase in costs due to different factors<br />
as a change in the costs <strong>of</strong> the inputs <strong>of</strong> production or in the indirect taxes. 14<br />
Let us move to the Marshall equilibrium. The two equilibrium conditions<br />
are now the optimality condition for a representative firm <strong>and</strong> the zero pr<strong>of</strong>it<br />
condition:<br />
p(X)+xp 0 (X) =c 0 (x), xp(X) =c(x)+F (2.14)<br />
Totally differentiating the system we can derive the comparative statics <strong>of</strong> a<br />
change in the constant marginal cost. The new effects are:<br />
dx<br />
dc = p0 x 2<br />
n∆ < 0<br />
dp<br />
dc = 2x2 p 02<br />
∆ > 0 dn<br />
dc = p0 x<br />
(2 − nE)<br />
∆<br />
Since ∆ ≡ x 2 p 02 (2p 0 + xp 00 ) > 0 by the second order condition, we can easily<br />
obtain that the cost increase raises the price less (more) than proportionally<br />
if E)0. Thenumber<strong>of</strong>firms is decreasing in the marginal cost except<br />
in the case <strong>of</strong> a highly convex dem<strong>and</strong> function.<br />
Hyperbolic dem<strong>and</strong>. As an example, let us look at the hyperbolic dem<strong>and</strong>:<br />
1<br />
p = P n<br />
J=1 x (2.15)<br />
j<br />
which can be derived from a st<strong>and</strong>ard logarithmic utility:<br />
à n<br />
!<br />
X<br />
U =log C j + C 0 (2.16)<br />
J=1<br />
14 For instance, in the linear case with a specific taxt s <strong>and</strong> an ad valorem tax t v<br />
we have:<br />
p =<br />
a<br />
n +1 + n c + t<br />
s<br />
<br />
n +1 1 − t v<br />
which shows that the price is decreasing in the number <strong>of</strong> firms <strong>and</strong> in both<br />
the taxes. For more results on tax incidence in oligopoly see Delipalla <strong>and</strong> Keen<br />
(1992), Myles (1995), <strong>and</strong> in presence <strong>of</strong> tax evasion Etro (1997, 1998a,b), Cowell<br />
(2004) <strong>and</strong> Marchese (2006).
54 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
where C j is consumption <strong>of</strong> good j <strong>and</strong> good 0 is the numeraire. 15 It can<br />
be easily verified that the Nash equilibrium is characterized by a production<br />
for each firm equal to x =(n − 1)/n 2 c, <strong>and</strong> by the following price <strong>and</strong> gross<br />
pr<strong>of</strong>its:<br />
c<br />
p =<br />
Π = 1 1 − 1/n<br />
n 2 (2.17)<br />
Notice that pr<strong>of</strong>its are now independent from the marginal cost, which is in<br />
line with our general result (since E =2/n implies dΠ/dc =0), while they<br />
decrease in the number <strong>of</strong> firms. In a Marshall equilibrium (assuming F 0 <strong>and</strong> g 0 (p) < 0: thefirst assumption implies<br />
that the dem<strong>and</strong> <strong>of</strong> firm i decreases in the price <strong>of</strong> firm i, <strong>and</strong> the remaining<br />
assumptions make sure that it increases with the prices <strong>of</strong> the other firms.<br />
Focusing on the case <strong>of</strong> a constant marginal cost, we then have the gross<br />
pr<strong>of</strong>its:<br />
⎡<br />
⎤<br />
nX<br />
π i =(p i − c)D ⎣p i , g(p j ) ⎦ − F (2.20)<br />
j=1,j6=i<br />
In Chapter 1 we developed an example based on the Logit dem<strong>and</strong>:<br />
15 Notice that the hyperbolic dem<strong>and</strong> is nested in the non linear one cited above<br />
for a =0<strong>and</strong> γ =1.
2.4 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 55<br />
D i =<br />
Ne−λp i<br />
P n<br />
j=1 e−λp j<br />
(2.21)<br />
which belongs to our class <strong>of</strong> dem<strong>and</strong> functions after setting g(p) =exp(−λp),<br />
that satisfies g 0 (p) < 0. Andersonet al. (1988) have shown that this dem<strong>and</strong><br />
is consistent with a representative agent maximizing the utility:<br />
µ 1 X n µ <br />
Cj<br />
U = C 0 − C j ln<br />
(2.22)<br />
λ<br />
N<br />
j=1<br />
when when P n<br />
j=1 C j = N <strong>and</strong> −∞ otherwise (total consumption for the<br />
n goods is exogenous), under the budget constraint C 0 + P n<br />
j=1 p jC j = Y ,<br />
with C 0 as the numeraire. This interpretation allows one to think <strong>of</strong> 1/λ as<br />
a measure <strong>of</strong> the variety-seeking behavior <strong>of</strong> the representative consumer.<br />
Other important cases derive from the class <strong>of</strong> dem<strong>and</strong> functions introduced<br />
by Spence (1976) <strong>and</strong> Dixit <strong>and</strong> Stiglitz (1977) <strong>and</strong> derived from<br />
the h maximization ³ <strong>of</strong> a utility function <strong>of</strong> a representative agent as U =<br />
Pn<br />
´i<br />
u C 0 ,V<br />
j=1 Cθ j under the budget constraint C 0 + P n<br />
j=1 p jC j = Y ,<br />
where C 0 is the numeraire, u(·) is quasilinear or homothetic, V (·) is increasing<br />
<strong>and</strong> concave, <strong>and</strong> θ ∈ (0, 1] parametrizes the substitutability between<br />
goods. Consider the utility function:<br />
⎡ ⎤ 1<br />
θ<br />
nX<br />
U = C0<br />
α ⎣ Cj<br />
θ ⎦<br />
(2.23)<br />
j=1<br />
with θ ∈ (0, 1) <strong>and</strong> α>0. In this case the constant elasticity <strong>of</strong> substitution<br />
between goods is 1/(1 − θ) <strong>and</strong> increases in θ: for this reason this model<br />
is <strong>of</strong>ten referred to as the CES (constant elasticity <strong>of</strong> substitution) model.<br />
Dem<strong>and</strong> for each good i =1, ..., n can be derived as:<br />
Yp − 1<br />
1−θ<br />
i<br />
D i =<br />
(1 + α) P n<br />
θ<br />
j=1 p− 1−θ<br />
j<br />
(2.24)<br />
which belongs to our general class after setting g(p) =p − 1−θ , which <strong>of</strong> course<br />
satisfies g 0 (p) < 0. Similar dem<strong>and</strong> functions <strong>and</strong> related models <strong>of</strong> price competition<br />
have been widely employed in many fields where imperfect competition<br />
plays a crucial role, including the new trade theory, the newkeynesian<br />
macroeconomics, the new open macroeconomy, the endogenous growth theory<br />
<strong>and</strong> the new economic geography. 16<br />
We now have to verify that the pr<strong>of</strong>it functions derived from this class<br />
<strong>of</strong> dem<strong>and</strong> functions are actually nested in our general model with gross<br />
16 Anderson et al. (1992) have provided a detailed analysis <strong>of</strong> the foundations for<br />
the Logit <strong>and</strong> CES dem<strong>and</strong> functions through three different approaches (rep-<br />
θ
56 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
pr<strong>of</strong>its Π (x i ,β i ). For this purpose, we will adopt a simple trick changing the<br />
strategic variable for each firm i from the price p i to its inverse x i ≡ 1/p i . 17<br />
Of course, choosing a price or its inverse is just a matter <strong>of</strong> mathematical<br />
definition, however it allows us to greatly simplify our discussion. First <strong>of</strong><br />
all, increasing x i =1/p i is now equivalent to reducing the price <strong>of</strong> firm i in<br />
both models <strong>of</strong> competition in quantities <strong>and</strong> in prices. Moreover, under our<br />
specification <strong>of</strong> the dem<strong>and</strong> functions, we can now define:<br />
µ 1<br />
h(x i )=g with h 0 (x i )=−(1/x 2 i )g 0 (1/x i ) > 0<br />
x i<br />
<strong>and</strong> rewrite gross pr<strong>of</strong>its as:<br />
µ <br />
1 1<br />
Π (x i ,β i )=µ<br />
− c D ,β<br />
x i x i<br />
i<br />
(2.25)<br />
The model belongs to our class <strong>of</strong> consistent market structures (2.1) under<br />
weak regularity conditions. Moreover SC holds as long as DD 12 0 is the elasticity <strong>of</strong> the direct dem<strong>and</strong>. Assuming that<br />
SC holds, we have the comparative statics results:<br />
dp<br />
dn ∝ p2 g(p)[D 2 +(p − c)D 12 ] < 0<br />
dp<br />
dc ∝ −p2 D 1 > 0<br />
resentative consumers models as those emphasized here, discrete choice models<br />
with stochastic utility <strong>and</strong> a multidimensional generalization <strong>of</strong> the Hotelling<br />
model) <strong>and</strong> <strong>of</strong> the existence <strong>of</strong> the related equilibria. For the case <strong>of</strong> an exponential<br />
subutility in the Dixit-Stiglitz preferences see Behrens <strong>and</strong> Murata (2007). I<br />
am thankful to Avinash Dixit to point this out.<br />
17 We borrowed this device from Mas-Colell et al. (1995, Ch. 12).
2.4 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 57<br />
while the effect <strong>of</strong> a change in the marginal cost on the pr<strong>of</strong>its is ambiguous.<br />
The Marshall equilibrium requires that all firms choose their prices optimally<br />
<strong>and</strong> that pr<strong>of</strong>its are driven to zero by endogenous entry:<br />
D [p, (n − 1)g(p)] + (p − c) D 1 [p, (n − 1)g(p)] = 0 (2.27)<br />
D [p, (n − 1)g(p)] (p − c) =F (2.28)<br />
Total differentiation <strong>of</strong> this equilibrium system generates the following comparative<br />
statics result:<br />
dp<br />
dc ∝ −g(p)D [2D 2 +(p − c)D 12 ] > 0<br />
while the effect <strong>of</strong> the marginal cost on the number <strong>of</strong> firmsisambiguous.<br />
Some examples. As we have seen in Chapter 1, in the case <strong>of</strong> a Logit<br />
dem<strong>and</strong> (2.21) under exogenous entry we have:<br />
p = c +<br />
n<br />
(n − 1)λ<br />
Π =<br />
N<br />
λ(n − 1) − F (2.29)<br />
while the endogenous entry equilibrium implies: 18<br />
p = c + F N + 1 λ<br />
n =1+ N λF<br />
(2.30)<br />
In the case <strong>of</strong> a CES dem<strong>and</strong> (2.24), the Nash equilibrium generates the<br />
following price <strong>and</strong> pr<strong>of</strong>its: 19<br />
p =<br />
c(n − θ)<br />
θ(n − 1)<br />
Π =<br />
Y (1 − θ)<br />
γ(n − θ)<br />
(2.31)<br />
This clearly implies a price decreasing in the number <strong>of</strong> firms <strong>and</strong> increasing<br />
more than proportionally in the marginal cost (dp/dc > 1). Gross pr<strong>of</strong>its for<br />
each firm are independent from the marginal cost, decreasing in the number <strong>of</strong><br />
firms <strong>and</strong> converging to zero when this number grows. Finally, in the Marshall<br />
equilibrium <strong>of</strong> the Dixit-Stiglitz model we have: 20<br />
18 The firstbestwouldrequireonefirm less than in the Marshall equilibrium. The<br />
second best under the zero pr<strong>of</strong>it constraint would require a price p = c +1/λ<br />
with N/F λ firms.<br />
19 In this case under specific <strong>and</strong>ad valorem taxation we have:<br />
p = (c + ts )(n − θ)<br />
(1 − t v )θ(n − 1)<br />
which implies overshifting <strong>of</strong> both taxes.<br />
20 The first best would require price equal to the marginal cost with Y (1−θ)/F (1+<br />
θα) firms. The second best under the zero pr<strong>of</strong>it constraint would require a price<br />
p = c/θ with Y (1 − θ)/F (1 + α) firms.
58 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
p =<br />
cY<br />
θ [Y − F (1 + α)]<br />
n =<br />
(1 − θ)Y<br />
(1 + α)F + θ (2.32)<br />
Notice that these equilibria can be compared with those that would<br />
emerge with the same isoelastic dem<strong>and</strong> function if firms were competing<br />
in quantities rather than in prices. 21 In that case one could solve for the<br />
Cournot equilibrium with an exogenous number <strong>of</strong> firms <strong>and</strong> obtain a price<br />
p = cn/θ(n − 1). This is higher than the price obtained above: competition<br />
in prices reduces the mark up <strong>and</strong> the pr<strong>of</strong>its <strong>of</strong> the firms compared to competition<br />
in quantities (this result holds in a more general set up than this).<br />
Finally, in all these cases the equilibrium price does not converge to the<br />
marginal cost when the number <strong>of</strong> firms increases (it converges to c+1/λ with<br />
the Logit dem<strong>and</strong> <strong>and</strong> to c/θ with the isoelastic dem<strong>and</strong>). This is possible<br />
because <strong>of</strong> product differentiation, which allows firms to maintain a certain<br />
degree <strong>of</strong> market power even if there are many competitors in the market; for<br />
this reason these kinds <strong>of</strong> models are <strong>of</strong>ten referred to as models <strong>of</strong> monopolistic<br />
competition - <strong>and</strong> in general equilibrium applications they are <strong>of</strong>ten<br />
employed neglecting the strategic interactions (so that the number <strong>of</strong> firms<br />
does not affect equilibrium prices <strong>and</strong> pr<strong>of</strong>its, <strong>and</strong> endogeneity <strong>of</strong> entry is<br />
irrelevant).<br />
2.4.3 <strong>Competition</strong> for the <strong>Market</strong><br />
A large class <strong>of</strong> models <strong>of</strong> investment in innovation or competition for the<br />
market can be studied within our general framework. For instance, in Chapter<br />
1westudiedasimplecontestwhereeveryfirmcouldobtainaninnovation<br />
with probability x i ∈ [0, 1] after investing x 2 i /2. The expected pr<strong>of</strong>its were:<br />
π i = x i<br />
n<br />
Y<br />
j=1,j6=i<br />
(1 − x j ) V − x2 i<br />
2 − F (2.33)<br />
That model was nested in our general framework, even if in such a case we<br />
would need a few steps to realize it:<br />
π i = x i e n<br />
j=1,j6=i log(1−xj) V − x2 i<br />
2 − F =<br />
= x i e − <br />
n<br />
j=1,j6=i log 1<br />
1−x j<br />
V − x2 i<br />
2 − F<br />
21 Maximizing the utility (2.23) one obtains the inverse dem<strong>and</strong>:<br />
Yx −(1−θ)<br />
i<br />
p i = n<br />
<br />
(1 + α)<br />
j=1 xθ j<br />
<strong>and</strong> therefore a pr<strong>of</strong>it function which is nested in our general specification (2.1).
2.5 Strategic Investments 59<br />
Now, setting h(x) =log[1/(1 − x)] which implies h 0 (x) =1/(1 − x) > 0, we<br />
can rewrite gross pr<strong>of</strong>its as:<br />
Π (x i ,β i )= x iV<br />
e β i<br />
− x2 i<br />
2<br />
(2.34)<br />
which is clearly nested in our model (2.1) <strong>and</strong> implies SS since Π 12 =<br />
−V/e β i < 0.<br />
22<br />
As we have seen in Chapter 1, <strong>and</strong> as one can easily verify from the first<br />
order condition under symmetry, the Nash equilibrium is characterized by an<br />
investment in innovation implicitly given by:<br />
x =(1− x) n−1 V (2.35)<br />
while in the Marshall equilibrium, where the number <strong>of</strong> firms reduces expected<br />
pr<strong>of</strong>its to zero, the investment is:<br />
x = √ 2F (2.36)<br />
Another related contest which is nested in our framework is a rent seeking<br />
contest in which agents invest to obtain rents with a probability given by<br />
their investment relative to the total one (Tullock, 1967). In Chapter 4 we will<br />
study more realistic forms <strong>of</strong> competition for the market where firms invest<br />
over time <strong>and</strong> innovations arrive according to a stochastic process depending<br />
on the investment <strong>of</strong> each firm (Loury, 1979). While that framework will allow<br />
us to consider further issues, many basic insights from the simple contest<br />
outlined here will be conserved.<br />
2.5 Strategic Investments<br />
Amainleitmotif <strong>of</strong> this book is about the behavior <strong>of</strong> market leaders in different<br />
forms <strong>of</strong> competitive environments. In the rest <strong>of</strong> this chapter we will<br />
approach this issue extending the framework analyzed until now to strategic<br />
investments or commitments by the leading firm. With strategic commitments<br />
we refer to any kind <strong>of</strong> preliminary decisions that affect the strategic<br />
condition <strong>of</strong> the leaders compared to the other firms. In the jargon <strong>of</strong> marketing,<br />
we may refer to all those commitments that affect the marketing mix,<br />
the so-called 4 P’s <strong>of</strong> marketing: product, price, place <strong>and</strong> promotion, here<br />
meaning quality <strong>of</strong> the good, costs, distribution <strong>and</strong> advertising (see Kotler,<br />
1999). In the jargon <strong>of</strong> strategy, we may refer to all those commitments that<br />
affect the competitive strategy <strong>and</strong> provide a competitive advantage to the<br />
leader (see Porter, 1985).<br />
22 This model can also be used as a foundation <strong>of</strong> a simple principle-agent model<br />
(for an introduction see Milgrom <strong>and</strong> Roberts, 1992) with which one can study<br />
hyerarchies within teams (see Goldfain, 2007).
60 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
More formally, in what follows we will study markets in which all firms<br />
compete simultaneously as before, but one <strong>of</strong> them, the leader, will have a<br />
chance to undertake a preliminary investment which will affect competition<br />
ex post. The purpose, <strong>of</strong> course, is to underst<strong>and</strong> what kind <strong>of</strong> decisions are<br />
taken by market leaders, whether they are going to induce an aggressive or<br />
an accommodating behavior, <strong>and</strong> how they affect equilibria. The pioneering<br />
analysis in this field is due to Dixit (1980) <strong>and</strong> Fudenberg <strong>and</strong> Tirole (1984),<br />
who focused on duopolies, while here we will consider the situation in which<br />
there is an exogenous number <strong>of</strong> firms n, possibly larger than two.<br />
Consider the following sequence <strong>of</strong> moves:<br />
1) in the first stage a leader, firm L, makes a strategic commitment on a<br />
variable k (we will <strong>of</strong>ten refer to this as to a strategic investment);<br />
2) in the second stage each follower chooses its own strategy x i <strong>and</strong> the<br />
leader chooses its own strategy x L after knowing the commitment <strong>of</strong> the<br />
leader. Therefore, all firms, the leader <strong>and</strong> the followers, play in Nash strategies<br />
in the second stage.<br />
In the second stage the pr<strong>of</strong>it <strong>of</strong> the leader is defined by:<br />
π L = Π L (x L ,β L ,k) − F (2.37)<br />
where, without loss <strong>of</strong> generality, we will assume that Π L 3 ≡ ∂Π L /∂k > 0:<br />
the variable k increases the pr<strong>of</strong>itability <strong>of</strong> the leader. The pr<strong>of</strong>it <strong>of</strong>eachother<br />
firm remains:<br />
π = Π (x, β) − F<br />
For a given strategic commitment, the second stage is characterized by<br />
the first order conditions for a Nash equilibrium. For the sake <strong>of</strong> simplicity,<br />
we follow Fudenberg <strong>and</strong> Tirole (1984) assuming that a unique equilibrium<br />
exists with symmetric strategies for all the firms except the leader <strong>and</strong> that<br />
there is entry <strong>of</strong> some followers for any feasible k. Therefore, we have the<br />
equilibrium conditions:<br />
Π L 1 (x L ,β L ,k)=0 Π 1 (x, β) =0 (2.38)<br />
In general, we will say that the investment makes the leader tough when<br />
Π L 13 > 0, that is a higher strategic investment k makes the leader more aggressive<br />
(increases x L ), <strong>and</strong> makes the followers less (more) aggressive under<br />
SS (SC). The investment makes the leader s<strong>of</strong>t when Π L 13 < 0.Inwhatfollows<br />
we will analyze many different kinds <strong>of</strong> investments, <strong>and</strong> in each application,<br />
there will be a cost for these investments. The leader will choose its investment<br />
by comparing its impact on the pr<strong>of</strong>it <strong>and</strong> its impact on the cost. Our<br />
interest, however, will be on the strategic effect, that is the effect <strong>of</strong> the<br />
investment <strong>of</strong> the leader on the behavior <strong>of</strong> the followers, defined as:<br />
SI(k) =Π L 2 (x L ,β L ,k) ∂β L<br />
∂k<br />
(2.39)
2.5 Strategic Investments 61<br />
If the cost <strong>of</strong> the strategic investment is given by some positive <strong>and</strong> increasing<br />
function f(k), the net pr<strong>of</strong>it <strong>of</strong>theleaderwillbe:<br />
π L (k) =Π L (x L ,β L ,k) − f(k) − F<br />
<strong>and</strong> the optimality condition will be:<br />
Π L 3 (x L ,β L ,k)+SI(k) =f 0 (k)<br />
It is clear that the strategic incentive is the interesting part for our purposes,<br />
since it tells us how the leader can exploit its commitment capacity in a<br />
strategic way to affect the equilibrium <strong>of</strong> the market <strong>and</strong> obtain more pr<strong>of</strong>its.<br />
To realize this, imagine what would happen if the leader could not choose k<br />
before competing with the other firms, but had to choose it simultaneously<br />
with the choice <strong>of</strong> the market strategies <strong>of</strong> all firms: then, the strategic incentive<br />
would not play any role in the choice <strong>of</strong> the investment (only the direct<br />
effect would remain). The importance <strong>of</strong> the commitment capacity relies exactly<br />
on the possibility <strong>of</strong> using the investment in a strategic way to affect<br />
the behavior <strong>of</strong> the other firms. When SI is positive we will say that there is<br />
a strategic incentive to overinvest, while when it is negative we will say that<br />
there is a strategic incentive to underinvest. Of course, overinvestment <strong>and</strong><br />
underinvestment should be thought relative to the direct incentive to invest.<br />
2.5.1 The Fudenberg-Tirole Taxonomy <strong>of</strong> Business Strategies<br />
Let us generalize the st<strong>and</strong>ard results <strong>of</strong> Fudenberg <strong>and</strong> Tirole (1984) on the<br />
strategic investment <strong>of</strong> a leader in duopoly to the case with an exogenous<br />
number <strong>of</strong> firms n. The two equilibrium first order conditions (2.38) can be<br />
easily differentiated to obtain ∂β L /∂k, <strong>and</strong> hence the strategic incentive:<br />
SI(k) = h0 (x L )Π L 2 ΠL 13 Π 12<br />
Ω<br />
(2.40)<br />
where Ω is positive by assumption <strong>of</strong> stability <strong>of</strong> the system. 23 The sign <strong>of</strong><br />
this incentive is the same as that <strong>of</strong> −Π 12 Π13, L <strong>and</strong> we have the following<br />
traditional result:<br />
Proposition 2.1. In a Nash equilibrium:<br />
1) when the strategic investment makes the leader tough (Π13 L ><br />
0), there is a strategic incentive to over- (under-) invest under<br />
strategic substitutability (complementarity);<br />
23 Here:<br />
Ω =<br />
Π11<br />
L <br />
Π11 +(n − 2) h 0 <br />
(x)Π<br />
(n − 1)h 0 12 + Π<br />
L<br />
(x)<br />
12 Π 12 > 0
62 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
2) when the strategic investment makes the leader s<strong>of</strong>t (Π13 L < 0),<br />
there is a strategic incentive to under- (over-) invest under strategic<br />
substitutability (complementarity).<br />
Now, imagine that in the absence <strong>of</strong> a strategic incentive to invest, the<br />
leader was going to choose an investment ¯k such that Π ¡ L x, β, ¯k ¢ = Π (x, β)<br />
for any x <strong>and</strong> β. 24 This is a neutrality assumption that allows to derive simple<br />
<strong>and</strong> interesting conclusions in a number <strong>of</strong> applications. It clearly implies<br />
that only the strategic incentive is going to induce the leader to behave<br />
in a different way from the other firms. In other words, only the strategic<br />
commitment can provide the leader with an advantage in the market <strong>and</strong> in<br />
the second stage we have:<br />
x L R x if <strong>and</strong> only if k R (Q)¯k when Π L 13 > ( 0). In such a case overinvestment is optimal<br />
when an aggressive behavior in the market induces a less aggressive behavior<br />
<strong>of</strong> the other firms (which requires SS: Π 12 < 0): this outcome corresponds<br />
to what has been called a “top dog” strategy in which the leading firm is<br />
aggressive to obtain non aggressive strategies <strong>of</strong> the other firms, a typical<br />
outcome <strong>of</strong> models <strong>of</strong> competition in quantities.<br />
However, when an aggressive behavior <strong>of</strong> a firm induces the other firms<br />
to be aggressive as well (which requires SC: Π 12 > 0), as in models <strong>of</strong> competition<br />
in prices, it is optimal to underinvest strategically: this corresponds<br />
to a “puppy dog” strategy where, in the words <strong>of</strong> Fudenberg <strong>and</strong> Tirole<br />
(1984), underinvestment “accommodates entry by turning the incumbent<br />
into a small, friendly, nonaggressive puppy dog.” The spirit <strong>of</strong> puppy dog<br />
strategies emerges in most models <strong>of</strong> competition in prices with product differentiation<br />
25 .Asanexample,Laffont et al. (1998) have shown that a puppy<br />
24 Within our specification <strong>of</strong> the cost function for the strategic investment, this<br />
<br />
requires Π3<br />
L x, β, ¯k = f 0 (¯k).<br />
25 As noticed by Tirole (1988), puppy dog strategies emerge in the Hotelling<br />
duopoly as well. Considering the location k 1 0 <strong>and</strong> ∂ 2 π i/∂p i∂k i<br />
is positive for firm 1 <strong>and</strong> negative for firm 2. Hencebothfirms have a strategic<br />
incentive to differentiate products.
2.5 Strategic Investments 63<br />
dog strategy emerges in (unregulated) markets for interconnected networks<br />
(for example the telecommunications industry) where an entrant chooses to<br />
invest strategically in geographical coverage before competing with the incumbent:<br />
then, the optimal strategy <strong>of</strong> the entrant is to underinvest to s<strong>of</strong>ten<br />
price competition. 26 A puppy dog behavior can emerge also in an indirect way.<br />
A typical example is a price protection policy implemented through a “mostfavored-customer<br />
clause”. This guarantees a firm’s customers that they will<br />
be reimbursed the price difference with the lowest price <strong>of</strong>fered by other firms:<br />
as shown by Tirole (1988) this policy s<strong>of</strong>tens price competition <strong>and</strong> increases<br />
pr<strong>of</strong>its.<br />
When the strategic investment makes the leader s<strong>of</strong>t (Π13 L < 0), the incentives<br />
take other directions: in the words <strong>of</strong> Fudenberg <strong>and</strong> Tirole (1984), the<br />
“fat cat strategy is overinvestment that accommodates entry by committing<br />
the incumbent to play less aggressively post entry. The lean <strong>and</strong> hungry strategy<br />
is underinvestment to be tougher.” A “lean <strong>and</strong> hungry look” emergesin<br />
case <strong>of</strong> SS (Π 12 < 0). As an example, consider our simple model <strong>of</strong> Chapter<br />
1 with competition for the market between an incumbent monopolist <strong>and</strong> an<br />
outsider. Because <strong>of</strong> the Arrow effect, the monopolist with positive pr<strong>of</strong>its<br />
from its leading technology had lower incentives to invest in innovation than<br />
the outsider, <strong>and</strong> higher current pr<strong>of</strong>its were inducing less investment by the<br />
incumbent <strong>and</strong> more by the outsider. In such a case, the incumbent would<br />
have liked to underinvest in pr<strong>of</strong>it enhancing strategies to have a strategic<br />
incentive to invest more in R&D.<br />
The “fat cat” strategy emerges in models <strong>of</strong> price competition (Π 12 ><br />
0) with a strategic investment that reduces the incentives to be aggressive,<br />
for instance, as we will see later on, with an investment in nonprice (or<br />
persuasive) advertising, which typically allows a firm to set high prices after<br />
having developed a goodwill. 27 For further discussion on the taxonomy <strong>of</strong><br />
strategic investment in duopolies, see the extensive treatment <strong>of</strong> Tirole (1988,<br />
Part II).<br />
2.5.2 Strategic Commitments with Endogenous Entry<br />
We will now follow Etro (2006,a) <strong>and</strong> assume that the number <strong>of</strong> potential<br />
entrants is great enough that a zero pr<strong>of</strong>it condition pins down the effective<br />
number <strong>of</strong> firms, n. Tobeprecise,wewilllookatthesubgameperfect<br />
equilibrium <strong>of</strong> the game with the following sequence <strong>of</strong> moves:<br />
1) in the first stage, firm L enters, pays the fixed cost F <strong>and</strong> chooses an<br />
investment k;<br />
26 See also Cambini <strong>and</strong> Valletti (2007).<br />
27 The quotation <strong>of</strong> Shakespeare from Julius Caesar (Act. 1, Sc. 2) introducing<br />
Fudenberg <strong>and</strong> Tirole (1984) is quite suggestive: “Letmehaveaboutmemen<br />
that are fat.”
64 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
2) in the second stage, after knowing the investment <strong>of</strong> the leader, all<br />
potential entrants simultaneously decide “in” or“out”: if a firm decides “in”,<br />
it pays the fixed cost F ;<br />
3) in the third stage all the firms that have entered choose their own<br />
strategy x i simultaneously.<br />
The equilibrium conditions are the two previous first order conditions<br />
(2.38), <strong>and</strong> the zero pr<strong>of</strong>it condition binding on the followers:<br />
Π (x, β) =F (2.41)<br />
We can now prove that a change in the strategic commitment by the leader<br />
does not affect the equilibrium strategies <strong>of</strong> all other firms, but reduces their<br />
equilibrium number. Let us use the definition β L ≡ (n − 1)h(x) to rewrite<br />
the equilibrium system (2.38)-(2.41) in terms <strong>of</strong> the three unknown variables<br />
x, x L <strong>and</strong> β L :<br />
Π 1 [x, h(x L ) − h(x)+β L ]=0<br />
Π1 L [x L ,β L ,k]=0<br />
Π [x, h(x L ) − h(x)+β L ]=F<br />
The second equation provides an implicit relationship x L = x L (β L ,k) with<br />
∂x L /∂β L = −Π12 L /ΠL 11 <strong>and</strong> ∂x L/∂k = −Π13 L /ΠL 11 > 0. Substituting this<br />
expression we obtain a system <strong>of</strong> two equations in two unknowns, x <strong>and</strong> β L :<br />
Π 1 [x, h(x L (β L ,k)) − h(x)+β L ]=0,<br />
Π [x, h(x L (β L ,k)) − h(x)+β L ]=F<br />
Totally differentiating the system <strong>and</strong> imposing stability, which requires<br />
Π L 11 − h 0 (x L )Π L 12 < 0, it follows that x = x(k), β L = β L (k) <strong>and</strong> x L =<br />
x L (β L (k),k) are the equilibrium functions with:<br />
dx<br />
dk =0<br />
dβ L<br />
dk = h 0 (x L )Π L 13<br />
Π L 11 − h0 (x L )Π L 12<br />
dx L<br />
dk = − Π L 13<br />
Π L 11 − h0 (x L )Π L 12<br />
<strong>and</strong> dn/dk =(dβ L /dk) /h(x). This shows that in a Marshall equilibrium, an<br />
increase in the strategic investment does not affect the equilibrium strategy<br />
<strong>of</strong> all the other firms but reduces their equilibrium number. In the initial<br />
stage, the strategic incentive becomes:<br />
SI(k) = h0 (x L )Π L 2 Π L 13<br />
Π L 11 − h0 (s)Π L 12<br />
(2.42)<br />
whose sign is just the sign <strong>of</strong> Π L 13. This delivers our main result:<br />
Proposition 2.3. In a Marshall equilibrium, when the strategic<br />
investment makes the leader tough (s<strong>of</strong>t), there is a strategic<br />
incentive to over- (under-) invest; moreover, the leader is always<br />
aggressive compared to the followers.
2.5 Strategic Investments 65<br />
Basically, whenever investment makes the leader tough (Π L 13 > 0) <strong>and</strong><br />
entry is endogenous, it is always optimal for the leader to adopt a “top<br />
dog” strategy with overinvestment in the first stage so as to be aggressive<br />
in the final stage. On the other side, when investment makes the leader s<strong>of</strong>t<br />
(Π L 13 < 0), we always have a “lean <strong>and</strong> hungry” look with underinvestment,<br />
but also in this case, the outcome in the final stage is an aggressive behavior<br />
<strong>of</strong> the leader.<br />
To underst<strong>and</strong> the intuition <strong>of</strong> this simple but general result, let us focus<br />
on the first case, in which investment makes the leader tough. Let us suppose<br />
that SC holds: this is the most interesting case because endogenous entry<br />
overturns the traditional results (but a similar mechanism works under SS<br />
as well). Under our assumptions a leader may accept the cost <strong>of</strong> underinvesting<br />
strategically (compared to the optimal direct investment) to become<br />
more accommodating, <strong>and</strong> this would be the optimal thing to do when the<br />
number <strong>of</strong> competitors is exogenous. Now, let us consider the consequences<br />
<strong>of</strong> an accommodating strategy when entry is endogenous. Since strategies<br />
are assumed to be complements, accommodation by the leader would induce<br />
accommodating strategies by the followers as well. The associated increase<br />
in expected pr<strong>of</strong>its would attract entry <strong>of</strong> other firms, which will also behave<br />
in an accommodating way. Since entry occurs as long as there are pr<strong>of</strong>itable<br />
opportunities to exploit, the followers must obtain zero pr<strong>of</strong>its in equilibrium.<br />
Therefore, the entry process induced by an accommodating strategy exhausts<br />
all possible gains for the followers. What about the leader? Its attempt to induce<br />
accommodation has the cost <strong>of</strong> distorting its strategy from the optimal<br />
direct level. Moreover, it wastes all the potential benefits from accommodation<br />
because it increases entry. Accordingly, underinvestment cannot increase<br />
the pr<strong>of</strong>its <strong>of</strong> the leader.<br />
Consider now an aggressive strategy induced by an initial overinvestment<br />
<strong>of</strong> the leader. Such a strategy may induce the rivals to be more aggressive<br />
as well, <strong>and</strong> this would reduce entry in the market. Therefore, the leader<br />
distorts its investment strategy from the directly optimal level but succeeds<br />
in reducing the negative externalities derived from the strategies <strong>of</strong> the rivals<br />
because <strong>of</strong> the reduction in their number. The optimal level <strong>of</strong> overinvestment<br />
trades <strong>of</strong>f the costs <strong>of</strong> the distortion in the investment level <strong>and</strong> the benefits<br />
<strong>of</strong> the reduction <strong>of</strong> the number <strong>of</strong> entrants.<br />
Finally, notice that the same argument would go through in the case the<br />
investment made the leader s<strong>of</strong>t, but in that case underinvestment would<br />
induce the optimal aggressive strategy.<br />
We will now apply the above results to some basic forms <strong>of</strong> strategic<br />
commitments as investments in cost reductions, advertising, financial decisions,<br />
bundling or price discrimination strategies, strategic contracts, strategic<br />
mergers <strong>and</strong> so on. There are many other applications that are not discussed<br />
in this chapter. Our focus will be limited to the applications with<br />
substantial relevance for the underst<strong>and</strong>ing <strong>of</strong> the behavior <strong>of</strong> market leaders
66 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
<strong>and</strong> for our future discussions <strong>of</strong> antitrust issues. We will emphasize how the<br />
results can drastically change according to whether we assume that entry is<br />
exogenous or endogenous, but we will mainly pay attention to the case <strong>of</strong> endogenous<br />
entry. After all, we do believe that entry <strong>of</strong> firmsisanendogenous<br />
choice in most markets, <strong>and</strong> not an exogenous fact.<br />
2.6 Cost Reductions <strong>and</strong> Signaling<br />
Our first application is to a situation where a firm can adopt preliminary<br />
investments to improve its production technology <strong>and</strong> hence reduce its costs.<br />
Traditional results on the opportunity <strong>of</strong> these investments for market leaders<br />
are ambiguous when the number <strong>of</strong> firms is exogenous, but, as we will show,<br />
they are not when entry is endogenous. From now on, we will assume for<br />
simplicity that marginal costs are constant. Here, the leader can invest k<br />
<strong>and</strong> reduce its marginal cost to c(k) > 0 with c 0 (k) < 0, while the marginal<br />
cost cannot be changed for all the other firms. One could think <strong>of</strong> the cost<br />
reducing investment as an investment in R&D to improve the production<br />
technology, but also in terms <strong>of</strong> learning by doing: past production reduces<br />
future costs. 28<br />
Consider first a model <strong>of</strong> quantity competition. The gross pr<strong>of</strong>it <strong>of</strong>the<br />
leader becomes:<br />
Π L (x L ,β L ,k)=x L p (x L ,β L ) − c(k)x L (2.43)<br />
Notice that in such a model, Π L 12 has an ambiguous sign, but we have:<br />
Π L 13 = −c 0 (k) > 0<br />
consequently the leader will overinvest in cost reductions when facing a fixed<br />
number <strong>of</strong> competitors (as long as SS holds), <strong>and</strong> will always overinvest <strong>and</strong><br />
produce more than the other firms when entry is endogenous.<br />
For instance, assume an inverse dem<strong>and</strong> p = a − X, a constant marginal<br />
cost c(k) =c − √ gk for the leader investing k, <strong>and</strong>c for the entrants, where<br />
g measures the productivity <strong>of</strong> the R&D investment, whose cost is f(k) =k.<br />
A Nash equilibrium with n firms would imply:<br />
x L =<br />
a − nc(k)+(n − 1)c a + c(k) − 2c<br />
, x =<br />
n +1<br />
n +1<br />
The optimal investment by the leader can be derived as:<br />
28 This is the typical case <strong>of</strong> the aircraft industry (Boeing, Airbus), the production<br />
<strong>of</strong> chips (Intel) <strong>and</strong> many other sectors with a fast technological progress. See<br />
Sutton (Ch. 14) for an analysis <strong>of</strong> these markets.
2.6 Cost Reductions <strong>and</strong> Signaling 67<br />
k =<br />
(a − c) 2 g<br />
[(n +1) 2 − ng] 2<br />
which clearly generates an equilibrium output for the leader that is higher<br />
than the one <strong>of</strong> the entrants (notice that SS holds in this example). The<br />
optimal investment is increasing in the productivity <strong>of</strong> the R&D technology,<br />
that is in g. Moreover, if this productivity is high enough, it is optimal to<br />
induce entry deterrence.<br />
The bias toward overinvestment in cost reducing technology aimed at an<br />
aggressive behavior in the market holds also when entry is endogenous, in<br />
which case the equilibrium production <strong>of</strong> the leader <strong>and</strong> <strong>of</strong> the entrants are:<br />
√<br />
F<br />
x L =<br />
1 − g , x = √ F<br />
<strong>and</strong> the leader induces such an equilibrium through the preliminary investment:<br />
k =<br />
gF<br />
(1 − g) 2<br />
in cost reductions. This implies the following rule for the optimal ratio between<br />
R&D spending k <strong>and</strong> sales <strong>of</strong> the leader px L :<br />
R&D<br />
Sales =<br />
g √ F<br />
(1 − g)(c + √ F )<br />
(2.44)<br />
Of course, this result requires g to be small enough, otherwise entry deterrence<br />
³<br />
would be optimal, <strong>and</strong> it would require an investment k = a − c − 3 √ ´2<br />
F /g.<br />
In this framework, the chance to undertake a strategic investment in a cost<br />
reducing technology leads to the same outcome we obtained in Section 1.2.1,<br />
when the leader could simply choose its output before the other firms <strong>and</strong><br />
marginal costs were increasing: the leader is aggressive to produce more than<br />
the other firms, but the cost <strong>of</strong> an aggressive strategy (increasing marginal<br />
costs <strong>of</strong> production there, costs <strong>of</strong> R&D investment here) limits the production<br />
<strong>of</strong> the leader. A lot <strong>of</strong> research has extended this model to the realistic<br />
case <strong>of</strong> spillovers <strong>of</strong> the R&D activity <strong>of</strong> the incumbent on the entrants (for<br />
instance, see Žigić et al., 2006 <strong>and</strong> V<strong>and</strong>ekerckhove <strong>and</strong> De Bondt, 2007),<br />
<strong>and</strong> the tendency toward overinvestment under endogenous entry holds also<br />
in that case. 29<br />
29 Assuming that investment k by the leader induces a marginal cost for the entrants<br />
c−χ √ gk, whereχ ∈ [0, 1) isameasure<strong>of</strong>thedegree<strong>of</strong>spillovers,theequilibrium<br />
with endogenous entry implies an investment:<br />
k =<br />
(1 − χ)2 gF<br />
[1 − g(1 − χ) 2 ] 2
68 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Consider now the model <strong>of</strong> price competition where the leader can invest<br />
to reduce its marginal costs in the same way <strong>and</strong> its pr<strong>of</strong>it function becomes:<br />
Π L (x L ,β L ,k)=[p L − c(k)] D (p L ,β L ) with p L =1/x L (2.45)<br />
Now we have:<br />
Π L 13 = c 0 (k)D 1 p 2 L > 0<br />
Accordingly, underinvestment in cost reductions emerges when entry is exogenous<br />
(since SC holds), but overinvestment is optimal when there is endogenous<br />
entry. Whenever this is the case, the leader wants to improve its<br />
cost function to be more aggressive in the market <strong>and</strong> sell its good at a lower<br />
price. Summarizing, we have: 30<br />
Proposition 2.4. Under both quantity <strong>and</strong> price competition<br />
with endogenous entry, a firm always has an incentive to overinvest<br />
in cost reductions <strong>and</strong> to be more aggressive than the others in the<br />
market.<br />
This theory <strong>of</strong> cost reducing investments aimed at inducing aggressive<br />
behavior toward the competitors <strong>and</strong> ultimately at decreasing prices, has<br />
been extended in a genuinely dynamic framework in an important work by<br />
Žigić et al. (2006). They depart from the static model <strong>of</strong> quantity competition<br />
analyzed above <strong>and</strong> study a dynamic duopoly in which the leader can<br />
invest over time to reduce the marginal cost gradually. The optimal accommodating<br />
strategy generates an increasing investment associated with a decreasing<br />
price. The optimal entry deterring strategy requires a heavy initial<br />
investment able to deter entry as soon as possible, <strong>and</strong> a lower investment<br />
in the subsequent monopolistic phase, which generates a decreasing price in<br />
the predatory phase <strong>and</strong> an increasing price in the monopolistic phase. The<br />
predatory strategy is optimal when the investment is productive enough (g<br />
is high enough) <strong>and</strong> the speed <strong>of</strong> adjustment <strong>of</strong> the marginal cost (namely <strong>of</strong><br />
its reduction with the investment) is high enough. However, the surprising<br />
that is decreasing in the spillovers, which dissipate R&D effort from the perspective<br />
<strong>of</strong> the leader. Only when spillovers are small enough (χ
2.6 Cost Reductions <strong>and</strong> Signaling 69<br />
result is that the sharp decrease in the equilibrium price due to the predatory<br />
investment in R&D leads to permanent gains for the consumers also in the<br />
monopolistic phase after predation (when potential entry still constrains the<br />
R&D activity).<br />
Our results can also be used to re-interpret models <strong>of</strong> predatory pricing<br />
through cost signaling. In a classic work <strong>of</strong> the modern industrial organization<br />
(<strong>and</strong> <strong>of</strong> the post-Chicago approach to antitrust), Milgrom <strong>and</strong> Roberts (1982)<br />
have studied the entry decision <strong>of</strong> an entrant in a duopoly with an incumbent<br />
that is already active in the market, <strong>and</strong> have introduced incomplete information:<br />
since the study <strong>of</strong> informational asymmetries is beyond the scope <strong>of</strong><br />
this book, we will just sketch their idea to emphasize the similarities with<br />
our approach. Imagine that the entrant does not know the cost <strong>of</strong> the leader,<br />
which can be a high cost or a low cost, but would like to enter only when<br />
facing a high cost leader. Milgrom <strong>and</strong> Roberts study under which conditions<br />
preliminary strategies <strong>of</strong> the leader induce entry deterrence. For instance, a<br />
low cost leader can signal its own efficiency through initial over-production<br />
or under-pricing (associated with a sacrifice <strong>of</strong> pr<strong>of</strong>its)aslongasthisisrelatively<br />
cheaper for the low cost leader compared to the high cost one. This<br />
sorting or single crossing condition, first pointed out by Spence (1974) in a<br />
different context, 31 is respected here exactly because the marginal pr<strong>of</strong>itability<br />
<strong>of</strong> production decreases with the marginal cost. In our terminology, this<br />
corresponds exactly to our condition Π L 13 > 0: when the marginal cost is<br />
lower (c(k) is lower because the investment k is higher), the marginal benefits<br />
<strong>of</strong> an aggressive strategy is higher. This means that the marginal cost<br />
<strong>of</strong> an aggressive strategy is lower for a low cost firm. Then, in a separating<br />
equilibrium, a low cost leader is initially aggressive overproducing enough to<br />
signal its efficiency <strong>and</strong> induce the follower not to enter, while a high cost<br />
leader does not imitate such a strategy because it is more pr<strong>of</strong>itable to behave<br />
monopolistically initially <strong>and</strong> accommodate entry subsequently. This result<br />
shows that cost reductions can have a strategic role also in the presence <strong>of</strong><br />
incomplete information about costs. 32<br />
Notice that even without exclusionary purposes, a leader may like to signal<br />
its own type to affect post-entry competition with incomplete information on<br />
costs. Under competition in quantities (<strong>and</strong> SS), a low cost leader may signal<br />
its efficiency to reduce the equilibrium output <strong>of</strong> the entrant <strong>and</strong> increase<br />
its own, but under price competition it is a high cost leader that wants to<br />
signal its inefficiency to induce high prices by the entrant <strong>and</strong> obtain high<br />
31 The initial application was to the signaling <strong>of</strong> productivity through higher education<br />
(which requires a lower relative effort for more productive agents). For an<br />
introduction to the economics <strong>of</strong> asymmetric information see Tirole (1988, Ch.<br />
9), Hirshleifer <strong>and</strong> Riley (1992) <strong>and</strong> Laffont <strong>and</strong> Tirole (1993).<br />
32 When the probability that the leader is low cost is high enough a pooling equilibrium<br />
occurs. In such a case, the high cost leader produces the same monopolistic<br />
output <strong>of</strong> the low cost leader, <strong>and</strong> the entrant does not enter anyway.
70 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
pr<strong>of</strong>its for both, a point first made by Fudenberg <strong>and</strong> Tirole (1984). Without<br />
developing the argument in technical details, we can point out that when<br />
entry is endogenous there can only be a gain from signaling efficiency for<br />
a low cost incumbent, since signaling a high cost would not s<strong>of</strong>ten price<br />
competition, but just induce further entry. In the spirit <strong>of</strong> our model, we<br />
can conclude by suggesting that also under incomplete information about<br />
costs, there is a role for a positive strategic investment in cost reductions (for<br />
signaling purposes) whenever entry in the market is endogenous. And this<br />
does not necessarily imply exclusionary aims.<br />
2.7 Advertising <strong>and</strong> Dem<strong>and</strong> Enhancing Investments<br />
We will now consider investments which affect the dem<strong>and</strong> function <strong>of</strong> a<br />
firm, such as nonprice advertising (aimed at br<strong>and</strong> positioning <strong>and</strong> at enhancing<br />
the goodwill), <strong>and</strong> investments for quality improvements or product<br />
differentiation. These investments tend to increase dem<strong>and</strong> <strong>and</strong> also reduce<br />
the substitutability between goods. 33 Under endogenous entry, the aim <strong>of</strong><br />
the leader is always to be aggressive in the market, but different strategies<br />
emerge under quantity <strong>and</strong> price competition.<br />
Consider a model <strong>of</strong> quantity competition characterized by the inverse<br />
dem<strong>and</strong> p (x L ,β L ,k) for the leader. The marginal effect <strong>of</strong> investment on<br />
inverse dem<strong>and</strong> is positive (p 3 > 0), while the one on its slope is negative<br />
(p 13 < 0), which implies that a higher investment not only increases dem<strong>and</strong>,<br />
but it also makes it more rigid. 34 In this case, its gross pr<strong>of</strong>it becomes:<br />
Π L (x L ,β L ,k)=x L [p (x L ,β L ,k) − c] (2.46)<br />
Consequently, we have:<br />
Π L 13 = p 3 (1 − η)<br />
where η ≡−x L p 31 /p 3 is the elasticity <strong>of</strong> the marginal effect <strong>of</strong> investment on<br />
price with respect to production. As long as this elasticity is less than unitary,<br />
which means that investment does not make dem<strong>and</strong> too rigid, we have Π L 13 ><br />
33 See Tirole (1988, Ch. 2 <strong>and</strong> Ch. 7) on product selection, quality <strong>and</strong> advertising,<br />
<strong>and</strong>onproductdifferentiation.<br />
34 This may not be the case for informative advertising (which informs consumers<br />
abour product price <strong>and</strong> availability) or other forms <strong>of</strong> investment that attract<br />
marginal consumers. Since these consumers are by definition more sensitive<br />
to price changes, the investment may increase both dem<strong>and</strong> <strong>and</strong> its elasticity<br />
(Becker <strong>and</strong> Murphy, 1993). In general, marketing studies suggest that investments<br />
in advertising make dem<strong>and</strong> more rigid for a price increase <strong>and</strong> more<br />
elastic for a price decrease (Kotler, 1999). A classic work in the field is Lambin<br />
(1970).
2.7 Advertising <strong>and</strong> Dem<strong>and</strong> Enhancing Investments 71<br />
0. While under exogenous entry the investment choice <strong>of</strong> the leader depends<br />
on many factors, under endogenous entry overinvestment takes place if <strong>and</strong><br />
only if η 0<br />
<strong>and</strong> D 13 > 0 <strong>and</strong> the gross pr<strong>of</strong>it becomes:<br />
Π L (x L ,β L ,k)=(p L − c) D (p L ,β L ,k) with p L =1/x L (2.48)<br />
where the crucial cross effect is:<br />
Π L 13 = − [D 3 +(p L − c)D 13 ] p 2 L < 0<br />
In this case with an exogenous number <strong>of</strong> firms the leader would overinvest<br />
to increase its price <strong>and</strong> exploit the induced increase in the price <strong>of</strong> the<br />
competitors. However, under endogenous entry the behavior <strong>of</strong> the leader<br />
radically changes <strong>and</strong> there is always underinvestment so as to reduce the<br />
pricebelowtheprice<strong>of</strong>thefollowers. 36<br />
35 The model can also be reinterpreted in terms <strong>of</strong> product differentiation. It is well<br />
known that, from the 1950s to the 1970s in US, established firms in the readyto-eat<br />
breakfast cereal industry rapidly increased the number <strong>of</strong> the br<strong>and</strong>s they<br />
<strong>of</strong>fered with aggressive purposes against further entry in the market.<br />
36 Vertical differentiation is another way to interpret our model. For instance, if dem<strong>and</strong><br />
depends on the price-quality ratio, according to some function ˆD(p L/k, β L )<br />
where k is quality, it is easy to derive Π L 13 < 0: committing to a high quality<br />
leads to choose high prices. Nevertheless, in Section 3.4.4 we will study a more<br />
realistic situation in which committing to high quality is the best strategy for a<br />
leader facing endogenous entry.
72 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Fudenberg <strong>and</strong> Tirole (1984) have introduced another simple example<br />
<strong>of</strong> investment in advertising that is nested in our framework <strong>and</strong> is derived<br />
from Schmalensee (1982). Imagine that firms compete in prices on the same<br />
customers, but the leader, through a costly investment in advertising k, can<br />
obtain an extra dem<strong>and</strong> D(k) from new customers, with D 0 (k) > 0. This<br />
simple stylized set up delivers a pr<strong>of</strong>it function for the leader:<br />
Π L (x L ,β L ,k)=(p L − c) D(k)+(p L − c) D (p L ,β L )<br />
with p L =1/x L<br />
while the pr<strong>of</strong>its for the other firmsarethesameasbefore.Thecrosseffect<br />
is now Π13 L = −D 0 (k)p 2 L < 0. Hence, as Fudenberg <strong>and</strong> Tirole (1984) noticed<br />
inthecase<strong>of</strong>tw<strong>of</strong>irms, “if the established firm chooses to allow entry, it<br />
will advertise heavily <strong>and</strong> become a fat cat in order to s<strong>of</strong>ten the entrant’s<br />
pricing behavior”, but, we add, when entry <strong>of</strong> firms is endogenous, the leader<br />
will underinvest in advertising to keep low prices while allowing some firms<br />
to enter in the market. Summarizing our results for nonprice advertising, we<br />
have:<br />
Proposition 2.5. Under quantity competition with endogenous<br />
entry, a firm has an incentive to overinvest in nonprice advertising<br />
as long as this does not make dem<strong>and</strong> too rigid; under price competition<br />
with endogenous entry the leader has always an incentive<br />
to underinvest in nonprice advertising.<br />
Once again this result overturns common wisdom obtained by duopoly<br />
models, especially under price competition.<br />
2.8 Debt <strong>and</strong> the Optimal Financial Structure<br />
We can also apply our results to the theory <strong>of</strong> corporate finance to study<br />
the strategic role <strong>of</strong> the financial structure. As shown by Br<strong>and</strong>er <strong>and</strong> Lewis<br />
(1986, 1988) <strong>and</strong> Showalter (1995, 1999) in models <strong>of</strong> duopolies with uncertainty,<br />
when product decisions are managed by the equity holders, debt can<br />
affect the marginal pr<strong>of</strong>itability, <strong>and</strong> hence there can be a role for a bias in<br />
the optimal financial structure, departing from the st<strong>and</strong>ard neutrality results<br />
<strong>of</strong> Modigliani <strong>and</strong> Miller (1958). 37 The outcome depends on the kind <strong>of</strong><br />
competition, but also on the kind <strong>of</strong> uncertainty.<br />
For finance to play a role in product market competition, we need to<br />
introduce uncertainty on pr<strong>of</strong>its. Imagine that the total financing requirement<br />
37 See Tirole (2006, Ch. 7) for a survey on the relation between corporate finance<br />
<strong>and</strong> product market competition, <strong>and</strong> Brealey <strong>and</strong> Myers (2002) for a general<br />
introduction to the theory <strong>of</strong> the optimal financial structure. Between many<br />
empirical analysis on alternative financing tools in different contexts, see the<br />
recent work <strong>of</strong> Cenciarini et al. (2006).
2.8 Debt <strong>and</strong> the Optimal Financial Structure 73<br />
for each firm is fixed <strong>and</strong> each firm has enough cash to finance production<br />
entirely without issuing debt. Furthermore, suppose that the credit market<br />
is perfectly competitive, so that lenders break even. In such a context, the<br />
Modigliani-Miller neutrality result holds only if the financial structure does<br />
not affectproductmarketcompetition.<br />
For simplicity, we will assume that the financial structure <strong>of</strong> the outsiders<br />
implies no debt. The leader, however, can adopt a different financial structure<br />
by issuing positive debt at a preliminary stage. Afterward, the equity holders<br />
<strong>of</strong> all firms choose their market strategies, uncertainty is resolved <strong>and</strong> pay<strong>of</strong>fs<br />
for equity holders <strong>and</strong> debt holders are assigned. Assume that the pr<strong>of</strong>it<br />
functions are disturbed by a r<strong>and</strong>om shock z ∈ [z¯, ¯z] independently <strong>and</strong> identically<br />
distributed according to the cumulative function G(z) with density<br />
g(z). The initial ownership <strong>of</strong> the leading firm can decide its debt level k to<br />
be repaid out <strong>of</strong> gross pr<strong>of</strong>its, if these are sufficient. Once this choice is taken,<br />
competition takes place, uncertainty is solved <strong>and</strong> each firm obtains its own<br />
pr<strong>of</strong>its net <strong>of</strong> the debt or goes bankrupt.<br />
If the gross pr<strong>of</strong>its <strong>of</strong> the leader can be written as R(x L ,β L ,z) with the<br />
usual notation, the value <strong>of</strong> equity, corresponding to the expected pr<strong>of</strong>its net<br />
<strong>of</strong> debt repayment can be written as:<br />
E(k) =Π L (x L ,β L ,k) − F =<br />
Z¯z<br />
ẑ<br />
[R(x L ,β L ,z) − F − k] g(z)dz (2.49)<br />
where the lower bound ẑ is such that gross pr<strong>of</strong>its are zero:<br />
R(x L ,β L , ẑ) − F = k<br />
Notice that dẑ/dk =1/R z (x L ,β L , ẑ). We assume usual properties for the<br />
pr<strong>of</strong>it function (R xx (x L ,β L ,z) < 0), <strong>and</strong> we also assume, without loss <strong>of</strong><br />
generality, that the r<strong>and</strong>om variable is chosen so that R z (x i ,β i ,z) > 0: this<br />
implies that the cut-<strong>of</strong>f level <strong>of</strong> the positive shock ẑ below which bankruptcy<br />
occurs is increasing in the debt level (dẑ/dk > 0).<br />
We could think <strong>of</strong> a model <strong>of</strong> competition in quantities where:<br />
R(x i ,β i ,z)=x i p(x i ,β i ,z) − c(x i ,z)<br />
with p z (x i ,β i ,z) > 0 <strong>and</strong> c z (x i ,z) < 0: a positive shock increases dem<strong>and</strong><br />
or reduces costs. In case <strong>of</strong> dem<strong>and</strong> uncertainty with the stochastic linear<br />
dem<strong>and</strong> p = z − X x j <strong>and</strong> zero marginal costs, we would have ẑ =(k +<br />
F )/x L + X x j , which is <strong>of</strong> course increasing in the debt level. In this example<br />
<strong>and</strong> generally under weak conditions, a positive shock increases the marginal<br />
pr<strong>of</strong>itability <strong>of</strong> production (R xz (x L ,β L ,z) > 0). We can also have a model <strong>of</strong><br />
competition in prices with:<br />
R(x i ,β i ,z)=[p i − c(z)] D (p i ,β i ,z)<br />
with p L =1/x L
74 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
<strong>and</strong> we allow explicitly for an impact <strong>of</strong> uncertainty on both dem<strong>and</strong> <strong>and</strong><br />
costs. Our assumptions are compatible with D z (1/x i ,β i ,z) > 0 <strong>and</strong> c z (z) <<br />
0: a positive shock increases dem<strong>and</strong> <strong>and</strong>/or reduces costs. Moreover, under<br />
mild conditions assumed in what follows, a positive dem<strong>and</strong> shock increases<br />
the marginal pr<strong>of</strong>itability <strong>of</strong> a price increase (R xz (x L ,β L ,z) < 0), while a<br />
positive cost shock always decreases it (R xz (x L ,β L ,z) > 0).<br />
In general we have:<br />
Π L 1 (x L ,β L ,k)=<br />
Z¯z<br />
ẑ<br />
R x (x L ,β L ,z)g(z)dz − [R(x L ,β L , ẑ) − k] dẑ<br />
dk<br />
whose last term is zero by the definition <strong>of</strong> ẑ. In any equilibrium, the optimal<br />
behavior <strong>of</strong> each firm would require that the expectation <strong>of</strong> its marginal pr<strong>of</strong>it<br />
issetequaltozero.Butnoticethatwhatisrelevantforafirm with a positive<br />
debt are the expected pr<strong>of</strong>its conditional on these being positive after debt<br />
repayment, <strong>and</strong> this affects substantially the marginal pr<strong>of</strong>its as well. When<br />
R xz (x L ,β L ,z) is positive, marginal pr<strong>of</strong>it increasesinẑ <strong>and</strong> hence in the debt<br />
level, <strong>and</strong> the opposite happens when R xz (x L ,β L ,z) is negative. As always,<br />
it is crucial to derive the sign <strong>of</strong> the cross effect: 38<br />
Π13 L (x L ,β L ,k)=−R x (x L ,β L , ẑ) dẑ<br />
dk =<br />
= −R x(x L ,β L , ẑ)<br />
R 0 if R xz (x L ,β<br />
R z (x L ,β L , ẑ)<br />
L ,z) R 0<br />
This implies that when the number <strong>of</strong> firms is exogenous <strong>and</strong> the leader<br />
accommodates entry, under SS there is a strategic incentive to issue debt<br />
when a positive shock increases marginal pr<strong>of</strong>its (R xz (x L ,β L ,z) > 0) <strong>and</strong><br />
under SC in the opposite case (R xz (x L ,β L ,z) < 0). For instance, under<br />
competition in quantities there is typically a strategic role for debt financing<br />
(Br<strong>and</strong>er <strong>and</strong> Lewis, 1986), while under competition in prices there is a role<br />
for debt financing only in the presence <strong>of</strong> dem<strong>and</strong> uncertainty, but not in case<br />
<strong>of</strong> cost uncertainty (Showalter, 1995). 39 Things are different, however, when<br />
entry takes place endogenously until expected pr<strong>of</strong>its are zero. In this case<br />
we can apply Prop. 2.3 <strong>and</strong> conclude with:<br />
Proposition 2.6. Under endogenous entry, a firm has an incentive<br />
to adopt debt financing to be more aggressive in the competition<br />
whenever a positive shock increases marginal pr<strong>of</strong>its.<br />
38 The sign <strong>of</strong> the marginal pr<strong>of</strong>it atitsboundsẑ <strong>and</strong> ¯z depends on the sign <strong>of</strong><br />
R xz (x L ,β L ,z). InparticularR x (x L ,β L , ẑ) Q 0 if R xz (x L ,β L ,z) R 0. Forfurther<br />
details see Etro (2006e). Notice that a bias toward debt financing is equivalent to<br />
a bias toward risk-taking behavior, a well known consequence <strong>of</strong> debt contracts<br />
(at least since Stiglitz <strong>and</strong> Weiss, 1981).<br />
39 Debt financing to deter entry can emerge with quantity competition <strong>and</strong> SS or<br />
with price competition <strong>and</strong> cost uncertainty (Showalter, 1999).
2.8 Debt <strong>and</strong> the Optimal Financial Structure 75<br />
In general, under quantity competition there is always a strategic bias<br />
toward debt financing, while under price competition the same bias emerges<br />
only when uncertainty affects costs, but not when it affects dem<strong>and</strong>. The<br />
intuition is again related with the role <strong>of</strong> debt financing in inducing a more<br />
aggressive behavior in the market, which is always desirable for the leader<br />
facing endogenous entry. Under quantity competition, debt induces the management<br />
to care only about the good states <strong>of</strong> the world (high dem<strong>and</strong> <strong>and</strong><br />
low costs) <strong>and</strong> therefore to choose aggressive strategies. Similarly, under price<br />
competition <strong>and</strong> dem<strong>and</strong> uncertainty a higher debt increases the marginal<br />
pr<strong>of</strong>itability <strong>of</strong> a higher price strategy. Accordingly, it helps implementing a<br />
more accommodating strategy in the market: just what a leader would like to<br />
do when facing exogenous entry, but the opposite <strong>of</strong> what would be desirable<br />
in front <strong>of</strong> endogenous entry. However, under cost uncertainty, the management<br />
decides the price to maximize pr<strong>of</strong>its conditional on a good state <strong>of</strong> the<br />
world, meaning low costs, which leads to a bias toward low prices: this is a<br />
suboptimal strategy with exogenous entry, but an optimal one with endogenous<br />
entry. 40<br />
To complete our analysis, notice that the initial ownership would actually<br />
choose debt to maximize the overall value <strong>of</strong> the firm, which corresponds to<br />
the equity value E (k) plus the debt value:<br />
D(k) =<br />
Z ẑ<br />
z<br />
¯<br />
[R(x L ,β L ,z) − F ] g(z)dz + k[1 − G(ẑ)]<br />
where the first term represents the expected repayment in the case <strong>of</strong> bankruptcy<br />
<strong>and</strong> the second one the expected repayment in case <strong>of</strong> successful outcome<br />
for the firm. Taking into account the dependence <strong>of</strong> the equilibrium on<br />
debt k, the value <strong>of</strong> the firm is then:<br />
V(k) =E(k)+D(k) =<br />
Z¯z<br />
z<br />
¯<br />
R [x L (k),β L (k),z] g(z)dz − F (2.50)<br />
which corresponds to the expected pr<strong>of</strong>its <strong>of</strong> the firm. When a positive shock<br />
increases the marginal pr<strong>of</strong>itability <strong>of</strong> an aggressive strategy, the optimal financial<br />
structure requires an amount <strong>of</strong> debt k ∗ that induces the management<br />
to behave as a Stackelberg leader in front <strong>of</strong> the other firms - as we will see<br />
40 Chevalier (1995) examines changes in supermarket prices in local markets after<br />
“leverage buyouts” <strong>and</strong> finds that prices decrease following an LBO in front <strong>of</strong><br />
rival firms which are not highly leveraged, while they increase when the LBO<br />
firm’s rivals are also highly leveraged. She associates the former result to predatory<br />
strategies <strong>and</strong> the latter to a s<strong>of</strong>tening <strong>of</strong> price competition, but she does<br />
not control for the endogeneity <strong>of</strong> entry in these local markets, which makes hard<br />
to evaluate the results.
76 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
in the next chapter, this is the best equilibrium the leader can aim for; more<br />
debt would induce an excessively aggressive strategy. When a positive shock<br />
decreases the marginal pr<strong>of</strong>itability <strong>of</strong> an aggressive strategy, the financial<br />
structure cannot improve the performance <strong>of</strong> the firm: in this case, for instance<br />
with price competition <strong>and</strong> dem<strong>and</strong> uncertainty, the optimal financial<br />
structure requires no debt. Summing up, the optimal ratio between the value<br />
<strong>of</strong> debt <strong>and</strong> the value <strong>of</strong> equity can be defined as:<br />
∙ ¸<br />
Debt<br />
Equity =max 0, D(k∗ )<br />
E(k ∗ (2.51)<br />
)<br />
Notice that this rule has been derived assuming a perfectly competitive credit<br />
market, free entry in the product market, no taxes <strong>and</strong> no bankruptcy costs,<br />
exactly as for the Modigliani-Miller theorem; further generalizatins could be<br />
considered. 41<br />
2.9 Network Externalities <strong>and</strong> Two-Sided <strong>Market</strong>s<br />
Many markets are characterized by network externalities, in the sense that<br />
dem<strong>and</strong> is enhanced by past production <strong>and</strong> the consequent diffusion <strong>of</strong> the<br />
product across customers. This may happen for cultural or social reasons,<br />
for instance because goods become fashionable when they have been already<br />
chosen by other customers, or because <strong>of</strong> technological reasons, for instance<br />
because the willingness to pay for a good by each consumer depends on how<br />
many other consumers have the same good. The last situation is typical <strong>of</strong><br />
advanced technological markets: in principle we may attach a high value<br />
to video phone communication, but until many <strong>of</strong> our friends will have a<br />
video phone, we are unlikely to attach a high value to owning one as well.<br />
The classic study <strong>of</strong> competition in this kind <strong>of</strong> markets is due to Katz <strong>and</strong><br />
Shapiro (1985). 42 Here we will focus on a more stylized model <strong>of</strong> the behavior<br />
<strong>of</strong> market leaders in the presence <strong>of</strong> network externalities.<br />
We will adopt the simplest model <strong>of</strong> quantity competition with homogeneous<br />
goods <strong>and</strong> introduce a time dimension. Imagine that in a first period the<br />
leader is alone in the market <strong>and</strong> produces k facing the inverse dem<strong>and</strong> p(k)<br />
<strong>and</strong> a marginal cost c. In the second period other firms compete in quantities<br />
<strong>and</strong> the leader faces the inverse dem<strong>and</strong> p(X)φ(k), whereX is total<br />
41 The model could be extended introducing bankruptcy costs <strong>and</strong> adding multiple<br />
periods to examine dynamic strategies for entry deterrence: as shown by a<br />
wide literature on the so-called “long purse” or “deep pocket” theory <strong>of</strong> predation,<br />
when initial aggressive strategies by the incumbent reduce the financing<br />
opportunities <strong>of</strong> the entrants, financial predation can indeed be optimal (see<br />
Holmstrom <strong>and</strong> Tirole 1997, Hart, 1995, <strong>and</strong> Tirole, 2006).<br />
42 See also Amir <strong>and</strong> Lazzati (2007).
2.9 Network Externalities <strong>and</strong> Two-Sided <strong>Market</strong>s 77<br />
production <strong>and</strong> φ(k) is some increasing function <strong>of</strong> past production, which<br />
is a measure <strong>of</strong> the diffusion <strong>of</strong> the good between consumers, <strong>and</strong> induces<br />
network externalities. The gross pr<strong>of</strong>it function for the leader becomes:<br />
Π L (x L ,β L ,k)=p(k)k − ck + δ [p (X)φ(k) x L − cx L ] (2.52)<br />
where δ ≤ 1 is the discount factor, while the net pr<strong>of</strong>it <strong>of</strong>theotherfirms is<br />
simply π i = x i p(X) − cx i − F . Since the other firms do not enjoy network<br />
effects, one can easily show that in a free entry equilibrium the future production<br />
x L (k) <strong>of</strong> the leader will be increasing in its initial production with<br />
∂x L /∂k = −cφ 0 (k)/φ(k) 2 p 0 (X) > 0. 43 Moreover, in equilibrium we have the<br />
cross effect:<br />
Π13 L = δcφ0 (k)<br />
> 0<br />
φ(k)<br />
which, according to our general principle, shows that the leader will always<br />
engage in initial overproduction to be more aggressive when the market opens<br />
up to endogenous entry. We can also derive a simple expression for the optimal<br />
initial production:<br />
p(k)+kp 0 (k) =c − δp ¡ X)φ 0 (k ¢ x L (k) − δ cφ0 (k)x L (k)<br />
φ(k)<br />
(2.53)<br />
This rule equates the marginal revenue <strong>of</strong> initial production to its effective<br />
marginal cost, which includes the myopic marginal cost c, a second term that<br />
represents the direct benefit due to the network effects on future dem<strong>and</strong><br />
(determining what is sometimes called a penetration price), <strong>and</strong> a last term<br />
representing the indirect (strategic) benefits due to the commitment to the<br />
adoption <strong>of</strong> a more aggressive strategy in the future. Notice that in the presence<br />
<strong>of</strong> network externalities, an incumbent expecting strong competition in<br />
the market may want to price well below marginal cost not with the purpose<br />
<strong>of</strong> excluding any other firm to enter in the market, but to be able to compete<br />
aggressively in the future: this is more likely when the marginal costs <strong>of</strong><br />
production are low <strong>and</strong> the discount factor is high. Summarizing we have:<br />
Proposition 2.7. In markets with network externalities an incumbent<br />
has an incentive to overproduce initially so as to be more<br />
aggressive when endogenous entry takes place in the future.<br />
Themodelabove,canbere-interpretedinaninterestingwaywhenwe<br />
assume that the externality function is simply φ(k) =k. Thisimpliesthatnet<br />
43 We focus on an interior equilibrium, but it is clear that a corner solution can<br />
emerge: such a tipping equilibrium is actually typical in markets with network<br />
effects (see Cremer et al., 2000).
78 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
pr<strong>of</strong>its in the competitive market are proportional to kx L .T<strong>of</strong>ix ideas, imagine<br />
that the firms under consideration produce local newspapers. The leader<br />
decides a capacity production for k copies <strong>of</strong> its local newspaper, but also<br />
sells advertising space on the newspaper in quantity x L <strong>and</strong> in competition<br />
with other newspapers (located elsewhere <strong>and</strong> with their own local readers).<br />
Of course, advertising is more valuable when a newspaper has more readers,<br />
<strong>and</strong> more precisely what matters is exactly the number <strong>of</strong> interactions between<br />
readers <strong>and</strong> advertisement, which is simply k · x L .Thisisthesimplest<br />
example <strong>of</strong> a two-sided market because newspapers sell two products (news<br />
<strong>and</strong> advertising) to different customers, <strong>and</strong> there are network effects between<br />
them (actually only in one direction in this example, since we assumed that<br />
readers are indifferent to the size <strong>of</strong> advertisement space on the newspapers).<br />
As first pointed out by Rochet <strong>and</strong> Tirole (2003) <strong>and</strong> Armstrong (2006),<br />
in such a two-sided market firms charge the different sides in different ways<br />
with the aim <strong>of</strong> enhancing network effects: in general the aim is to get on<br />
board many agents from the side whose size creates more value for the other<br />
side. In our example, for instance, the direct effect <strong>of</strong> the sales <strong>of</strong> newspapers<br />
(<strong>and</strong> maybe related bundled gadgets) on the pr<strong>of</strong>its from advertising induces<br />
a production beyond the myopic monopolistic output level. However, here we<br />
want to point out a new strategic element: a leader facing competition on one<br />
side (advertising), will have an additional indirect incentive to overproduce<br />
on the other side (newspapers), to enhance the value <strong>of</strong> the platform <strong>and</strong> to<br />
be aggressive in the competition with other firms (for the advertising). 44<br />
Similar situations emerge in many multi-sided markets where platforms<br />
compete on the volume <strong>of</strong> transactions between different groups <strong>of</strong> buyers <strong>and</strong><br />
sellers (think <strong>of</strong> credit cards, operating systems) 45 <strong>and</strong> multiple factors can<br />
44 One can verify that the same happens under price competition, which is the usual<br />
assumption in models <strong>of</strong> two-sided markets. However, under SC, overproduction<br />
by the leader is strictly related with the endogeneity <strong>of</strong> entry. When the number<br />
<strong>of</strong> competitors is exogenous, a leader would like to commit to (relatively) high<br />
prices for the newspapers so as to be accommodating in the competition for<br />
advertising space against other newspapers: only when entry is endogenous the<br />
need <strong>of</strong> being aggressive in the advertising market induces to price newspapers<br />
at a (relatively) low price. See Section 6.1.2 for further discussion.<br />
45 For instance, consider a variant <strong>of</strong> the previous example where both sides are<br />
now charged for each interaction, <strong>and</strong> c is the marginal cost <strong>of</strong> an interaction, so<br />
that:<br />
Π L (x L,β L ,k)=[p(k)+p(X) − c] · k · x L<br />
In case the leader is just a monopolist, k <strong>and</strong> x would be chosen to satisfy the<br />
Rochet-Tirole (2003) optimality condition:<br />
p(k)+p(x) − c = p(k)+p(x)<br />
(k)+(x) = p(k)<br />
(k) = p(x)<br />
(x)
2.10 Bundling 79<br />
induce different strategic behavior toward different sides. <strong>Market</strong> relations<br />
easily become complex when network effects act in both directions (in the<br />
case <strong>of</strong> informative advertising, readers may have positive externalities from<br />
more advertising in the newspapers), <strong>and</strong> especially when one or both sides<br />
engage in multi-homing (in case <strong>of</strong> national newspapers, readers may read<br />
more than one <strong>of</strong> them). In Chapter 6 we will discuss some <strong>of</strong> these issues<br />
within concrete applications.<br />
2.10 Bundling<br />
There has been a lot <strong>of</strong> attention in the economic literature on the rationale<br />
for bundling products rather than selling them separately. 46 A fundamental<br />
reason for this is that many antitrust cases have focused on such a practice as<br />
an anti-competitive one. Therefore, in this section we will try to underst<strong>and</strong><br />
when market leaders adopt bundling as a strategic device for exclusionary<br />
purposes.<br />
According to the traditional leverage theory <strong>of</strong> tied good sales, monopolists<br />
would bundle their products with others for competitive or partially<br />
competitive markets to extend their monopolistic power. Such a view as<br />
been criticized by the Chicago school (Bork, 1993, Posner, 2001) because<br />
it would erroneously claim that a firm can artificially increase monopolistic<br />
pr<strong>of</strong>its from a competitive market. Bundling should have different motivations,<br />
as price discrimination or creation <strong>of</strong> joint economies, whose welfare<br />
consequences are ambiguous <strong>and</strong> sometimes even positive. Whinston (1990)<br />
has changed the terms <strong>of</strong> the discussion trying to verify how a monopolist can<br />
affect the strategic interaction with its competitors in a secondary market by<br />
bundling. His main finding is that bundling tends to strengthen price competition<br />
against these competitors, therefore the only reason why a monopolist<br />
could bundle is to deter entry in the secondary market. However, here we<br />
will show that, when entry is endogenous, bundling may become the optimal<br />
“top dog” (aggressive) strategy.<br />
where (x) =−p(x)/xp 0 (x) is the elasticity <strong>of</strong> dem<strong>and</strong>: the side whose dem<strong>and</strong><br />
is more elastic should be charged relatively more because this keeps dem<strong>and</strong><br />
on both sides balanced <strong>and</strong> maximizes the volume <strong>of</strong> interactions for a given<br />
total price. Now, imagine that the leading platform competes on one side, but<br />
can commit to output k on the other side. Since Π13<br />
L = p 0 (k)k
80 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Imagine that a monopolistic market is characterized by zero costs <strong>of</strong> production<br />
<strong>and</strong> unitary dem<strong>and</strong> at price v, which corresponds to the valuation<br />
<strong>of</strong> the good alone. Another market is characterized by st<strong>and</strong>ard price competition,<br />
a fixed cost F <strong>and</strong> a constant marginal cost c. Grosspr<strong>of</strong>its for the<br />
monopolist without bundling are:<br />
π M = v +(p M − c) D (p M ,β M ) − F (2.54)<br />
while pr<strong>of</strong>its for the other firms are π i =(p i − c) D (p i ,β i ) − F .InBertr<strong>and</strong><br />
equilibrium with endogenous entry the monopolist enjoys the pr<strong>of</strong>its π M = v.<br />
Under bundling, dem<strong>and</strong> for the monopolist is constrained by dem<strong>and</strong><br />
for the other good, which is assumed less than unitary. The bundle price<br />
corresponds to P M = v 0 + p M ,wherev 0 ≥ v is the valuation <strong>of</strong> the primary<br />
good when bundled with a secondary good <strong>of</strong> the same firm: this maybe<br />
higher for efficiency reasons, complementarities or network externalities <strong>of</strong><br />
different kind. In such a case, the pr<strong>of</strong>its for the monopolist become:<br />
π MB =(P M − c) D(P M − v 0 ,β M ) − F 0 =(p M + v 0 − c) D (p M ,β M ) − F 0<br />
where F 0 ≤ F is the fixed cost <strong>of</strong> production in case <strong>of</strong> bundling: this may<br />
also be lower than before because <strong>of</strong> cost efficiencies. The other firms have<br />
the same objective function as before. In Bertr<strong>and</strong> equilibrium the monopolist<br />
chooses the price P M = p M + v 0 satisfying:<br />
(P M − c)D 1 [p M , (n − 1)g(p)] + D [p M , (n − 1)g(p)] = 0 (2.55)<br />
while each one <strong>of</strong> the other firms chooses p satisfying:<br />
(p − c)D 1 [p, g(p M )+(n − 2)g(p)] + D [p, g(p M )+(n − 2)g(p)] = 0 (2.56)<br />
If endogenous entry holds, the number <strong>of</strong> firms satisfies also:<br />
(p − c)D [p, g(p M )+(n − 2)g(p)] = F (2.57)<br />
so that the pr<strong>of</strong>it <strong>of</strong> the monopolist bundling the two goods becomes π MB =<br />
(P M − c) D [p M , (n − 1)g(p)]. Noticethatifwedefine β = g(p M )+(n−2)g(p)<br />
the equilibrium spillovers received by the entrants as a consequence <strong>of</strong> the<br />
price chosen by their competitors, the equilibrium conditions (2.56)-(2.57)<br />
jointly determine p <strong>and</strong> β independently from the price <strong>of</strong> the monopolist.<br />
Using β M = β + g(p) − g(p M ) we can rewrite the equilibrium first order<br />
condition <strong>of</strong> the monopolist as an implicit expression for p M = p M (v 0 ),<strong>and</strong><br />
immediately derive that the equilibrium price <strong>of</strong> the secondary good decided<br />
by the monopolist has to be decreasing in v 0 . 47<br />
47 In particular we have:<br />
dp M<br />
= −D 1 [p M ,β+ g(p) − g(p M )]<br />
< 0<br />
dv 0 ∆
2.10 Bundling 81<br />
Clearly, bundling is optimal if π MB > π M . We need to verify under<br />
which conditions this happens. Before doing that, let us look at the way in<br />
which bundling changes the strategy <strong>of</strong> the monopolist. Since ∂π MB /∂p M −<br />
∂π M /∂p M = v 0 D 1 < 0, bundling makes the monopolist tough. This implies<br />
that the monopolist is led to reduce the effective price in the secondary market<br />
by choosing a low price <strong>of</strong> the bundle. Since SC holds, a price decrease<br />
by the monopolist induces the other firms to reduce their prices. Under exogenous<br />
entry, as in the Whinston (1990) model with two firms, this reduces<br />
pr<strong>of</strong>its <strong>of</strong> all firms in the secondary market, hence bundling is never optimal<br />
unless it manages to deter entry. Under endogenous entry, however, this result<br />
can change: bundling can now be an effective device to outplace some<br />
<strong>of</strong> the other firms without fully deterring entry in the secondary market, but<br />
creating some pr<strong>of</strong>its for the monopolist in this market through an aggressive<br />
strategy. In particular, bundling is optimal if the low price <strong>of</strong> the bundle<br />
increases pr<strong>of</strong>its in the competitive market more than it reduces them in the<br />
monopolistic one. It is easy to verify that bundling is optimal if:<br />
[p M (v 0 ) − c] D [p M (v 0 ),β M ] − F 0 >v− v 0 D [p M (v 0 ),β M ]<br />
whose left h<strong>and</strong> side is the gain in pr<strong>of</strong>its in the competitive market <strong>and</strong><br />
whose right h<strong>and</strong> side is the loss in pr<strong>of</strong>its in the monopolistic market:<br />
Proposition 2.8. Under price competition with endogenous entry<br />
in a secondary market, a monopolist in a primary market can<br />
have an incentive to bundle both goods to be aggressive.<br />
It is important to remark that, in this case, bundling does not need to<br />
have an exclusionary purpose as assumed by the leverage theory <strong>of</strong> tied good<br />
sales. The reduction in the price <strong>of</strong> the two bundled goods together can also<br />
benefit consumers. This is even more likely when they are complements, when<br />
there are network externalities between products, or when bundling creates<br />
efficiency effects.<br />
Bundling is an example <strong>of</strong> a discrete strategy: a firm either bundles two<br />
goods or not. A similar story can be used to evaluate a related discrete<br />
strategy, the choice <strong>of</strong> product compatibility <strong>and</strong> system compatibility, orinteroperability:<br />
as Tirole (1988, p. 335) has correctly noticed, “a manufacturer<br />
that makes its system incompatible with other systems imposes a de facto<br />
tie-in.” Typically, product compatibility s<strong>of</strong>tens price competition because<br />
consumers can mix <strong>and</strong> match products <strong>of</strong> different firms: these products endogenously<br />
become complements, while they would be substitutes in case <strong>of</strong><br />
incompatibility. Since price cuts are more pr<strong>of</strong>itable when competing products<br />
are substitutes rather than complements, interoperability s<strong>of</strong>tens price<br />
competition.<br />
where ∆ ≡ 2D 1 +(p M +v 0 −c)[D 11 −g 0 (p M)D 12]−g 0 (p M)D 2 < 0 by the stability<br />
<strong>of</strong> the equilibrium system. In other words, the price <strong>of</strong> the bundle increases less<br />
than proportionally with v 0 or the monopolist <strong>of</strong>fers the bundle with a discount<br />
on the secondary good compared to its competitors.
82 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
Therefore, according to the st<strong>and</strong>ard outcome under price competition<br />
with an exogenous number <strong>of</strong> competitors, the only reason why a leader would<br />
choose a low level <strong>of</strong> interoperability would be to induce their exit from the<br />
market. On the contrary, our results suggest that, when entry in the market<br />
is endogenous, a leader may favour a limited level <strong>of</strong> interoperability for<br />
adifferent purpose than entry deterrence: just because this strategy would<br />
strengthen price competition <strong>and</strong> enhance the gains from a low pricing strategy<br />
in the system competition, that is the competition between alternative<br />
systems.<br />
2.11 Vertical Restraints<br />
Vertical restraints are agreements or contracts between vertically related<br />
firms. They include franchise fees, that specify a non-linear payment <strong>of</strong> the<br />
downstream firm for the inputs provided by the upstream firm with a fixed<br />
fee <strong>and</strong> a variable part (so that the average price is decreasing in the number<br />
<strong>of</strong> units bought), quantity discounts <strong>and</strong> various forms <strong>of</strong> rebates, that<br />
<strong>of</strong>ten play a similar role to the one <strong>of</strong> the francise fees, exclusivity clauses<br />
<strong>and</strong> other minor restraints. When these restraints improve the coordination<br />
<strong>of</strong> a vertical chain, they are typically welfare improving, however, when they<br />
affect interbr<strong>and</strong> competition, that is competition between different products<br />
<strong>and</strong> different vertical chains, they can induce adverse consequences on consumers:<br />
namely they can be used to keep prices high <strong>and</strong>, therefore, they<br />
should be punished by the antitrust authorities. This is the st<strong>and</strong>ard result<br />
<strong>of</strong> the theory <strong>of</strong> strategic vertical restraints in interbr<strong>and</strong> competition (Bonanno<br />
<strong>and</strong> Vickers, 1988; Rey <strong>and</strong> Stiglitz, 1988), which suggests that, as long<br />
as firms compete in prices, a firm has incentives to choose vertical separation<br />
<strong>and</strong> charge his retailer a francise fee together with a wholesale price above<br />
marginal cost to induce an accommodating behavior.<br />
Consider an upstream firm that produces a good at marginal cost c <strong>and</strong><br />
fixed cost F , <strong>and</strong> delegates its distribution on the market to a downstream<br />
firm through a contract implying a fixed fee Υ <strong>and</strong> a wholesale price w for the<br />
good. The downstream firm sells this same good at the price p D to maximize<br />
net pr<strong>of</strong>its:<br />
π D =(p D − w)D(p D ,β D ) − Υ (2.58)<br />
while the other firms, that are vertically integrated <strong>and</strong> face the same cost<br />
structure, have the st<strong>and</strong>ard pr<strong>of</strong>it function π i =(p i − c)D(p i ,β i ) − F .The<br />
upstream firm can preliminarily choose the optimal contract, meaning the<br />
wholesale price w <strong>and</strong> the fee Υ that maximize net pr<strong>of</strong>its:<br />
π U =(w − c)D(p D ,β D )+Υ − F (2.59)
2.11 Vertical Restraints 83<br />
It is always optimal to choose w such that the pr<strong>of</strong>its <strong>of</strong> the downstream firm<br />
are maximized, <strong>and</strong> the fee that fully expropriates these pr<strong>of</strong>its. Of course,<br />
achoicew = c would be neutral for the market outcome. However, Bonanno<br />
<strong>and</strong> Vickers (1988) have shown that, if competition is between an exogenous<br />
number <strong>of</strong> firms, it is optimal to choose a high wholesale price w>cto s<strong>of</strong>ten<br />
price competition, <strong>and</strong> increase prices compared to the outcome in which the<br />
firm is vertically integrated. This is the classic example <strong>of</strong> an anti-competitive<br />
vertical restraint adopted by a market leader through strategic delegation <strong>of</strong><br />
accommodating pricing. 48<br />
When entry in the market is endogenous, the market leader cannot operate<br />
as above, because high wholesale prices would put the downstream firm<br />
out <strong>of</strong> the market. A market leader can still gain from delegating pricing decisions,<br />
but the optimal contract is now radically different. In particular, we<br />
know from our general results, that competition in prices with endogenous<br />
entry between the downstream firm <strong>and</strong> the other firms would lead to a price<br />
p D (w) increasing in the wholesale price for the downstream firm, a price for<br />
the other firms p = p D (c) <strong>and</strong> an endogenous value for β; moreover,bothp<br />
<strong>and</strong> β would be independent from w, <strong>and</strong>β D (w) =β + g(p) − g(p D (w)). One<br />
can verify that the optimal contract solves the problem:<br />
max π U =(w − c)D [p D (w),β D (w)] + Υ − F<br />
{w,Υ}<br />
s.v. : π D =[p D (w) − w] D [p D (w),β D (w)] − Υ ≥ 0<br />
<strong>and</strong> requires a wholesale price for the retailer smaller than the marginal cost<br />
<strong>and</strong> implicitly given by: 49<br />
w ∗ = c + (p D − c)D 2 g 0 (p D )<br />
84 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
In such a case, the vertical restraint leads to a lower price for the consumers<br />
<strong>and</strong> there is no ground for conjecturing any anti-competitive behavior.<br />
50 Therefore, also in the case <strong>of</strong> vertical restraints affecting interbr<strong>and</strong><br />
competition, entry conditions are crucial to derive proper conclusions.<br />
2.12 Price Discrimination<br />
When firms sell the same good at different prices for different consumers, they<br />
are adopting a policy <strong>of</strong> price discrimination, which is <strong>of</strong>ten regarded as an<br />
anti-competitive practice by antitrust authorities dealing with exclusionary<br />
or exploitative abuses by dominant firms: for this reason, in this sction we will<br />
try to underst<strong>and</strong> the motivations for the adoption <strong>of</strong> price discrimination. 51<br />
Typically, this increases pr<strong>of</strong>itability, but it also allows to differentiate prices<br />
betweenconsumerswithadifferent willingness to pay. Moreover, notice that<br />
price discrimination requires a certain commitment, because similar goods<br />
mustbesoldnotonlyatdifferent prices for different consumers, but also<br />
in different packages <strong>and</strong> with different advertising. In theory, when firms<br />
enjoy perfect information on the preferences <strong>of</strong> the consumers <strong>and</strong> can set a<br />
price equal to the maximum willingness to pay <strong>of</strong> each consumer, they can<br />
fully extract the consumer surplus, something known as first degree price<br />
discrimination. However, the usual forms <strong>of</strong> discrimination are much more<br />
limited.<br />
A large literature has focused on the more realistic case <strong>of</strong> incomplete<br />
information, in which firms <strong>of</strong>fer different deals <strong>and</strong> customers choose their<br />
favorite:atypicalexample<strong>of</strong>thissecond degree price discrimination involves<br />
price-quantity bundles, <strong>of</strong>ten involving quantity discounts. A market <strong>of</strong>ten<br />
analyzed in the literature is the insurance market, where high risk types dem<strong>and</strong><br />
more insurance <strong>and</strong> low risk types dem<strong>and</strong> less insurance (but firms do<br />
not know who is who). In such a market, simple price competition with free<br />
entry leads to a market failure because high risk types drive out low risk types<br />
<strong>and</strong> the market collapses (Akerl<strong>of</strong>, 1970). In such a case, different prices for<br />
different quantities naturally emerge in a competitive framework. Rothschild<br />
<strong>and</strong> Stiglitz (1976) have shown that a free entry equilibrium <strong>of</strong> this kind is<br />
characterized by low risk types accepting limited insurance associated with a<br />
low price <strong>and</strong> high risk types accepting full insurance at a higher price. Such<br />
a separating equilibrium works because high risk types prefer their contract<br />
rather than imitating the low risk types <strong>and</strong> obtain cheaper but limited in-<br />
50 A similar result emerges also in models <strong>of</strong> competition in quantities, but this<br />
is less surprising since it confirms the outcome <strong>of</strong> delegation games with an<br />
exogenous number <strong>of</strong> competitors.<br />
51 See Tirole (1988, Ch. 3) for an introduction to price discrimination.
2.12 Price Discrimination 85<br />
surance. 52 On the contrary, a pooling equilibrium cannot exist because a firm<br />
may deviate by <strong>of</strong>fering a contract which is pr<strong>of</strong>itable if accepted just by the<br />
low risk types.<br />
The relevance <strong>of</strong> this theory <strong>of</strong> competitive price discrimination due to<br />
asymmetric information has been challenged on empirical ground because we<br />
rarely observe a positive correlation between risk <strong>and</strong> insurance, even in the<br />
automobile insurance market (Chiappori <strong>and</strong> Salanie’, 2000). However, this<br />
should not surprise because this (as many other insurance markets) is not<br />
a one shot market, but is characterized by short term contracts which are<br />
periodically updated. In a dynamic version <strong>of</strong> the Rothschild-Stiglitz model,<br />
pooling equilibria with experience rating naturally emerge (Etro, 2000), <strong>and</strong><br />
they exactly mimic the bonus-malus policy that characterizes this market<br />
everywhere: initial contracts are st<strong>and</strong>ard for anybody, but future contracts<br />
are updated in a Bayesian fashion according to the performance <strong>of</strong> the drivers<br />
(which is public knowledge). 53 Notice that also this form <strong>of</strong> dynamic price<br />
differentiation based on observable features (the accident record) is a form <strong>of</strong><br />
price discrimination, still emerging in an equilibrium with endogenous entry.<br />
When firms discriminate on the basis <strong>of</strong> observable characteristics, we talk<br />
about third degree price discrimination. Wecanprovideasimpleexample<br />
<strong>of</strong> the role <strong>of</strong> this form <strong>of</strong> price discrimination within our framework. For<br />
simplicity, imagine that all firms compete simultaneously for a common set<br />
<strong>of</strong> consumers, whose dem<strong>and</strong> is D A (p i ,β i ) for each firm i, <strong>and</strong> the leader<br />
also serves a local market with dem<strong>and</strong> D B (p i ) (weassumethathastoserve<br />
52 While Rothschild <strong>and</strong> Stiglitz (1976) limited their analysis to two types <strong>of</strong> consumers,<br />
the model can be extended to multiple types: in such a case the equilibrium<br />
price function is nonlinear <strong>and</strong> can involve quantity discounts (Etro,<br />
1999, 2000). Hence, the empirical literature started with Cawley <strong>and</strong> Philipson<br />
(1999), who tested (<strong>and</strong> rejected) the convexity <strong>of</strong> the price function in insurance<br />
marketsishighlymisleading: contrary to their erroneous claim, bulk discounts<br />
can perfectly characterize the equilibrium price <strong>of</strong> competitive insurance markets<br />
with asymmetric information (exacly as they can characterize the optimal<br />
monopolistic price discrimination; see Maskin <strong>and</strong> Riley, 1984).<br />
53 To gain insights on the nature <strong>of</strong> the pooling equilibrium in a two period<br />
Rothschild-Stiglitz model (Etro, 2000), let us suppose that a pooling contract<br />
is <strong>of</strong>fered in the first period by all firms. The usual problem is that a new firm<br />
may deviate by <strong>of</strong>fering a contract which is pr<strong>of</strong>itable if accepted just by the low<br />
risk types. In the one period model this kind <strong>of</strong> contract always exists. In the<br />
two period model, however, the high risk types have a new incentive to accept<br />
a similar deviation (<strong>and</strong> make it unpr<strong>of</strong>itable), since by doing that, they would<br />
gain a reputation as low risk types <strong>and</strong> the associated second period contract<br />
with cheap full insurance. If agents are patient enough, any deviation is accepted<br />
by both types <strong>and</strong> so it is not pr<strong>of</strong>itable: hence the pooling contract is an equilibrium.<br />
Of course, also separating equilibria with immediate revelation <strong>of</strong> the<br />
risk types exist for low discounting.
86 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
both markets simultaneously). The leader can commit to a policy <strong>of</strong> price<br />
discrimination, <strong>and</strong> then choose two separate prices p A L <strong>and</strong> pB L for the same<br />
good sold at different kinds <strong>of</strong> customers. The marginal cost <strong>of</strong> production is<br />
c for all firms. The pr<strong>of</strong>its <strong>of</strong> the leader are then:<br />
π L = p A LD A (p A L,β L )+p B L D B (p B L ) − c[D B (p B L )+D A (p A L,β L )] − F (2.61)<br />
while the pr<strong>of</strong>its <strong>of</strong> the other firms are simply:<br />
π i = ¡ p A i<br />
− c ¢ D A (p A i ,β i ) − F<br />
Otherwise the leader can adopt a uniform pricing policy <strong>and</strong> choose a unique<br />
price p L for both kinds <strong>of</strong> customers, with the same pr<strong>of</strong>it functionasabove.<br />
The idea behind the commitment to discriminate is that price discrimination<br />
requires a small preliminary investment in package diversification <strong>and</strong><br />
separate advertising for the products sold for the different kind <strong>of</strong> customers.<br />
Consider the case <strong>of</strong> an exogenous number <strong>of</strong> firms. Choosing price discrimination,<br />
the leader sets the two prices, say p A L >pB L , <strong>and</strong> obtains monopolistic<br />
pr<strong>of</strong>its in the local market <strong>and</strong> (given symmetry) the same pr<strong>of</strong>its<br />
as the other firms in the symmetric Bertr<strong>and</strong> equilibrium for the common<br />
market. Choosing uniform pricing, the leader chooses an intermediate price<br />
p L ∈ (p B L ,pA L ) in Bertr<strong>and</strong> equilibrium, <strong>and</strong> SC implies that also the other<br />
firms will reduce their equilibrium prices. Ultimately, the leader reduces its<br />
pr<strong>of</strong>its in the local market <strong>and</strong> strengthens competition in the common market.<br />
Clearly, in this case, price discrimination is the optimal choice, since<br />
it allows the leader to maximize pr<strong>of</strong>its in the local market <strong>and</strong> to s<strong>of</strong>ten<br />
competition in the common one.<br />
Consider endogenous entry now. Under price discrimination, all firms<br />
choosethesamepricep A L in the common market <strong>and</strong> entry drives pr<strong>of</strong>its<br />
to zero in this market, while the leader enjoys only its monopolistic pr<strong>of</strong>its<br />
in the local market setting the optimal price p B L . Assume again that the dem<strong>and</strong><br />
conditions are such that p A L >pB L . In this case, by adopting uniform<br />
pricing, the leader will choose an intermediate price between p B L <strong>and</strong> pA L ,<strong>and</strong><br />
will obtain two results: on one side, pr<strong>of</strong>its in the local market will decrease<br />
because pricing is above monopolistic pricing, on the other side, pr<strong>of</strong>its in the<br />
common market will increase because the leader is endogenously committed<br />
to aggressive pricing, which is always optimal in a market where entry is<br />
endogenous. If the former loss is smaller than the latter gain, it is optimal to<br />
adopt uniform pricing rather than committing to price discrimination. 54<br />
This simple example is just aimed a suggesting that price discrimination<br />
can have a role in s<strong>of</strong>tening price competition (compared to uniform pricing)<br />
inducing negative consequences for consumers: this effect, however, is less<br />
54 Notice that this can happen because the loss from a small deviation from monopolistic<br />
pricing is a second order loss, while the gain in the common market is<br />
a first order gain.
2.13 <strong>Antitrust</strong> <strong>and</strong> Horizontal Mergers 87<br />
likely to emerge in markets where entry is endogenous, since in these markets<br />
an aggressive uniform pricing strategy can be optimal. In conclusion, we may<br />
have a possible new case for the association <strong>of</strong> price discrimination by market<br />
leaders with anti-competitive purposes. 55<br />
2.13 <strong>Antitrust</strong> <strong>and</strong> Horizontal Mergers<br />
We have seen that even when they face endogenous entry <strong>of</strong> competitors,<br />
market leaders can obtain positive pr<strong>of</strong>its by adopting certain strategic commitments.<br />
One may think that a preliminary merger with other firms <strong>and</strong><br />
a subsequent cooperation in the strategic decisions may serve a similar role.<br />
When the number <strong>of</strong> firms in the market is given, this is typically the case.<br />
Moreover, a merger induces a more accommodating behavior which exerts<br />
an indirect effect on the other firms. When SS holds the other firms become<br />
more aggressive, when SC holds they become more accommodating as well: 56<br />
for this reason, loosely speaking, mergers tend to be more pr<strong>of</strong>itable under<br />
competition in prices. However, once again, the situation changes when entry<br />
is endogenous. In such a case the merger can affect entry, which creates a<br />
new effect, <strong>of</strong>ten taken into account in antitrust policy considerations, but<br />
not in the theory <strong>of</strong> mergers until now. 57 In our context, a merger induces<br />
accommodation by the merged firm, which attracts entry <strong>and</strong> reduces the<br />
pr<strong>of</strong>its <strong>of</strong> the merged firm: consequently, there is no any strategic rationale<br />
for mergers when entry in the market is endogenous. 58<br />
Consider a merger between two firms, say firms k <strong>and</strong> j. Thenetpr<strong>of</strong>its<br />
<strong>of</strong> the merged firms become:<br />
π Merger = Π (x k ,β k )+Π ¡ x j ,β j<br />
¢<br />
− ˜F<br />
55 Notice that a different situation emerges if the dem<strong>and</strong> conditions are such that<br />
under price discrimination we have p A L
88 2. Strategic Commitments <strong>and</strong> Endogenous Entry<br />
where ˜F is the new fixed cost <strong>of</strong> production. Using the fact that β j = β k +<br />
h(x k ) − h(x j ) for k, j =1, 2, wehavethefirst order conditions:<br />
Π 1 (x k ,β k )+Π 2 (x j ,β j )h 0 (x k )=0 k, j =1, 2 (2.62)<br />
which clearly shows an accommodating behavior for both strategies. As we<br />
know, such a behavior creates a strategic disadvantage when entry in the<br />
market is endogenous. The equilibrium after the merger is characterized by<br />
two identical strategies for the merged firm, x k = x j = x M ,astrategyfor<br />
the followers x, <strong>and</strong> respective spillovers β M <strong>and</strong> β such that:<br />
Π 1 (x M ,β M )+Π 2 (x M ,β M )h 0 (x M )=Π 1 (x, β) =0, Π(x, β) =F<br />
This implies x M β: the equilibrium strategy <strong>of</strong> the other firms<br />
isalwaysthesameafterthemerger,buttheaccommodatingbehavior<strong>of</strong>the<br />
merged entity induces further entry so as to decrease its gross pr<strong>of</strong>its below<br />
those <strong>of</strong> each independent firm. Nevertheless, the merger can still be pr<strong>of</strong>itable<br />
if π Merger > 0, whichrequires ˜F < 2Π(x M ,β M ). In a market where entry<br />
is endogenous, the only way a merger can be pr<strong>of</strong>itable is by creating cost<br />
efficiencies. 59 This conclusion exactly matches the informal insights <strong>of</strong> the<br />
Chicago school on horizontal mergers (Bork, 1993, Posner, 2001), <strong>and</strong> can be<br />
summarized as follows:<br />
Proposition 2.10. In a market with endogenous entry, a horizontal<br />
merger induces accommodating behavior <strong>of</strong> the merged firm <strong>and</strong><br />
attracts entry <strong>of</strong> other firms: the merger is pr<strong>of</strong>itable if <strong>and</strong> only if<br />
it creates enough cost efficiencies to compensate for the strategic<br />
disadvantage <strong>of</strong> the merged firm.<br />
Notice that in models <strong>of</strong> competition in quantities <strong>and</strong> prices, as long<br />
as the merged firm does not deter entry, the equilibrium after the merger<br />
implies the same total production or the same price indexes as before (see<br />
also Chapter 3). Therefore, consumer surplus is not affected by the merger.<br />
Since the latter takes place only when there are significant cost efficiencies,<br />
it follows that horizontal mergers in markets where entry is endogenous are<br />
welfare improving. 60<br />
59 For instance, in the linear model <strong>of</strong> competition in quantities <strong>of</strong> Section 1.1,<br />
the merged firm would produce the same as the two separate firms, therefore<br />
the merger could be pr<strong>of</strong>itable only if ˜F < F. In the model with imperfect<br />
substitutability <strong>of</strong> Section 1.2.2, a merger between two firms would lead them to<br />
produce 2 −b times as before <strong>and</strong> to reach the joint pr<strong>of</strong>its π Merger =(2− b)(2 +<br />
b − b 2 )F/2 − ˜F which are positive if product differentiationisstrongenough(b<br />
or ˜F small).<br />
60 The Erkal-Piccinin model extends the analysis to more complex dem<strong>and</strong> functions:<br />
under competition in prices with a dem<strong>and</strong> system derived from the<br />
quadratic utility function (2.11), a merger increases the prices <strong>of</strong> the merged
2.14 Conclusions 89<br />
2.14 Conclusions<br />
This chapter has examined Nash <strong>and</strong> Marshallian competition within a general<br />
framework, <strong>and</strong> it has studied the strategic incentives <strong>of</strong> market leaders<br />
to undertake preliminary investments that can affect competition. A main result<br />
<strong>of</strong> this investigation has been that the behavior <strong>of</strong> market leaders facing<br />
endogenous entry is always biased toward the implementation <strong>of</strong> aggressive<br />
strategies. As we noticed in the examples <strong>of</strong> Chapter 1, this result confirms<br />
what we found in models <strong>of</strong> Stackelberg competition with endogenous entry,<br />
that is in models where the leader does not undertake full fledged investments<br />
to constraint subsequent decisions, but simply commits to strategies before<br />
the other firms. Since the ultimate results are analogous, we can safely look at<br />
models <strong>of</strong> Stackelberg competition with endogenous entry as reduced forms<br />
<strong>of</strong> the more general models <strong>of</strong> Marshallian competition with strategic investments<br />
analyzed in this chapter. The advantages <strong>of</strong> the first kind <strong>of</strong> models<br />
are that they are simpler, they allow to derive clear welfare comparisons with<br />
the corresponding models <strong>of</strong> Marshallian competition, <strong>and</strong> they allow further<br />
extensions. For this reason, in the next chapter we will move on to the study<br />
<strong>of</strong> general models <strong>of</strong> Stackelberg competition in the market with <strong>and</strong> without<br />
endogenous entry. In Chapter 4 we will do the same for general models <strong>of</strong><br />
Stackelberg competition for the market with <strong>and</strong> without endogenous entry.<br />
firms <strong>and</strong> reduces the prices <strong>of</strong> the other firms while increasing entry (nevertheless,<br />
in the absence <strong>of</strong> cost efficiencies, the impact on consumer surplus is<br />
typically negative).
3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous<br />
Entry<br />
In the 1930s, Stackelberg (1934) pioneered the study <strong>of</strong> a market structure<br />
where a firm has a leadership over the rivals. <strong>Market</strong> leaders obtain a stable<br />
advantage on the followers when they are first movers in the choice <strong>of</strong> the<br />
strategy. It is well known that the commitment <strong>of</strong> the leaders may not be<br />
credible when initial strategies can be easily revised over time. However, a<br />
commitment represents a credible advantage in markets with a short horizon<br />
or when strategies are costly to change. For instance, in some markets a<br />
certain production level is associated with preliminary investments in the<br />
preparation <strong>of</strong> projects, machinery, <strong>and</strong> on the allocation <strong>of</strong> different inputs.<br />
It may be costly to change these factors <strong>of</strong> production afterwards: being a<br />
first mover on the output choice in these markets can represent a fundamental<br />
strategic advantage. In other markets, prices are sticky in the short run due<br />
to small menu costs or to costly acquisition <strong>of</strong> information, or because a<br />
price change can induce adverse reputational effects on the perception <strong>of</strong> the<br />
customers: in these cases, being the first mover in the price choice provides<br />
the leader with a credible commitment in the short run. Finally, in sectors<br />
where firms compete for the market, a preliminary investment in research<br />
<strong>and</strong> development represents a solid commitment to an innovation strategy.<br />
In general, the economic concept <strong>of</strong> leadership associated with the timing <strong>of</strong><br />
the decisions can be seen as a simple representation <strong>of</strong> situations in which<br />
preliminary investments, as those studied in the previous chapter, provide a<br />
strategic advantage to a firm.<br />
The modern game theoretic analysis <strong>of</strong> competition between market leaders<br />
<strong>and</strong> followers started from the seminal contribution <strong>of</strong> Dixit (1980), 1 who<br />
focused, as with most <strong>of</strong> the subsequent literature, on a duopoly. When a market<br />
leader faces a single follower, two basic situations can emerge. If the fixed<br />
costs <strong>of</strong> production for the follower are high, the leader finds it optimal to<br />
deter entry, for instance by choosing a high output level that leaves too little<br />
dem<strong>and</strong> for the follower, or by choosing a low price against which the follower<br />
cannot pr<strong>of</strong>itably compete. If the fixed cost <strong>of</strong> production is low enough, for<br />
instance if it is zero, the leader cannot pr<strong>of</strong>itably deter entry, <strong>and</strong> has to<br />
compete on the market with the entrant. There are two possible outcomes,<br />
<strong>and</strong> they depend on the form <strong>of</strong> competition, or more precisely, on the kind<br />
1 See also Spence (1977) <strong>and</strong> Dixit (1979).
92 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
<strong>of</strong> strategic interaction. Under strategic substitutability, that typically holds<br />
with competition in quantities, the leader is aggressive: for instance, produces<br />
a lot so as to gain market share compared to the rival. Under strategic complementarity,<br />
that typically holds with competition in prices, the leader is<br />
accommodating: for instance, chooses a high price so as to induce the rival to<br />
choose a high price as well. Ultimately, whether strategic complementarity or<br />
substitutability holds is an empirical question, but its answer is not obvious,<br />
as it is <strong>of</strong>ten not obvious what the strategic variables are that are under the<br />
control <strong>of</strong> firms in the real world. For this reason the results <strong>of</strong> the theory<br />
appear too vague to be <strong>of</strong> practical interest for unambiguous descriptions <strong>of</strong><br />
the behavior <strong>of</strong> market leaders <strong>and</strong> for policy recommendations.<br />
The above analysis <strong>of</strong> the competition between a leader <strong>and</strong> a follower, as<br />
already said, holds when the fixed cost <strong>of</strong> production for the latter is low, so<br />
that entry deterrence is not an option for the leader. However, in this exact<br />
situation, net pr<strong>of</strong>its for the followers are likely to be high <strong>and</strong> they could<br />
attract other firms in the market. For this reason, an analysis limited to an<br />
exogenous number <strong>of</strong> firms (the leader <strong>and</strong> a single follower, or two followers,<br />
or any other given number <strong>of</strong> followers) can be quite misleading. In most<br />
markets, we can regard the number <strong>of</strong> competitors as an endogenous variable,<br />
which depends on the interaction between the market leader <strong>and</strong> the other<br />
firms, <strong>and</strong> not as an exogenous variable. In this chapter, we will examine the<br />
case in which a leader faces an endogenous number <strong>of</strong> followers. The results,<br />
based on Etro (2002, 2008), are quite simpler: the leader always behaves in an<br />
aggressive way, choosing higher production or lower prices than the followers.<br />
In particular, if each firm has a pr<strong>of</strong>it function Π(x i ,X −i ), where the aggregate<br />
statistics X −i = P j6=i x j summarize the strategies <strong>of</strong> the other firms,<br />
an interior equilibrium can be characterized quite simply. The choice <strong>of</strong> each<br />
entrant satisfies the normal optimality condition Π 1 =0, while the choice <strong>of</strong><br />
the leader satisfies Π 1 = Π 2 . For instance, under competition in quantities<br />
<strong>and</strong> homogenous goods, this implies that the entrants equate marginal cost<br />
<strong>and</strong> marginal revenue, while the leader equates marginal cost <strong>and</strong> price. Its<br />
pr<strong>of</strong>its are positive because production is in the region <strong>of</strong> increasing average<br />
costs. We will also verify under which conditions the leader finds it optimal<br />
to be so aggressive as to deter entry, <strong>and</strong> we will see that the conditions for<br />
such an outcome are not very dem<strong>and</strong>ing: under competition in quantities<br />
<strong>and</strong> homogenous goods the equilibrium implies just one firm,theleader,as<br />
long as there are increasing, constant or even slightly decreasing returns to<br />
scale. 2<br />
The analysis <strong>of</strong> Stackelberg competition with endogenous entry is somewhat<br />
related with three older theoretical frameworks. The first is the initial<br />
literature on entry deterrence associated with the so-called Bain-Modigliani-<br />
Sylos Labini framework. However, even if the initial contributions by Sylos<br />
2 As we have already seen in Chapter 1, with a linear dem<strong>and</strong> p = a − X <strong>and</strong> a<br />
constant marginal cost c, the equilibrium implies the limit price p = c +2 √ F .
3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry 93<br />
Labini (1956), Bain (1956) <strong>and</strong> Modigliani (1958) took in consideration the<br />
effects <strong>of</strong> entry on the behavior <strong>of</strong> market leaders, they were not developed<br />
in a coherent game theoretic framework <strong>and</strong> were substantially limited to<br />
the case <strong>of</strong> competition with perfectly substitute goods <strong>and</strong> constant or decreasing<br />
marginal costs (which not by chance, as we will see, are sufficient<br />
conditions for entry deterrence).<br />
The second is the dominant firm theory, which tried to explain the pricing<br />
decision <strong>of</strong> a market leader facing a competitive fringe <strong>of</strong> firms taking<br />
as given the price <strong>of</strong> the leader. 3 Assuming that the supply <strong>of</strong> this fringe is<br />
increasing in the price, the dem<strong>and</strong> <strong>of</strong> the leader is total market dem<strong>and</strong> net<br />
<strong>of</strong> this supply. The pr<strong>of</strong>it maximizing price <strong>of</strong> the leader is above marginal<br />
cost but constrained by the competitive fringe. While such a model is not<br />
fully consistent with rational behavior <strong>of</strong> the parts in a game theoretic perspective,<br />
it provides interesting insights on the behavior <strong>of</strong> market leaders<br />
under competitive pressure.<br />
The third is the theory <strong>of</strong> contestable markets by Baumol et al. (1982),<br />
which focuses mainly on homogenous goods <strong>and</strong> shows that, in the absence<br />
<strong>of</strong> sunk costs <strong>of</strong> entry, the possibility <strong>of</strong> “hit <strong>and</strong> run” strategies by potential<br />
entrants is compatible only with an equilibrium price equal to the average<br />
cost. One <strong>of</strong> the main implications <strong>of</strong> this result is that “one firm can be<br />
enough” for competition when there are aggressive potential entrants. 4<br />
None <strong>of</strong> these frameworks provides indications on the behavior <strong>of</strong> market<br />
leaders in general contexts, but nevertheless they have been quite helpful in<br />
providing insights on the role <strong>of</strong> competitive pressure in markets with leaders.<br />
3 See Carlton <strong>and</strong> Perl<strong>of</strong>f (2004) <strong>and</strong> Viscusi et al. (2005, Ch. 6) for an introduction<br />
<strong>and</strong> Kahai et al. (1996) for an empirical application to the case <strong>of</strong> AT&T. See<br />
also the work <strong>of</strong> Gaskins (1971) on dynamic limit pricing under threat <strong>of</strong> entry;<br />
I am grateful to Avinash Dixit for attracting my attention on this work.<br />
4 Baumol et al. (1982) note that the contestable outcome can be described as the<br />
game in which firms first choose prices simultaneously <strong>and</strong> then choose output (or<br />
capacity) if they enter (choosing positive output implies entry decision). They<br />
also claim that the theory <strong>of</strong> perfect contestable market can be viewed as a<br />
generalization <strong>of</strong> the Bertr<strong>and</strong> game to markets with increasing returns to scale.<br />
Inthecase<strong>of</strong>alineardem<strong>and</strong>p = a − X <strong>and</strong> a constant marginal cost c, the<br />
contestable-market equilibrium requires a price <strong>of</strong> the incumbent equal to the<br />
average cost (p = a − x = c + F/x), therefore:<br />
p = 1 2<br />
<br />
a + c − <br />
(a − c) 2 − 4F<br />
which is always lower than the equilibrium price under Stackelberg competition<br />
in quantities with endogenous entry. The contestable-market equilibrium can be<br />
also interpreted as a Stackelberg equilibrium in prices with endogenous entry <strong>and</strong><br />
homogenous goods. Of course, our theory applies beyond the case <strong>of</strong> homogenous<br />
goods.
94 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
In this chapter we will develop a general theory <strong>of</strong> Stackelberg competition<br />
with endogenous entry within the framework developed in the previous chapter,<br />
<strong>and</strong> we will analyze complex situations where there are multiple leaders,<br />
where the leadership itself is endogenous, where there are multiple strategies<br />
to be chosen, <strong>and</strong> where there are more general pr<strong>of</strong>it functions. Finally<br />
we will analyze a few applications concerning collusive cartels <strong>and</strong> antitrust<br />
policy, strategic export promotion <strong>and</strong> privatizations.<br />
The chapter is organized as follows. Section 3.1 studies pure Stackelberg<br />
competition where entry is exogenous, while Section 3.2 studies Stackelberg<br />
competition with endogenous entry. Section 3.3 applies these models to general<br />
forms <strong>of</strong> competition in quantities <strong>and</strong> in prices. Section 3.4 extends the<br />
model in different directions. Section 3.5 derives some implications for collusion<br />
between firms. Section 3.6-7 concludes our analysis with a digression on<br />
commitments created by government policy as state aids to exporting firms<br />
<strong>and</strong> privatizations. Section 3.8 concludes.<br />
3.1 Stackelberg Equilibrium<br />
In this section we will study a general version <strong>of</strong> a simple <strong>and</strong> well known<br />
game: Stackelberg competition. The number <strong>of</strong> firms in the market, n, is<br />
exogenous, for instance because legal or institutional constraints limit entry,<br />
or because a certain technology is available only for a limited number <strong>of</strong><br />
firms, or is protected by intellectual property rights. What is crucial for the<br />
following analysis is that no other firms can enter in the market even if this<br />
is pr<strong>of</strong>itable. One <strong>of</strong> the firms, the leader, can choose its own strategy before<br />
the other firms. These other firms, defined as followers, choose simultaneously<br />
their own strategies taking as given the strategy <strong>of</strong> the leader. Therefore, this<br />
is a Stackelberg game with one leader <strong>and</strong> n−1 followers, <strong>and</strong> we are looking<br />
for its subgame perfect equilibrium.<br />
Imagine that each firm i has the pr<strong>of</strong>it function:<br />
π i = Π(x i ,β i ) − F with β i =<br />
nX<br />
j=1,j6=i<br />
h(x j ) (3.1)<br />
where Π is unimodal in the first argument x i , which is the strategy <strong>of</strong> the<br />
same firm, <strong>and</strong> decreasing in the second argument β i , which summarizes the<br />
strategies <strong>of</strong> the other firms through a positive <strong>and</strong> increasing function h(·).<br />
In Chapter 1 we analyzed a few examples <strong>of</strong> this environment: models<br />
<strong>of</strong> Stackelberg competition in quantities with linear dem<strong>and</strong> <strong>and</strong> with homogenous<br />
goods or imperfect substitutability between goods, models with<br />
U-shaped average cost functions, models <strong>of</strong> competition in prices with a Logit<br />
dem<strong>and</strong>, <strong>and</strong> simple models <strong>of</strong> competition for the market. In those examples<br />
the leader in the market was exploiting the first mover advantage in different
3.1 Stackelberg Equilibrium 95<br />
ways. For instance, in models <strong>of</strong> competition in quantities <strong>and</strong> <strong>of</strong> competition<br />
for the market we found out that the leader was aggressive compared to the<br />
followers (producing or investing more), while in models <strong>of</strong> competition in<br />
prices the leader was accommodating (choosing higher prices <strong>and</strong> producing<br />
less). Here we generalize those findings in a rule for the behavior <strong>of</strong> the market<br />
leaders.<br />
We will focus on the case in which interior equilibria emerge, that is all<br />
firms are active in the market <strong>and</strong> obtain positive pr<strong>of</strong>its, <strong>and</strong> the leader does<br />
not find it optimal to deter entry. This case emerges whenever the fixed costs<br />
are low enough. 5 We can define the equilibrium in the following way:<br />
Definition 3.1. A Stackelberg Equilibrium between n firms is such that 1)<br />
each follower chooses its strategy x to maximize its pr<strong>of</strong>its given the spillovers<br />
β from the other firms <strong>and</strong> the strategy <strong>of</strong> the leader x L ;2)theleaderchooses<br />
its strategy x L to maximize its pr<strong>of</strong>its under rational expectations on β L ;3)<br />
β =(n − 2)h(x)+h(x L ) <strong>and</strong> β L =(n − 1)h(x).<br />
As usual, the equilibrium can be solved by backward induction. Given<br />
the strategy <strong>of</strong> the leader, defined as x L , all the followers choose their own<br />
strategies to satisfy the first order conditions:<br />
Π 1 (x i ,β i )=0 for any i (3.2)<br />
In this kind <strong>of</strong> game, for a given number <strong>of</strong> firms, a pure-strategy equilibrium<br />
exists if the reaction functions are continuous or do not have downward jumps<br />
(see Vives, 1999). Unfortunately this may not be the case due to the presence<br />
<strong>of</strong> fixed costs, but weak conditions for existence have been studied for many<br />
applications. 6 In this general framework we will just assume the existence<br />
<strong>of</strong> a unique symmetric equilibrium where all the followers choose the same<br />
strategy x. 7 In the symmetric equilibrium we have:<br />
Π 1 [x, (n − 2)h(x)+h(x L )] = 0 (3.3)<br />
This expression provides the strategy <strong>of</strong> the follower x as a function <strong>of</strong> the<br />
strategy <strong>of</strong> the leader <strong>and</strong> <strong>of</strong> the number <strong>of</strong> firms, x = x(x L ,n). Totally<br />
differentiating the equilibrium first order condition, it follows that:<br />
dx<br />
Π 12 (x, β) h 0 (x L )<br />
=<br />
dx L − [Π 11 (x, β)+(n − 2)h 0 (x)Π 12 (x, β)]<br />
5 The next section will deal with the case in which the fixed costs <strong>of</strong> production<br />
are high enough (or the number <strong>of</strong> potential entrants is high enough), that only<br />
a limited <strong>and</strong> endogenous number <strong>of</strong> firms actually enters in the market.<br />
6 For instance, see Amir <strong>and</strong> Lambson (2000) on Cournot games with perfectly<br />
substitute goods <strong>and</strong> Vives (1999) for a survey.<br />
7 This happens in all <strong>of</strong> our examples <strong>and</strong>, in general, under a st<strong>and</strong>ard contraction<br />
condition, Π 11 +(n − 2)h 0 (x) |Π 12 | < 0. Thisalwaysholdsforn =2.Withmore<br />
than one follower, weaker conditions for uniqueness are available for particular<br />
models.
96 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
whose denominator is positive under the assumption <strong>of</strong> stability. Hence, a<br />
more aggressive strategy <strong>of</strong> the leader (an increase in x L ) makes the followers<br />
more aggressive under the assumption <strong>of</strong> SC (Π 12 > 0), <strong>and</strong> less aggressive<br />
under SS (Π 12 < 0).<br />
In the first stage, the leader takes this into account <strong>and</strong> maximizes:<br />
π L = Π L (x L ,β L ) − F (3.4)<br />
where β L =(n − 1)h [x(x L ,n)]. Therefore, in the case <strong>of</strong> an interior solution,<br />
we obtain the first order condition:<br />
Π1 L (x L ,β L )+Π2 L (x L ,β L ) ∂β L<br />
=0 (3.5)<br />
∂x L<br />
<strong>and</strong> we assume that the second order condition is satisfied. Using our expression<br />
for dx/dx L we have:<br />
Π1 L (x L ,β L )= (n − 1)h0 (x)h 0 (x L )Π 12 (x, β) Π2 L (x L,β L )<br />
[Π 11 (x, β)+(n − 2)h 0 (3.6)<br />
(x)Π 12 (x, β)]<br />
whose term on the right h<strong>and</strong> side has the sign <strong>of</strong> Π 12 (x, β). Comparing the<br />
equilibrium condition for the followers <strong>and</strong> that for the leader, it is immediate<br />
to derive:<br />
Proposition 3.1. A Stackelberg equilibrium with exogenous entry<br />
implies that the leader is aggressive compared to the followers<br />
under strategic substitutability <strong>and</strong> accommodating under strategic<br />
complementarity.<br />
The intuition for this result is straightforward. 8 When the leader foresees<br />
that a more aggressive strategy will induce the followers to be more aggressive,<br />
it is optimal to be accommodating, which happens under SC. When the<br />
leader foresees that a more aggressive strategy will induce the followers to<br />
be more accommodating, it is then optimal to be aggressive, which happens<br />
under SS. From this general principle we can make sense <strong>of</strong> our results in<br />
Chapter 1: in the models <strong>of</strong> competition in quantities, the leader was aggressive<br />
because higher production was pushing toward a lower production<br />
for the followers, while in the model <strong>of</strong> competition in prices, the leader was<br />
accommodating because a higher price was pushing toward higher prices for<br />
the followers as well, increasing the pr<strong>of</strong>its <strong>of</strong> all firms. As we will see later on<br />
in detail, these are the typical outcomes <strong>of</strong> these two forms <strong>of</strong> competition,<br />
while competition for the market can lead to a different behavior <strong>of</strong> the leader<br />
depending on many market features, a point that we will revisit in the next<br />
chapter.<br />
8 Contrary to the model <strong>of</strong> Fudenberg <strong>and</strong> Tirole (1984), here we do not have<br />
a preliminary investment that affectsthestrategy,butwehaveapreliminary<br />
strategy tout court. In the terminology <strong>of</strong> the last chapter, it is as if we are<br />
alwaysinthecasewhereΠ L 13 > 0.
3.2 Stackelberg Equilibrium with Endogenous Entry 97<br />
3.2 Stackelberg Equilibrium with Endogenous Entry<br />
Letusmovenowtothecaseinwhichthenumber<strong>of</strong>firms in the market is not<br />
an exogenous variable, but it actually depends on the pr<strong>of</strong>itable opportunities<br />
in the market. As long as there are positive pr<strong>of</strong>its to be made in the market,<br />
firms enter <strong>and</strong> compete with the leader <strong>and</strong> the other firms. 9 Therefore,<br />
the number <strong>of</strong> competitors is endogenously determined by the technological<br />
conditions, by the nature <strong>of</strong> the strategic interaction, <strong>and</strong> by the preliminary<br />
strategy <strong>of</strong> the leader.<br />
More precisely, following Etro (2002, 2008), we will look at the subgame<br />
perfect equilibrium <strong>of</strong> the game with the following sequence <strong>of</strong> moves:<br />
1) in the first stage, a leader, firm L, enters,paysthefixed cost F <strong>and</strong><br />
chooses its own strategy x L ;<br />
2) in the second stage, after knowing the strategy <strong>of</strong> the leader, all potential<br />
entrants simultaneously decide “in” or“out”: if a firm decides “in”,<br />
it pays the fixed cost F ;<br />
3) in the third stage all the followers that have entered choose their own<br />
strategy x i (hence, the followers play simultaneously).<br />
We can define the new equilibrium in the following way:<br />
Definition 3.2. A Stackelberg equilibrium with endogenous entry is such<br />
that 1) each follower chooses its strategy x to maximize its pr<strong>of</strong>its given the<br />
spillovers β from the other firms; 2) the number <strong>of</strong> firms n is such that<br />
all followers make non negative pr<strong>of</strong>its <strong>and</strong> entry <strong>of</strong> another follower would<br />
induce negative pr<strong>of</strong>its for all <strong>of</strong> them; 3) the leader chooses its strategy x L<br />
to maximize its pr<strong>of</strong>its under rational expectations on x <strong>and</strong> n; 4)β =(n −<br />
2)h(x)+h(x L ) <strong>and</strong> β L =(n − 1)h(x).<br />
To characterize this equilibrium we look at the last stage again. In this<br />
stage, in the case <strong>of</strong> an interior equilibrium, we still have a st<strong>and</strong>ard first<br />
order condition for the followers:<br />
Π 1 [x, (n − 2)h(x)+h(x L )] = 0 (3.7)<br />
Since dΠ/∂dn = Π 2 h(x) < 0 under our assumptions, entry reduces gross<br />
pr<strong>of</strong>its until they reach the fixed costs <strong>and</strong> further entry is not pr<strong>of</strong>itable<br />
anymore. Therefore, ignoring the integer constraint on the number <strong>of</strong> firms,<br />
we can impose the endogenous entry condition as a zero pr<strong>of</strong>it condition:<br />
9 The exogeneity <strong>of</strong> the leadership, that is <strong>of</strong> the identity <strong>of</strong> the leader <strong>and</strong> also<br />
<strong>of</strong> the number <strong>of</strong> leaders, can be a realistic description for markets with an<br />
established dominant firm, or where entry at an earlier stage was not possible for<br />
technological or legal reasons, for liberalized markets that were once considered<br />
natural monopolies or those where intellectual property rights play an important<br />
role. Later, we will extend the model to multiple leaders <strong>and</strong> to endogenous<br />
leadership.
98 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
Π [x, (n − 2)h(x)+h(x L )] = F (3.8)<br />
Leaving a formal treatment to the Appendix, we will provide here an intuitive<br />
<strong>and</strong> constructive argument to characterize the Stackelberg equilibrium<br />
with endogenous entry which will be useful in the applications <strong>of</strong> the next<br />
section. The system (3.7)-(3.8) can be thought <strong>of</strong> as determining the behavior<br />
<strong>of</strong> the followers in the second <strong>and</strong> third stages, namely as determining x <strong>and</strong><br />
n as functions <strong>of</strong> the leader’s first stage action. But we can also look at these<br />
two equations in a different way: they can be solved for the two unknowns<br />
x <strong>and</strong> β. The pair (x, β) will only depend on the fixed cost <strong>of</strong> production<br />
<strong>and</strong> not on the strategy <strong>of</strong> the leader. Given (x, β), there is a unique locus <strong>of</strong><br />
(x L ,n) pairs that satisfy the equilibrium relation β =(n − 2)h(x)+h(x L ).In<br />
other words, the strategy <strong>of</strong> the followers is independent from the strategy <strong>of</strong><br />
the leader, while their number must change with the latter. The invariance<br />
property (dx/dx L =0) is quite important since it shows that what matters<br />
for the leader is not the reaction <strong>of</strong> each single follower to its strategy, but<br />
the effect on entry. This is exactly the opposite <strong>of</strong> what happened in the<br />
Stackelberg equilibrium. When entry is exogenous the leader takes as given<br />
the number <strong>of</strong> followers <strong>and</strong> looks at the reaction <strong>of</strong> their strategies to its own<br />
strategy. When entry is endogenous the leader takes as given the strategies<br />
<strong>of</strong> the followers <strong>and</strong> looks at the reaction <strong>of</strong> their number to its own strategy.<br />
Let us now move to the first stage <strong>and</strong> study the choice <strong>of</strong> the leader. As<br />
long as entry takes place, the perceived spillovers <strong>of</strong> the leader can be written<br />
as<br />
β L =(n − 1)h(x) =(n − 2)h(x)+h(x)+h(x L ) − h(x L )= (3.9)<br />
= β + h(x) − h(x L )<br />
which depends on x L only through the last term, since we have just seen that<br />
the pair (x, β) does not depend on x L . We can use this result to verify when<br />
entry <strong>of</strong> followers takes place or does not. It is immediate that entry does not<br />
occur for any strategy <strong>of</strong> the leader x L above a cut-<strong>of</strong>f ¯x L such that n =2<br />
or, substituting in (3.9), such that:<br />
β = h(¯x L ) (3.10)<br />
which clearly implies ¯x L ≥ x. Entry occurs whenever x L < ¯x L .Insucha<br />
case, the leader chooses the optimal strategy to maximize:<br />
π L = Π L [x L ,β+ h(x) − h(x L )] − F (3.11)<br />
which delivers the first order condition: 10<br />
10 Notice that the second order condition is:<br />
D L = Π L 11 − 2Π L 12h 0 (x L ) − Π L 2 h 00 (x L )+Π L 22h 0 (x L ) 2 < 0<br />
that we assume to be satisfied at the interior optimum.
3.2 Stackelberg Equilibrium with Endogenous Entry 99<br />
Π L 1 [x L , (n − 1)h(x)] = Π L 2 [x L , (n − 1)h(x)] h 0 (x L ) (3.12)<br />
In this case the equilibrium values for x L , x <strong>and</strong> n are given by the<br />
system <strong>of</strong> three equations (3.7)-(3.8) <strong>and</strong> (3.12). In general, the pr<strong>of</strong>itfunction<br />
perceived by the leader is an inverted U relation in x L for any strategy below<br />
the entry deterrence level ¯x L , <strong>and</strong> it takes positive values just for x L >x.<br />
Beyond the cut-<strong>of</strong>f ¯x L , it is downward sloping (as long as the market is not a<br />
natural monopoly). Hence, the strategy ¯x L is optimal only if it provides higher<br />
pr<strong>of</strong>its than at the local optimal strategy for x L < ¯x L (see the Appendix for<br />
the details). If we focus our attention on the qualitative behavior <strong>of</strong> the firms,<br />
we can conclude as follows:<br />
Proposition 3.2. A Stackelberg equilibrium with endogenous<br />
entry always implies that the leader is aggressive compared to each<br />
follower,<strong>and</strong>eachfollowereitherdoesnotenterorchoosesthesame<br />
strategy as in the Marshall equilibrium.<br />
The main result is that when entry in a market is endogenous, the leader<br />
<strong>of</strong> this market behaves always in an aggressive way, independently from the<br />
kind <strong>of</strong> strategic interaction that takes place with the followers. In particular,<br />
an accommodating behavior, which is typical <strong>of</strong> models <strong>of</strong> price competition<br />
(where SC holds) when entry is exogenously limited, will never emerge when<br />
the decision to enter in the market is endogenously taken by a sufficiently<br />
large number <strong>of</strong> potential entrants. Of course, this result is reminiscent <strong>of</strong><br />
what we found in the previous chapter: there leaders were always undertaking<br />
preliminary investments that were inducing an aggressive behavior in the<br />
market, here they directly undertake aggressive strategies in a preliminary<br />
stage. We can conclude that the aggressiveness <strong>of</strong> leaders facing endogenous<br />
entry is a fairly general result.<br />
Comparative statics. We could now investigate the way our equilibria are<br />
affected when we change some <strong>of</strong> the parameters, as we did in the previous<br />
chapter for the Nash <strong>and</strong> Marshall equilibria. Unfortunately, the comparative<br />
statics with respect to a generic parameter affecting the pr<strong>of</strong>it functions are<br />
quite complicated for both the Stackelberg equilibrium <strong>and</strong> the Stackelberg<br />
equilibrium with endogenous entry. In the second case, we can make some<br />
progress focusing on changes in the fixed cost. It turns out that results are<br />
typically the opposite if SS or SC holds. For simplicity, let us assume Π 22 ≥ 0,<br />
which will hold in most <strong>of</strong> our examples. 11 The main results are summarized<br />
in:<br />
Proposition 3.3. Consider a Stackelberg equilibrium with endogenous<br />
entry where entry occurs. Under SS, a) if Π 12 >Π 11 /h 0 (x),<br />
11 Π 22 > 0 holds in the case <strong>of</strong> quantity competition <strong>and</strong> perfectly substitute goods<br />
as long as dem<strong>and</strong> is convex, in our examples <strong>of</strong> price competition, <strong>and</strong> in the<br />
patent races <strong>of</strong> the next chapter.
100 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
the strategy <strong>of</strong> each firm is increasing <strong>and</strong> the number <strong>of</strong> firms is<br />
decreasing in F , b) otherwise, the strategy <strong>of</strong> entrants (leader) is<br />
increasing (decreasing) in F . Under SC, c) if Π12 L < (=)Π22Π L 1 L /Π2 L ,<br />
the strategy <strong>of</strong> entrants <strong>and</strong> their number are decreasing while the<br />
strategy <strong>of</strong> the leader is increasing in (independent from) F , d)<br />
otherwise, the strategy <strong>of</strong> each firm is decreasing in F .<br />
These results are more interesting when we interpret entry as a “general<br />
equilibrium” phenomenon determined by the pr<strong>of</strong>its available in other sectors.<br />
In this case, F can be re-interpreted as the pr<strong>of</strong>its available in another<br />
sector <strong>and</strong> a no arbitrage condition between sectors determine the entry decisions.<br />
As in the Marshall equilibrium case, a positive shock in another sector<br />
(increasing F ) tends to reduce entry <strong>and</strong> induce more aggressive strategies<br />
by the entrants under SS <strong>and</strong> more accommodating strategies under SC, but<br />
the strategies <strong>of</strong> the leaders may react in the opposite way (or remain unchanged).<br />
In the next section we will verify these results in models <strong>of</strong> quantity<br />
<strong>and</strong> price competition, <strong>and</strong> briefly in a simple model <strong>of</strong> competition for the<br />
market (generalized in the next chapter).<br />
3.3 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the<br />
<strong>Market</strong><br />
In the previous sections we characterized equilibria in markets with pure<br />
Stackelberg competition <strong>and</strong> with Stackelberg competition <strong>and</strong> endogenous<br />
entry in a general way. In Chapter 1 we analyzed a number <strong>of</strong> simple applications.<br />
In this section we will adopt an intermediate level <strong>of</strong> sophistication.<br />
3.3.1 <strong>Competition</strong> in Quantities<br />
The classic model <strong>of</strong> leadership due to Stackelberg (1934) is associated with<br />
competition in quantities <strong>and</strong> one firmcommittingtoitsownoutputbefore<br />
the other firms. Let us consider this situation under the following specification<br />
<strong>of</strong> the pr<strong>of</strong>it function:<br />
π i = x i p (x i ,β i ) − c(x i ) − F (3.13)<br />
where x i is the output <strong>of</strong> firm i, we allow for imperfect substitutability between<br />
goods (the inverse dem<strong>and</strong> is decreasing in both arguments) <strong>and</strong> we<br />
employ a general cost function.<br />
Exogenous entry. Let us first focus on the case <strong>of</strong> exogenous entry. Given<br />
the output <strong>of</strong> the leader x L , the equilibrium output <strong>of</strong> each follower will<br />
satisfy the pr<strong>of</strong>it maximizing condition:<br />
p (x, β)+xp 1 (x, β) =c 0 (x)
3.3 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 101<br />
where we remember that β = (n − 2)h(x) +h(x L ). The leader is aware<br />
that its strategy affects the choice <strong>of</strong> the followers according to the impact<br />
dx/dx L that can be derived from the above condition, <strong>and</strong> can choose its<br />
output taking this into account. 12 In Chapter 1 we solved for the Stackelberg<br />
equilibrium in the cases <strong>of</strong> a quadratic cost function with a linear dem<strong>and</strong><br />
<strong>and</strong> in the case <strong>of</strong> a linear cost function with a linear dem<strong>and</strong> <strong>and</strong> imperfect<br />
substitutability between goods. Beyond those examples things are already<br />
quite complex.<br />
To obtain more useful results, let us focus on the st<strong>and</strong>ard case <strong>of</strong> homogenous<br />
goods <strong>and</strong> constant marginal costs. Totally differentiating the equilibrium<br />
condition:<br />
p(X)+xp 0 (X) =c<br />
where X is total output, we obtain:<br />
dx −(1 − E)<br />
=<br />
dx L [n − E(n − 1)]<br />
Here E ≡−xp 00 (X)/p 0 (X) is the elasticity <strong>of</strong> the slope <strong>of</strong> the inverse dem<strong>and</strong><br />
with respect to the production <strong>of</strong> a follower, which we already encountered<br />
in the previous chapter, <strong>and</strong> which measures the degree <strong>of</strong> convexity <strong>of</strong> the<br />
dem<strong>and</strong> function. For instance, in the case <strong>of</strong> a linear dem<strong>and</strong>, like the one we<br />
studied in the example <strong>of</strong> Chapter 1, this elasticity was zero: in that case, an<br />
increase in the output <strong>of</strong> the leader was reducing the output <strong>of</strong> each follower<br />
by 1/n. A negative impact emerges whenever this elasticity is small enough,<br />
but for a high enough elasticity, the impact may turn out to be positive.<br />
Given the perceived reaction <strong>of</strong> the followers, the leader chooses its output<br />
to maximize pr<strong>of</strong>its π L =[p(X) − c] x L − F , which provides the optimality<br />
condition:<br />
∙<br />
p(X)+x L p 0 (X) 1+(n − 1) dx ¸<br />
= c<br />
dx L<br />
Joining the two equilibrium first order conditions <strong>and</strong> using the slope <strong>of</strong><br />
the reaction function, we can easily obtain a new general expression for the<br />
equilibrium output <strong>of</strong> the leader as a function <strong>of</strong> the equilibrium output <strong>of</strong><br />
the followers: 13<br />
12 If fixed costs <strong>of</strong> production are high enough, the leader can engage in entry<br />
deterrence, but now we focus on the case in which entry takes place.<br />
13 We can also solve for the equilibrium price under Stackelberg competition:<br />
p(X) =<br />
c<br />
1 − 1/ L [n − E(n − 1)]<br />
where L = −p(X)/p 0 (X)x L is the elasticity <strong>of</strong> dem<strong>and</strong> perceived by the leader.<br />
We could also calculate the market share <strong>of</strong> the leader, which is larger than 50%<br />
whenever E
102 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
x L = x [n − E(n − 1)] (3.14)<br />
We can easily verify that in the case <strong>of</strong> a linear dem<strong>and</strong> (E =0), the leader<br />
produces n times the output <strong>of</strong> the followers, a result we already encountered<br />
in Chapter 1. When the dem<strong>and</strong> is concave the leader produces even more<br />
than that, while in case <strong>of</strong> a convex dem<strong>and</strong> the leader produces less than<br />
that. Finally, notice that in the particular case in which E =1the first mover<br />
advantage disappears <strong>and</strong> the leader produces exactly the same as each one<br />
<strong>of</strong> the followers. This is not such an extreme result, as we will see in the<br />
following example.<br />
Consider the case <strong>of</strong> a hyperbolic dem<strong>and</strong> p =1/X, which can be derived<br />
from the logarithmic utility (2.16). After some tedious calculations, the<br />
Stackelberg equilibrium can be solved for the production levels:<br />
x L = 2n − 3<br />
4c(n − 1)<br />
x = 2n − 3<br />
4c(n − 1) 2<br />
Accordingly, the equilibrium price <strong>and</strong> the gross pr<strong>of</strong>its for the leader <strong>and</strong><br />
the followers are:<br />
p =<br />
2c(n − 1)<br />
2n − 3<br />
Π L =<br />
1<br />
4(n − 1)<br />
Π =<br />
1<br />
4(n − 1) 2<br />
First <strong>of</strong> all, notice that in the case where there are just two firms, the first<br />
mover advantage disappears: the choices <strong>of</strong> the two firms are strategically<br />
neutral in the Cournot duopolistic equilibrium (rather than complements or<br />
substitutes), <strong>and</strong> there is not an alternative commitment that can increase<br />
the pr<strong>of</strong>its <strong>of</strong> the leader. 14 When the number <strong>of</strong> firms increases, the output<br />
<strong>of</strong> the leader increases compared to the one<strong>of</strong>thefollowers:indeed,wecan<br />
verify that x L =(n − 1)x, whichsatisfies our general rule (3.14) for any<br />
number <strong>of</strong> firms. It follows that, with the exception <strong>of</strong> the duopoly case, we<br />
are always in a region where SS holds. Finally, one can also verify that total<br />
production is the same as under Cournot competition when there are just<br />
two firms, but it is higher whenever the number <strong>of</strong> firmsislargerthantwo.<br />
Endogenous entry. Let us move to the case <strong>of</strong> endogenous entry in the<br />
model <strong>of</strong> quantity competition with a leadership. Consider again the general<br />
pr<strong>of</strong>it function (3.13). The equilibrium first order condition for the followers<br />
<strong>and</strong> the endogenous entry condition are:<br />
p (x, β)+xp 1 (x, β) =c 0 (x)<br />
xp (x, β) =c(x)+F<br />
14 This is in line with our previous general result, since under this dem<strong>and</strong> function<br />
the elasticity <strong>of</strong> the inverse dem<strong>and</strong> is E =2x/X, whichsatisfies our general<br />
rule (3.14) for n =2only when x L = x.
3.3 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 103<br />
<strong>and</strong> they pin down the production <strong>of</strong> the followers x <strong>and</strong> their spillovers β<br />
independently from the production <strong>of</strong> the leader. Consequently, the pr<strong>of</strong>its <strong>of</strong><br />
the leader can be rewritten as:<br />
π L = x L p (x L ,β L ) − c(x L ) − F<br />
= x L p [x L ,β+ h(x) − h(x L )] − c(x L ) − F<br />
whose maximization delivers the optimality condition:<br />
p(x L ,β L )+x L [p 1 (x L ,β L ) − p 2 (x L ,β L )h 0 (x L )] = c 0 (x L ) (3.15)<br />
This relation provides the equilibrium production <strong>of</strong> the leader if goods are<br />
poor substitutes or the marginal cost is increasing enough, conditions that<br />
guarantee the existence <strong>of</strong> an interior solution. It emerges quite clearly that<br />
the leader is going to produce more than any follower.<br />
In particular, when goods are homogenous <strong>and</strong> the inverse dem<strong>and</strong> is simply<br />
p(X), the equilibrium condition for the leader boils down to an equation<br />
between the price <strong>and</strong> its marginal cost. In such a case, the equilibrium is<br />
fully charcterized by the following conditions:<br />
p(X) =<br />
c0 (x)<br />
1 − 1/ = c(x)+F<br />
x<br />
= c 0 (x L ) (3.16)<br />
where the first equality is a traditional mark up rule for the followers (with <br />
elasticity <strong>of</strong> dem<strong>and</strong>), the second equality is the endogenous entry condition,<br />
<strong>and</strong>thethirdonedefines the pricing rule <strong>of</strong> the leader. Notice that while the<br />
followers produce below the optimal scale (defined by the equality between<br />
marginal <strong>and</strong> average cost), the leader produces above this scale <strong>and</strong> obtains<br />
positive pr<strong>of</strong>its thanks to the increasing marginal costs.<br />
In Chapter 1 we studied an example <strong>of</strong> this result in the case <strong>of</strong> linear<br />
dem<strong>and</strong> (p = a − X) <strong>and</strong> linearly increasing marginal cost (equal to dx),<br />
where pr<strong>of</strong>its were given by (1.18). The equilibrium output <strong>of</strong> the leader <strong>and</strong><br />
the followers were:<br />
x L = 1+d<br />
d<br />
r<br />
2F<br />
2+d<br />
x =<br />
r<br />
2F<br />
2+d<br />
This simple set up with homogenous goods allows us to compare welfare<br />
under alternative forms <strong>of</strong> competition, namely Marshallian competition <strong>and</strong><br />
Stackelberg competition with endogenous entry. Since we know from Mankiw<br />
<strong>and</strong> Whinston (1986) that the Cournot case is characterized by too many<br />
firms producing too little, it is clear that Stackelberg competition does better<br />
on both dimensions. Hence, it is welfare improving to assign a leadership<br />
position to some firms despite this will give them a dominant position with<br />
associated extra-pr<strong>of</strong>its. This is a general result since our model implies that<br />
total production is always the same under Stackelberg <strong>and</strong> Cournot competition<br />
when there is endogenous entry, but a leader produces more than the
104 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
followers <strong>and</strong> consequently there are fewer firms in the Stackelberg case. The<br />
associated reduction in wasted fixed costs comes back in the form <strong>of</strong> pr<strong>of</strong>its<br />
for the leader. In conclusion, consumer surplus is the same, but welfare is<br />
higher under Stackelberg competition with endogenous entry:<br />
Proposition 3.4. Under endogenous entry <strong>and</strong> homogenous goods,<br />
as long as there is entry <strong>of</strong> some followers, Stackelberg competition<br />
in quantities is always Pareto superior with respect to Cournot<br />
competition.<br />
Another simple example <strong>of</strong> the aggressive behavior <strong>of</strong> the leader that we<br />
analyzed in Chapter 1 emerged in the model with product differentiation<br />
(dem<strong>and</strong> p i = a − (1 − b)x i − bX) <strong>and</strong> constant marginal cost (c), where<br />
pr<strong>of</strong>its were given by (1.25), <strong>and</strong> the equilibrium output <strong>of</strong> the leader <strong>and</strong> the<br />
followers were:<br />
x L =<br />
2 − b √ √<br />
F x = F<br />
2(1 − b)<br />
Again the leader produces more than the follower <strong>and</strong> sells at a price above<br />
its marginal cost. The consequence is that entry <strong>of</strong> followers is reduced. Since<br />
consumers value product differentiation in such a model the welfare consequences<br />
are more complex. Nevertheless, the reduction in the price <strong>of</strong> the<br />
leader more than compensates the reduction in the number <strong>of</strong> varieties <strong>and</strong><br />
consumer surplus is strictly increased by the leadership. 15 Therefore, in this<br />
case the consumers strictly gain from the aggressive pricing strategy <strong>of</strong> the<br />
leader even if this induces some firms to exit <strong>and</strong> reduces the number <strong>of</strong><br />
varieties provided in the market.<br />
Let us now move to the kind <strong>of</strong> equilibrium that can emerge when the<br />
interior solution characterized above does not maximize the pr<strong>of</strong>its <strong>of</strong> the<br />
leader. When goods are homogenous or highly substitute, or when the marginal<br />
cost is decreasing, constant or not too much increasing, the optimality<br />
for the leader implies a corner solution with entry deterrence <strong>and</strong>:<br />
15 Using the quadratic utility function (2.11) <strong>and</strong> the related dem<strong>and</strong> function, in<br />
equilibrium we have:<br />
⎡<br />
⎤<br />
U = Y + 1 n<br />
⎣ x 2 i + b <br />
x i x j<br />
⎦<br />
2<br />
i=1 i<br />
j6=i<br />
where Y is the exogenous income <strong>of</strong> the representative agent. One can verify<br />
that the gain in consumer surplus from the presence <strong>of</strong> a leader when entry is<br />
endogenous is:<br />
∆U =<br />
b(2 − b)F<br />
8(1 − b) > 0<br />
<strong>and</strong>thegaininwelfareis∆W = ∆U + π L . I am thankful to Nisvan Erkal <strong>and</strong><br />
Daniel Piccinin for insightful discussions on this point.
3.3 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 105<br />
xp [x, h(¯x L )] = c(x)+F ⇐⇒ ¯x L = β − x (3.17)<br />
We saw an example <strong>of</strong> this outcome in Chapter 1 within the basic model<br />
with homogeneous goods, linear dem<strong>and</strong> (p = a − X) <strong>and</strong> constant marginal<br />
costs c, where pr<strong>of</strong>its were given by (1.2). In that case, the equilibrium output,<br />
produced entirely by the leader was:<br />
¯x L = a − c − 2 √ F<br />
Moreover,inthatcasewenoticedthat welfare was greater under Stackelberg<br />
competition with entry deterrence rather than Cournot competition<br />
with free entry because total production was reduced but the pr<strong>of</strong>its <strong>of</strong> the<br />
leader <strong>and</strong> the savings in fixed costs were enough to compensate the lower<br />
consumer surplus.<br />
Another simple case emerges with the hyperbolic dem<strong>and</strong> (p = 1/X)<br />
<strong>and</strong> with constant marginal cost c. Now, the Stackelberg equilibrium with<br />
endogenous entry requires entry deterrence with production:<br />
³<br />
1 − √ ´2<br />
F<br />
¯x L =<br />
c<br />
In the case <strong>of</strong> general dem<strong>and</strong> functions for homogenous goods, we can actually<br />
find a simple sufficient condition for entry-deterrence which only depends<br />
on the shape <strong>of</strong> the cost function:<br />
Proposition 3.5. Under endogenous entry <strong>and</strong> homogenous goods,<br />
whenever marginal costs <strong>of</strong> production are constant or decreasing,<br />
Stackelberg competition in quantities always delivers entrydeterrence<br />
with only the leader in the market.<br />
This result can contribute to clarify the old debate on limit pricing. Entry<br />
deterrence through this forms <strong>of</strong> limit pricing is the equilibrium strategy for<br />
leaders facing endogenous entry for any dem<strong>and</strong> function as long as goods<br />
are homogenous (or highly substitutable) <strong>and</strong> returns to scale are constant<br />
or decreasing. 16 As both our examples show, the entry deterrence production<br />
is decreasing in the fixed cost, since this cost helps the leader to exclude<br />
the rivals. When the fixed cost diminishes the equilibrium output <strong>of</strong> the<br />
leader increases, <strong>and</strong> when it approaches zero, the equilibrium approaches<br />
the competitive outcome with a price equal to the marginal cost (indeed<br />
both ¯x L = a − c in the case <strong>of</strong> linear dem<strong>and</strong> <strong>and</strong> ¯x L =1/c inthecase<strong>of</strong><br />
hyperbolic dem<strong>and</strong> correspond to a price p = c). Nevertheless, this efficient<br />
output level is still entirely produced by one single firm, the leader.<br />
16 This corresponds to the result <strong>of</strong> the contestable markets theory <strong>of</strong> Baumol et al.<br />
(1982). However, that theory generates a limit price which implies zero pr<strong>of</strong>its<br />
for the leader. For instance, with the hyperbolic dem<strong>and</strong> the limit pricing would<br />
equate inverse dem<strong>and</strong> <strong>and</strong> average cost (p =1/x = c + F/x), which implies<br />
x =(1− F )/c > ¯x L.
106 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
3.3.2 <strong>Competition</strong> in Prices<br />
The role <strong>of</strong> price leadership is <strong>of</strong>ten underestimated for two main reasons.<br />
The first is that commitments to prices are hardly credible when it is easy<br />
<strong>and</strong> relatively inexpensive to change prices. While this is true for long term<br />
commitments, it is also true that short term commitments can be credible<br />
in most markets. The macroeconomic literature on price stickiness has developed<br />
a number <strong>of</strong> arguments on why this may be the case, ranging from small<br />
menu costs <strong>of</strong> price adjustments to costs in the acquisition <strong>of</strong> information to<br />
reoptimize. The second reason for which a price leadership may poorly describe<br />
the behavior <strong>of</strong> market leaders is probably more pervasive <strong>and</strong> relies<br />
on the absence <strong>of</strong> a first mover advantage in simple models <strong>of</strong> competition<br />
in prices. For instance, in st<strong>and</strong>ard duopolies, a price leader obtains lower<br />
pr<strong>of</strong>its than its follower, <strong>and</strong> for this reason neither one nor the other firm<br />
would like to be a leader: there is actually a second mover advantage. As we<br />
will see, this result disappears <strong>and</strong> the first mover advantage is back when<br />
entry in the market is endogenous.<br />
In our general class <strong>of</strong> models with price competition the pr<strong>of</strong>it function<br />
is given by:<br />
π i =(p i − c) D (p i ,β i ) − F (3.18)<br />
where p i is <strong>of</strong> course the price <strong>of</strong> firm i, <strong>and</strong> the dem<strong>and</strong> function is decreasing<br />
in both arguments, with β i = P j6=i<br />
g(p) for some positive <strong>and</strong> decreasing<br />
function g. Notice that the model is nested in our framework (3.1) after<br />
setting x i =1/p i as the strategic variable (see Section 2.4.2 for a discussion).<br />
Exogenous entry. Let us consider first the case <strong>of</strong> exogenous entry. Stackelberg<br />
equilibrium with n firms is characterized by the following equilibrium<br />
optimality conditions for the followers <strong>and</strong> the leader: 17<br />
D (p, β)+(p − c)D 1 (p, β) =0<br />
∙<br />
µ ¸ dβL<br />
D (p L ,β L )+(p L − c) D 1 (p L ,β L )+D 2 (p L ,β L ) =0 (3.19)<br />
dp L<br />
where dβ L /dp L < 0 can be derived by the optimality condition <strong>of</strong> the followers<br />
as long as SC holds. While the equilibrium conditions soon become quite<br />
complex, the positive last term shows that the leader chooses a price above<br />
the one <strong>of</strong> the followers, inducing a general increase in prices compared to<br />
the Nash-Bertr<strong>and</strong> equilibrium between the same firms.Thechoice<strong>of</strong>ahigh<br />
price by the leader is aimed at s<strong>of</strong>tening price competition, but it also leads<br />
the followers to make more pr<strong>of</strong>its by choosing a lower price <strong>and</strong> stealing<br />
market shares from the leader.<br />
17 If fixed costs <strong>of</strong> production are high enough, the leader can engage in entry<br />
deterrence, but here we will focus on the case in which entry is accommodated.
3.3 <strong>Competition</strong> in Quantities, in Prices <strong>and</strong> for the <strong>Market</strong> 107<br />
Endogenous entry. Let us now look at the Stackelberg equilibrium with<br />
endogenous entry. The optimality condition for the followers <strong>and</strong> the endogenous<br />
entry condition are:<br />
D (p, β)+(p − c)D 1 (p, β) =0<br />
(p − c) D (p, β) =F<br />
<strong>and</strong> they pin down the price <strong>of</strong> the followers p <strong>and</strong> their spillovers β =(n −<br />
1)g(p), so that the pr<strong>of</strong>it <strong>of</strong> the leader becomes:<br />
π L =(p L − c)D [p L , (n − 1)g(p) − g(p L )] − F =<br />
=(p L − c)D [p L ,β+ g(p) − g(p L )] − F<br />
Pr<strong>of</strong>it maximization delivers the equilibrium condition:<br />
D(p L ,β L )+(p L − c)[D 1 (p L ,β L ) − D 2 (p L ,β L )g 0 (p L )] = 0 (3.20)<br />
which implies a lower price p L than the price <strong>of</strong> the followers, since the last<br />
term is negative. This is a crucial result by itself since we are quite familiar<br />
with associating price competition <strong>and</strong> accommodating leaders setting<br />
higher prices than the followers: this st<strong>and</strong>ard outcome collapses under endogenous<br />
entry. Moreover, the leader is now obtaining positive pr<strong>of</strong>its, while<br />
each follower does not gain any pr<strong>of</strong>its: the first mover advantage is back.<br />
In Chapter 1 we have seen an example based on the Logit dem<strong>and</strong> (2.21),<br />
where the equilibrium prices were:<br />
p L = c + 1 λ<br />
p = c + 1 λ + F N<br />
Moreover, using the micr<strong>of</strong>oundation pointed out by Anderson et al. (1992)<br />
in terms <strong>of</strong> the quasilinear utility (2.22), one can show that this equilibrium<br />
is Pareto efficient compared to the correspondent Marshall equilibrium: the<br />
reduction in the price <strong>of</strong> the leader reduces entry, leaves unchanged consumer<br />
surplus <strong>and</strong> increases firms’ pr<strong>of</strong>its, inducing an increase in total welfare.<br />
In the case <strong>of</strong> the isoelastic dem<strong>and</strong> (2.24) derived in the last chapter from<br />
the utility function (2.23), we obtain the following prices:<br />
p L = c θ<br />
p =<br />
cY<br />
θ [Y − F (1 + α)]<br />
where <strong>of</strong> course the leader applies a lower mark up than each follower. 18 It<br />
can be verified that in any version <strong>of</strong> the Dixit-Stiglitz model where 1/(1−θ)<br />
18 As we already noticed, we could analyze competition in quantities within the<br />
same model - one can obtain the inverse dem<strong>and</strong> from (2.23). Since there is<br />
product differentiation, also that case would entail a higher output for the leader,<br />
<strong>and</strong> consequently a lower price.
108 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
is the constant elasticity <strong>of</strong> substitution between goods <strong>and</strong> c is the marginal<br />
cost <strong>of</strong> production, as long as entry is endogenous, the leader will choose the<br />
price p L = c/θ <strong>and</strong> the followers will choose a higher price. Indeed, free entry<br />
pins down the price index that is perceived by the leader, whose optimization<br />
problem is <strong>of</strong> the following kind:<br />
max(p L − c)D L ∝ (p L − c)p − 1<br />
1−θ<br />
L<br />
which always delivers the price above. As a consequence, the leader produces<br />
more than each follower <strong>and</strong> the number <strong>of</strong> followers is reduced compared<br />
to the Marshall equilibrium. Once again, however, consumer surplus is not<br />
changed because the price index is unaffected. Since the leader obtains positive<br />
pr<strong>of</strong>its, overall welfare is increased. We summarize these results as follows:<br />
Proposition 3.6. In a model <strong>of</strong> price competition with Logit dem<strong>and</strong><br />
or Dixit-Stiglitz dem<strong>and</strong> <strong>and</strong> endogenous entry, a leader sells<br />
its variety at a lower price than the entrants, inducing a Pareto improvement<br />
in the allocation <strong>of</strong> resources.<br />
In all <strong>of</strong> these models we can verify the existence <strong>of</strong> an unambiguous<br />
ranking <strong>of</strong> market structures from a welfare point <strong>of</strong> view. Indeed, from the<br />
best to the worst case for welfare we have: 1) endogenous entry with a leader;<br />
2) endogenous entry without a leader; 3) barriers to entry without a leader;<br />
4) barriers to entry with a leader. If we look at consumer surplus only, case<br />
1) <strong>and</strong> 2) deliver the same utility for the consumers, but the rest <strong>of</strong> the<br />
ranking is unchanged. This welfare results have important consequences for<br />
the evaluation <strong>of</strong> market leaders <strong>and</strong> for antitrust policy: we will return on<br />
them in Chapter 5.<br />
3.3.3 <strong>Competition</strong> for the <strong>Market</strong><br />
In Chapter 1 we studied a simple model <strong>of</strong> competition for the market where<br />
firms were investing to obtain a reward V . Under the specification:<br />
π i = x i<br />
n<br />
Y<br />
j=1,j6=i<br />
(1 − x j ) V − x2 i<br />
2<br />
(3.21)<br />
with x i investment in R&D for firm i, we found that because <strong>of</strong> SS, a Stackelberg<br />
equilibrium was characterized by a leader investing more than each<br />
follower, while a Stackelberg equilibrium with endogenous entry was characterized<br />
by only the leader investing:<br />
√<br />
2F<br />
¯x L =1−<br />
V<br />
These results are not general, since more realistic descriptions <strong>of</strong> the market<br />
for innovations can lead to different results. Nevertheless, as we will see in the<br />
next chapter, which focuses entirely on competition for the market, a leader<br />
always invests more than any other firm when entry is endogenous.
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple Strategies 109<br />
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple<br />
Strategies<br />
The results <strong>of</strong> the previous sections can be extended in many directions to be<br />
able to describe market structures in a more realistic way. This section will<br />
consider a few directions: introducing a technological asymmetry between<br />
the leader <strong>and</strong> the followers, extending the model to multiple leaders, endogenizing<br />
the same leadership status, allowing for multiple strategies <strong>and</strong><br />
considering more general pr<strong>of</strong>it functions. Our main focus, at this point, will<br />
be on the case where entry is endogenous, which we believe to be more relevant<br />
in most markets.<br />
3.4.1 Asymmetries Between Leader <strong>and</strong> Followers<br />
In this section, following Etro (2002), we assume that the leader has the pr<strong>of</strong>it<br />
function:<br />
π L = Π L (x L ,β L ,K) − F<br />
where K is a new parameter specific totheleader(itmaywellbeavector<br />
<strong>of</strong> parameters). The basic assumptions are Π3<br />
L ≡ ∂Π L /∂K > 0 <strong>and</strong><br />
Π L (x, β, 0) = Π i (x, β). Notice that, while this specification may look like the<br />
one analyzed in Chapter 2, here we are talking about an exogenous parameter<br />
K, not an endogenous one. We are interested in underst<strong>and</strong>ing how exogenous<br />
asymmetries affect the behavior <strong>of</strong> market leaders, <strong>and</strong> not how market<br />
leaders endogenously create asymmetries to affect their behavior (which was<br />
the purpose <strong>of</strong> the analysis <strong>of</strong> the previous chapter).<br />
A first mover advantage is <strong>of</strong>ten associated with some asymmetry between<br />
the leader <strong>and</strong> the followers. For instance, in the simple model <strong>of</strong> competition<br />
for the market <strong>of</strong> Chapter 1 we extended the basic model to consider leaders<br />
that were also incumbent monopolists with a flow <strong>of</strong> current pr<strong>of</strong>its affecting<br />
their expected pr<strong>of</strong>its. In other cases, it is natural to link the first mover<br />
advantage with some technological or market advantage, for instance a lower<br />
marginal cost c(K) for the leader (with c 0 (K) < 0), or other differences as<br />
those suggested in the previous chapter.<br />
In general, when entry is endogenous we obtain a strategy <strong>of</strong> the leader<br />
which depends on K, x L = x L (K), <strong>and</strong> hence the number <strong>of</strong> entrants, but<br />
not their individual strategy, also depends on K. One can show:<br />
Proposition 3.7. An asymmetric Stackelberg equilibrium with<br />
endogenous entry implies that the leader is aggressive whenever<br />
Π13 L ≥ Π23(Π L 1 L /Π2 L ) or K is small enough.<br />
The intuition is the following: an increase in the advantage <strong>of</strong> the leader<br />
(that is in K) induces a higher incentive to aggressiveness if it raises the
110 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
marginal benefit from it more than the change in its marginal cost. Indeed<br />
the sufficient condition could be rewritten as ∂(Π1 L /Π2 L )/∂K ≤ 0, thatis<br />
the marginal rate <strong>of</strong> substitution between x L <strong>and</strong> β L is decreasing in K. If<br />
this condition does not hold, it means that x 0 L (K) < 0, therefore for a great<br />
enough K (a strong enough asymmetry) the leader may be accommodating<br />
(x L (K) <br />
0 <strong>and</strong> Π23 L = 0, <strong>and</strong> under competition in prices we have Π13 L > 0 <strong>and</strong><br />
Π23 L < 0. Similarly one can examine other kinds <strong>of</strong> exogenous asymmetries<br />
(as those we examined in the previous chapter on the dem<strong>and</strong> side, in the<br />
financial structure, in complementary markets, <strong>and</strong> so on) <strong>and</strong> verify how the<br />
incentives <strong>of</strong> the leader to be aggressive are changed.<br />
3.4.2 Multiple Leaders<br />
Until now we considered a simple game with just one leader playing in the<br />
first stage. Here we will consider the case in which multiple leaders play<br />
simultaneously in the first stage. Hence the timing <strong>of</strong> the game becomes the<br />
following: 1) in the first stage, m leaders simultaneously choose their own<br />
strategies; 2) in the second stage, potential entrants decide whether to enter<br />
or not; 3) in the third stage each one <strong>of</strong> the n − m followers that entered<br />
chooses its own strategy. In the next section we will discuss how to endogenize<br />
m.<br />
When entry is endogenous we should consider two different situations: one<br />
in which entry <strong>of</strong> followers is not deterred in equilibrium <strong>and</strong> one in which the<br />
leaders deter entry. Consider first the case in which the number <strong>of</strong> leaders m is<br />
small enough, or the cost <strong>of</strong> deterrence is large enough that entry <strong>of</strong> followers<br />
takes place in equilibrium. In such a case, the behavior <strong>of</strong> the leaders can be<br />
characterized in a similar fashion to our basic analysis. Moreover, contrary<br />
to what happens when the number <strong>of</strong> firms n in the market is exogenous (in<br />
that case the number <strong>of</strong> leaders m affects their strategic interaction, their<br />
strategies <strong>and</strong> their pr<strong>of</strong>its), with endogenous entry the number <strong>of</strong> leaders<br />
does not affect their strategies, still given by (3.12), <strong>and</strong> their pr<strong>of</strong>its:<br />
Proposition 3.8. Under Stackelberg competition with m leaders,<br />
as long as there is endogenous entry <strong>of</strong> some followers, each leader<br />
is aggressive compared to each follower <strong>and</strong> its strategy <strong>and</strong> pr<strong>of</strong>its<br />
are independent from m.<br />
This confirms the spirit <strong>of</strong> our results with a single leader. Each one <strong>of</strong><br />
the leaders now behaves in an aggressive way compared to the followers <strong>and</strong><br />
also independently from the other leaders. For instance, under competition<br />
in quantities <strong>and</strong> increasing marginal cost, each leader produces the same
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple Strategies 111<br />
output that equates the marginal cost to the price, <strong>and</strong> the equilibrium price<br />
equates the optimal mark up <strong>of</strong> the followers to the fixed cost <strong>of</strong> production.<br />
While the pr<strong>of</strong>it <strong>of</strong> each leader is not affected by the number <strong>of</strong> leaders,<br />
thenumber<strong>of</strong>entrantsisclearlydecreasedbyanincreaseinthenumber<strong>of</strong><br />
leaders.<br />
The situation is more complicated if there is entry deterrence in equilibrium.<br />
In the case <strong>of</strong> an exogenous number <strong>of</strong> firms, entry deterrence is a sort<br />
<strong>of</strong> public good for the leaders, which may introduce free-riding issues in their<br />
behavior. Gilbert <strong>and</strong> Vives (1986) have analyzed this issue in a model with<br />
m leaders facing a potential entrant, while Tesoriere (2006) has extended the<br />
model in Etro (2002) to analyze the case <strong>of</strong> m leaders facing endogenous entry.<br />
The result can easily be seen through a simple example with two leaders.<br />
Let us analyze a model <strong>of</strong> competition in quantities with a linear inverse<br />
dem<strong>and</strong> p = a − X, constant marginal cost c, m =2<strong>and</strong> endogenous entry<br />
<strong>of</strong> followers. Remember that the entry deterring output in this model is ¯x =<br />
a − c − 2 √ F . Consider the best response <strong>of</strong> one leader, say L 1 .Iftheoutput<br />
<strong>of</strong> the other leader, say L 2 , is already above the entry deterrence level, x L2 ><br />
¯x, the best strategy is clearly x L1 =0. However, whenever the output <strong>of</strong><br />
the second leader is below the entry deterring level, it is always optimal to<br />
produce at least enough to deter the entry <strong>of</strong> any follower, which implies<br />
x L1 ≥ ¯x − x L2 . Nevertheless, it may be optimal to produce more than this<br />
when the st<strong>and</strong>ard Cournot best response, namely x L1 =(a − c − x L2 ) /2,<br />
generates a higher output than the level that is sufficient to deter entry, which<br />
happens for x L2 >a− c − 4 √ F . An analogous rule drives the best response<br />
for the second leader. In summary, the best response function for a leader L i<br />
with i, j =1, 2 is:<br />
x Li (x Lj )=<br />
(<br />
h<br />
max<br />
)<br />
0<br />
i<br />
if x Lj ≥ ¯x<br />
if x Lj < ¯x<br />
a − c − 2 √ F − x Lj ; a−c−xLj<br />
2<br />
that can be rewritten as:<br />
⎧<br />
⎨ 0 x Lj ≥ a − c − 2 √ ⎫<br />
F<br />
x Li (x Lj )= (a − c − x<br />
⎩<br />
Lj ) /2 if x Lj ∈ [a − c − 4 √ F ; a − c − 2 √ ⎬<br />
F )<br />
a − c − 2 √ F − x Lj<br />
x Lj
112 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
x Li = a − c − 2 √ F , x Lj =0<br />
Moreover, there are other possible equilibria with both the leaders active<br />
in the market. We need to distinguish two cases depending on the size <strong>of</strong> the<br />
fixed cost. When the fixed cost is high enough, the st<strong>and</strong>ard equilibrium <strong>of</strong> the<br />
Cournot duopoly is an equilibrium <strong>of</strong> Stackelberg competition in quantities<br />
with endogenous entry <strong>and</strong> two leaders, since it implies a high enough output<br />
so that further entry is deterred. Since in a symmetric Cournot duopoly each<br />
firm (each leader here) produces:<br />
x L1 = x L2 = a − c<br />
3<br />
this equilibrium requires that pr<strong>of</strong>its are positive for both firms, or F <<br />
(a − c) 2 /9, <strong>and</strong> that total output is enough to deter entry <strong>of</strong> any follower,<br />
2(a − c)/3 > ¯x or F>(a − c) 2 /36. Whenthefixed cost is lower than this last<br />
cut-<strong>of</strong>f, however, the two best response functions overlap in an intermediate<br />
region where aggregate production is just enough to deter entry, <strong>and</strong> we have<br />
a continuum <strong>of</strong> equilibria with x L1 + x L2 = a − c − 2 √ F <strong>and</strong> such that both<br />
firms produce enough to obtain positive pr<strong>of</strong>its. This requires:<br />
x L1 = a − c − 2 √ h<br />
F − x L2 <strong>and</strong> x L2 ∈ a − c − 4 √ F ;2 √ i<br />
F<br />
In summary, Stackelberg equilibria in quantities with two leaders <strong>and</strong><br />
endogenous entry in the case <strong>of</strong> linear dem<strong>and</strong> <strong>and</strong> marginal cost are always<br />
characterized by entry deterrence with the following possible configurations<br />
<strong>of</strong> production by the leaders:<br />
⎧<br />
⎪⎨<br />
x Li = a − c − 2 √ F , x Lj =0 for any F< 4(a−c)2<br />
x L1 =(a − c) /3 <strong>and</strong> x L2 =(a − c) /3 if F ∈<br />
⎪⎩ any x Li = a − c − 2 √ h<br />
F − x Lj ∈ a − c − 4 √ F ;2 √ i<br />
F<br />
h 25<br />
(a−c)<br />
2<br />
36<br />
; (a−c)2<br />
9<br />
i<br />
if F< (a−c)2<br />
36<br />
Tesoriere (2006) generalizes this example to m leaders, showing that endogenous<br />
entry <strong>of</strong> followers is always deterred, 20 <strong>and</strong> there is always an equilibrium<br />
with just one leader producing the entry deterrence output <strong>and</strong> the<br />
remaining leaders producing zero. Furthermore, the symmetric Cournot equilibrium<br />
between all the m leaders can be an equilibrium when total Cournot<br />
output <strong>of</strong> the m leaders exceeds the entry deterrent output, <strong>and</strong> there can<br />
be a continuum <strong>of</strong> equilibria with aggregate production equal to the entry<br />
deterrent level when the fixed cost <strong>of</strong> production is low enough. Hence, underinvestment<br />
in entry deterrence cannot occur when entry is endogenous,<br />
while overinvestment in entry deterrence can occur (but leaders always obtain<br />
strictly positive pr<strong>of</strong>its). Once again, this outcome remains in the spirit<br />
<strong>of</strong> our results about the aggressive behavior <strong>of</strong> market leaders.<br />
20 As we have seen from the case <strong>of</strong> a single leader, under constant marginal costs,<br />
entry deterrence occurs for any dem<strong>and</strong> function.<br />
⎫<br />
⎪⎬<br />
⎪⎭
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple Strategies 113<br />
3.4.3 Endogenous Leadership<br />
After developing a Stackelberg model with multiple leaders <strong>and</strong> endogenous<br />
entry <strong>of</strong> followers, it is natural to verify what happens when there is endogenous<br />
entry <strong>of</strong> leaders as well.<br />
The simplest way to endogenize the number <strong>of</strong> leaders is by adding an<br />
initial stage <strong>of</strong> the game where firms decide simultaneously whether or not to<br />
become a leader. 21 Any firm can make an investment, say I, whichprovides<br />
the status <strong>of</strong> a leader in the market, while any firm that does not invest<br />
can only enter in the market as a follower: in other words, commitment to<br />
strategies is costly. As Prop. 3.8 suggests, as long as there is entry <strong>of</strong> followers,<br />
it must be that all leaders obtain the same level <strong>of</strong> positive pr<strong>of</strong>its (which<br />
is independent from the number <strong>of</strong> leaders m). Therefore, if the investment<br />
needed to become a leader is small enough, there must always be incentives<br />
to invest to become leaders when this does not deter entry <strong>of</strong> followers. Then,<br />
consider the largest number <strong>of</strong> leaders compatible with some entry, say M.<br />
Given this number <strong>of</strong> leaders, another firm may invest in leadership <strong>and</strong><br />
subsequently engage in Nash competition with only the other leaders (entry<br />
<strong>of</strong> followers is now deterred by construction). If such an entry is pr<strong>of</strong>itable,<br />
the equilibrium must imply only leaders in the market <strong>and</strong> an endogenous<br />
number m ∗ >M derived from a free entry condition with a fixed cost F + I<br />
(clearly this happens whenever the cost <strong>of</strong> leadership is zero or small enough).<br />
If this is not the case, the only equilibrium implies m ∗ = M firms investing in<br />
leadership <strong>and</strong> a residual competitive fringe <strong>of</strong> followers: once again, as Prop.<br />
3.8 still implies, all leaders would be aggressive compared to each follower.<br />
Another interesting situation emerges when entry is sequential, leading<br />
to a hierarchical leadership. While a general treatment <strong>of</strong> sequential games<br />
is complex, Vives (1988) <strong>and</strong> Anderson <strong>and</strong> Engers (1994) have fully characterized<br />
sequential competition in quantities with linear costs <strong>and</strong> isoelastic<br />
dem<strong>and</strong>, <strong>and</strong> with an exogenous number <strong>of</strong> firms. 22 Their analysis makes clear<br />
that in the case <strong>of</strong> endogenous entry the only possible equilibrium would imply<br />
entry deterrence. 23<br />
21 See Hamilton <strong>and</strong> Slutsky (1990).<br />
22 See Prescott <strong>and</strong> Visscher (1977) for an early discussion. Economides (1993) studies<br />
free entry in a game with simultaneous entry at the first stage <strong>and</strong> sequential<br />
quantity decisions between the entrants.<br />
23 Tesoriere (2006) studies the following extension <strong>of</strong> Etro (2002): at a first preplay<br />
stage firms simultaneously decide whether or not to enter the market <strong>and</strong><br />
at which period t ∈ T , then at each stage t =1, 2, ..., T ,eachfirm that has<br />
chosen to enter at stage t decides how much to produce, knowing the production<br />
chosen by all the firms that entered in the previous periods, taking as given that<br />
<strong>of</strong> the other firms that enter at the same time t, <strong>and</strong> anticipating correctly the<br />
strategies <strong>of</strong> the later movers. Focusing on the case <strong>of</strong> constant marginal costs, he<br />
shows that at any Subgame Perfect Nash Equilibrium, endogenous entry occurs
114 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
3.4.4 Multiple Strategies<br />
In this section we will show that a weaker version <strong>of</strong> the result on aggressive<br />
leaders generalizes when firms choose multiple strategic variables.<br />
Imagine that each firm chooses a vector <strong>of</strong> K ≥ 1 strategic variables<br />
x i =[x i1 ,x i2 , ..., x iK ] ∈ < K + , <strong>and</strong> its well behaved pr<strong>of</strong>it function can be<br />
written as:<br />
⎡<br />
⎤<br />
π i = Π ⎣x i ; X h(x j ) ⎦ − F (3.22)<br />
j6=i<br />
with h : < K + → < + differentiable <strong>and</strong> increasing in all its arguments. Examples<br />
are models in which the leader sells more than one good, models with<br />
multimarket competition or competition in quality <strong>and</strong> price, models with<br />
multiple inputs in the production function, patent races with multiple investments<br />
<strong>and</strong> so on. Clearly these are very important cases in real world<br />
industries. Results on the behavior <strong>of</strong> leaders in similar markets with an exogenous<br />
number <strong>of</strong> firms are complicated <strong>and</strong> ambiguous since they depend<br />
on all the possible cross derivatives <strong>and</strong> therefore on many specific properties<br />
<strong>of</strong> the markets (see Bulow et al., 1985). Nevertheless, under endogenous<br />
entry, a weaker version <strong>of</strong> our result still holds.<br />
Define firm i “on average” more aggressive than firm j if h(x i ) >h(x j ).<br />
Then, in equilibrium we have a vector x for the followers which is independent<br />
from the leader’s strategies, <strong>and</strong> the following equilibrium conditions for the<br />
strategies <strong>of</strong> the leader:<br />
∂Π L (x L ,β L )<br />
∂x Lk<br />
=<br />
µ ∂h(x)<br />
∂x Lk<br />
∂Π L (x,β L )<br />
∂β L<br />
≤ 0 for all k (3.23)<br />
These conditions do not imply that the leader is more aggressive in all the<br />
strategies, but that it must be more aggressive in some strategies. Moreover,<br />
they allow us to derive:<br />
immediately <strong>and</strong> simultaneously, <strong>and</strong> that any <strong>of</strong> the following configurations is<br />
an equilibrium:<br />
1) one <strong>of</strong> the firms enters in t =1<strong>and</strong> produces the entry deterring output;<br />
2) m firms enter in t =1<strong>and</strong> produce the entry deterring output in aggregate,<br />
when the Cournot equilibrium with m firms would imply a lower aggregate<br />
output than the entry deterring output;<br />
3) m firms enter in t = 1 <strong>and</strong> produce as in a Cournot equilibrium with<br />
m firms, when this implies an aggregate output which is larger than the entry<br />
deterring output.<br />
No other configuration is sustainable as equilibrium, but the above characterization<br />
<strong>of</strong> equilibria already exhibits multiplicity. At any <strong>of</strong> the possible equilibria,<br />
however, only market leaders produce: when the leadership is endogenous,<br />
sequential production is never observed.
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple Strategies 115<br />
Proposition 3.9. A Stackelberg equilibrium with endogenous entry<br />
<strong>and</strong> multiple strategic variables always implies that the leader<br />
is on average more aggressive than each follower.<br />
To see how to use this result we develop two examples.<br />
Capital/labour choices. Let us extend the simplest model <strong>of</strong> competition<br />
in quantities with a production function using two inputs, say capital k i<br />
<strong>and</strong> labour l i according to a Cobb-Douglas specification x i = ki αlη i with<br />
0
116 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
leader will be more aggressive than the followers on average, which means<br />
that h(θ L ,q L ) >h(θ, q). But this implies g(1/θ L ) >g(1/θ) or, using the fact<br />
that g is a decreasing function, that θ L >θ. We can then conclude that in a<br />
Stackelberg equilibrium in price <strong>and</strong> quality with endogenous entry the leader<br />
supplies a good with a better quality-price ratio than each other follower. 24<br />
3.4.5 General Pr<strong>of</strong>it Functions<br />
In this chapter we examined the behavior <strong>of</strong> firms with a first mover advantage<br />
over their competitors in the choice <strong>of</strong> the market strategy. A general result<br />
that emerges in the presence <strong>of</strong> endogenous entry is that leaders tend to<br />
behave in an aggressive way, in particular they choose lower prices <strong>and</strong> higher<br />
output than their followers. While we noticed that the spirit <strong>of</strong> this result is<br />
robust to a number <strong>of</strong> extensions, at least under some regularity conditions,<br />
we are aware that we had to impose a considerable amount <strong>of</strong> symmetry in<br />
the general model adopted in this book to obtain the simple results described<br />
until now. 25 For instance, our simple results do not apply to models where<br />
pr<strong>of</strong>its depend on the number <strong>of</strong> firms in a more complex way 26 or when<br />
conjectural variations <strong>of</strong> the firms are not restricted to the Nash case.<br />
Nevertheless, as shown in Etro (2008), also in a more general framework<br />
there is still a tendency <strong>of</strong> the market leaders to be aggressive toward a fringe<br />
<strong>of</strong> competitors that endogenously enter in the market. In particular, generalizing<br />
our analysis to the case <strong>of</strong> dem<strong>and</strong> functions exhibiting strong forms <strong>of</strong><br />
loveforvariety,wehaveverified that the tendency toward an aggressive pricing<br />
<strong>of</strong> the leaders facing endogenous entry remains, but the behavior <strong>of</strong> the<br />
followersisnowaffected. For instance, consider the classic case <strong>of</strong> imperfect<br />
substitutability with linear dem<strong>and</strong>s p i = a − x i + b P j6=i x j, that we studied<br />
many times in this book <strong>and</strong> that can be derived from a quadratic utility<br />
function as (2.11). Inverting the system, we obtain the direct dem<strong>and</strong>s:<br />
b<br />
1−b<br />
D i = a − p i +<br />
1+b(n − 1)<br />
P<br />
j6=i (p j − p i )<br />
(3.24)<br />
24 Similarly, in a generalized version <strong>of</strong> the Logit model (1.34) with dem<strong>and</strong> for good<br />
i equal to D i = Ne u n<br />
<br />
i<br />
/<br />
j=1 eu j<br />
,whereu i = q i − λp i parameterizes utility<br />
from purchasing good i, a leader facing endogenous entry would choose quality<br />
<strong>and</strong> price so as to provide higher utility to the consumers than the followers.<br />
Analogous results emerge if quality does not affect marginal costs but it affects<br />
fixed costs. On quality choices in the Logit model see also Anderson et al. (1992,<br />
Ch. 7).<br />
25 For a critique to the generality <strong>of</strong> our characterization <strong>of</strong> Stackelberg equilibria<br />
with endogenous entry, <strong>and</strong> to the implications that can be drawn from it, see<br />
Encaoua (2006).<br />
26 As with the dem<strong>and</strong> function <strong>of</strong> Shubik (1980), which, however, does not generate<br />
love for variety (see Erkal <strong>and</strong> Piccinin, 2007a).
3.4 Asymmetries, Multiple Leaders <strong>and</strong> Multiple Strategies 117<br />
It can be verified that the pr<strong>of</strong>it function associated with this case is not<br />
nested in our general framework. Nevertheless, it is well behaved <strong>and</strong> it is<br />
decreasing in the number <strong>of</strong> firms for given strategies. Since prices are strategic<br />
complements, the Stackelberg equilibrium with an exogenous number <strong>of</strong><br />
firms is characterized by a higher price for the leader compared to the followers.<br />
Contrary to this, the Stackelberg equilibrium with endogenous entry is<br />
characterized by a lower price for the leader compared to the followers. Moreover,<br />
the price <strong>of</strong> the leader is below the equilibrium price in the Marshall<br />
equilibrium, while the price <strong>of</strong> the followers is above it <strong>and</strong> the number <strong>of</strong><br />
products is reduced. 27 In the long run, prices turn into strategic substitutes:<br />
the reduction in the price <strong>of</strong> the leader induces the followers to increase their<br />
prices. 28<br />
Finally, we hope that these simple models <strong>of</strong> market leadership could be<br />
useful for normative purposes. Underst<strong>and</strong>ing the way markets work under<br />
different entry conditions is important not only to derive policy implications<br />
for competition policy, a topic on which we will turn in Section 3.5 <strong>and</strong> in<br />
Chapter 5, but also to underst<strong>and</strong> how government policy should deal with a<br />
number <strong>of</strong> issues concerning foreign markets <strong>and</strong> domestic ones, a hot topic<br />
in the days <strong>of</strong> intense globalization, on which we will focus in Sections 3.6<br />
<strong>and</strong> 3.7.<br />
27 Assume zero marginal costs. The optimality condition <strong>of</strong> the followers <strong>and</strong> the<br />
endogenous entry condition imply the following equilibrium relation between the<br />
price <strong>of</strong> the followers p <strong>and</strong> the number <strong>of</strong> firms n:<br />
<br />
F (1 − b)[1 + b(n − 1)]<br />
p =<br />
[1 + b(n − 2)<br />
A reduction in the price <strong>of</strong> the leader p L reduces entry <strong>and</strong>, according to this<br />
relation, it increases the price <strong>of</strong> the followers. The pr<strong>of</strong>it <strong>of</strong>theleaderis:<br />
<br />
<br />
p L<br />
b(n − 1)<br />
π L =<br />
a − p L +<br />
1+b(n − 1)<br />
1 − b (p − pL) − F<br />
where both n <strong>and</strong> p depend on p L .Since∂π L /∂nb pL =p< 0, itisoptimalfor<br />
the leader to reduce the number <strong>of</strong> firms compared to the Marshall equilibrium.<br />
This implies a lower price <strong>of</strong> the leader <strong>and</strong> a higher price <strong>of</strong> the followers in<br />
the Stackelberg equilibrium with endogenous entry compared to the price <strong>of</strong> the<br />
Marshall equilibrium.<br />
28 These result derive from joint work with Nisvan Erkal <strong>and</strong> Daniel Piccinin. Notice<br />
that with a Shubik dem<strong>and</strong> a leader facing endogenous entry would reduce its<br />
price <strong>and</strong> the followers would reduce their prices as well (prices are strategic<br />
complements in both the short <strong>and</strong> long run). As a consequence the number <strong>of</strong><br />
varieties provided in the market would decrease. Nevertheless, consumer surplus<br />
would strictly increase because <strong>of</strong> the generalized reduction in prices.
118 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
3.5 <strong>Antitrust</strong> <strong>and</strong> Collusion<br />
Our analysis <strong>of</strong> the behavior <strong>of</strong> a market leader <strong>and</strong> <strong>of</strong> multiple market leaders<br />
in this <strong>and</strong> in the previous chapter has been useful to introduce our discussions<br />
<strong>of</strong> antitrust issues concerning abuse <strong>of</strong> dominance. Nevertheless, the same<br />
principles can be exploited to investigate other antitrust issues as well. In<br />
this section we will focus on price fixing cartels.<br />
One <strong>of</strong> the main objectives <strong>of</strong> antitrust policy is the elimination <strong>of</strong> forms <strong>of</strong><br />
collusion between firms aimed at increasing prices. As we have seen in Chapter<br />
1, a collusive cartel for the choice <strong>of</strong> prices or quantities between an exogenous<br />
number <strong>of</strong> firms ends up increasing prices <strong>and</strong> harming consumers. When a<br />
restricted number <strong>of</strong> firms collude, they can still implement accommodating<br />
strategies <strong>and</strong> increase their equilibrium prices <strong>and</strong> pr<strong>of</strong>its (especially if they<br />
act as leaders). The reaction <strong>of</strong> the other firms to their collusive strategies can<br />
be either aggressive under SS or accommodating under SC, but the outcome<br />
is qualitatively similar to the previous one: when it takes place, collusion in<br />
a market with an exogenous number <strong>of</strong> firms tends to harm consumers. This<br />
book does not have much to add to this important principle. In this section<br />
we will examine a different, but related, issue: the impact <strong>of</strong> collusion between<br />
arestrictednumber<strong>of</strong>firms in a market where entry is endogenous. In such<br />
a case, collusion has unusual effects.<br />
More formally, let us consider a collusive cartel between m firms, where<br />
their strategies x k for k =1, 2, ..., m, are chosen to maximize the joint pr<strong>of</strong>its:<br />
mX<br />
π Cartel = Π(x k ,β k ) − mF (3.25)<br />
k=1<br />
while the other firms i = m +1, ..., n, maximize their simple pr<strong>of</strong>its π i =<br />
Π(x i ,β i ) − F <strong>and</strong> enter until these net pr<strong>of</strong>its are zero.<br />
In a hypothetical Nash equilibrium between the cartel <strong>and</strong> the outsider<br />
firms, each member <strong>of</strong> the cartel would implement an accommodating strategy<br />
according to the joint optimality conditions:<br />
Π 1 (x k ,β k )+<br />
mX<br />
q=1,q6=k<br />
Π 2 (x q ,β q )h 0 (x k )=0 for k =1, 2, ..., m (3.26)<br />
while the outsiders would stick to the usual optimality conditions Π 1 (x i ,β i )=<br />
0. Notice that the accommodating strategies <strong>of</strong> the members <strong>of</strong> the cartel<br />
would attract entry until the cartel becomes a lossmaker: in Marshall equilibrium,<br />
a simple commitment to collusion is not pr<strong>of</strong>itable when entry is<br />
endogenous (this is another application <strong>of</strong> our results in Chapter 2, since the<br />
collusive commitment makes the members <strong>of</strong> the cartel more accommodating).<br />
However, a commitment to join in a cartel can be pr<strong>of</strong>itable when the<br />
members <strong>of</strong> the cartel act as leaders in the competition with the other firms.
3.5 <strong>Antitrust</strong> <strong>and</strong> Collusion 119<br />
More formally, consider a game in which the cartel plays first, then the followers<br />
enter, <strong>and</strong> finally the followers play simultaneously. In this case, the<br />
optimality condition <strong>of</strong> the followers <strong>and</strong> their zero pr<strong>of</strong>it condition pin down<br />
their strategy x <strong>and</strong> their spillovers β independently from the strategies <strong>of</strong><br />
the cartel. 29 Therefore, taking into account that the expected spillover <strong>of</strong> a<br />
member <strong>of</strong> the cartel is β k = P j6=k h(x j)=β + h(x) − h(x k ),theoptimal<br />
strategies <strong>of</strong> the cartel solve the problem:<br />
max<br />
x 1,...,x m<br />
π Cartel =<br />
mX<br />
Π [x k ,β+ h(x) − h(x k )] − mF (3.27)<br />
k=1<br />
The corresponding optimality conditions are:<br />
Π 1 (x k ,β k )=Π 2 (x k ,β k )h 0 (x k ) for k =1, 2, ..., m (3.28)<br />
But these conditions exactly correspond to the condition defining the equilibrium<br />
strategy <strong>of</strong> a leader (or more leaders) in the Stackelberg equilibrium with<br />
endogenous entry, namely (3.12). On this basis, we can apply all the results<br />
derived in the rest <strong>of</strong> this chapter. In the case <strong>of</strong> competition in quantities, a<br />
collusive cartel in a market where entry is endogenous would coordinate an<br />
increase in the output <strong>of</strong> its members so as to increase their market shares<br />
<strong>and</strong> improve the allocation <strong>of</strong> resources. In the case <strong>of</strong> competition in prices,<br />
the cartel would coordinate a reduction <strong>of</strong> the prices <strong>of</strong> its members to increase<br />
their market shares, <strong>and</strong> this would lead to an improvement in the<br />
allocation <strong>of</strong> resources. 30 We can summarize our result as follows:<br />
Proposition 3.10. In a market with endogenous entry, a collusive<br />
cartel is not effective unless it acts as a leader: in such a case, as<br />
long as there is endogenous entry <strong>of</strong> some followers, each member<br />
<strong>of</strong> the cartel is aggressive compared to each follower.<br />
Paradoxically, collusion by cartels acting as leaders in markets where entry<br />
is endogenous turns out to be pr<strong>of</strong>itable, sustainable 31 <strong>and</strong> also procompetitive.<br />
This result should not be overemphasized from a policy point <strong>of</strong><br />
view. It suggests that harmful collusion between a restricted number <strong>of</strong> firms<br />
<strong>of</strong> a market cannot occur when there is endogenous entry <strong>of</strong> other firms in the<br />
market - as already pointed out within the Chicago view (Bork, 1993, Posner,<br />
2001). However, most <strong>of</strong> the time, collusive cartels involve all the firms <strong>of</strong> an<br />
oligopolistic market <strong>and</strong> are harmful to consumers: their avoidance should be<br />
the main focus <strong>of</strong> antitrust authorities.<br />
29 We focus on the case in which the number <strong>of</strong> members <strong>of</strong> the cartel is small <strong>and</strong><br />
entry takes place in equilibrium. If this is not the case, the cartel deters entry.<br />
30 Under competition for the market an R&D cartel acting as a leader under endogenous<br />
entry would enhance investments in R&D for its members.<br />
31 Since the cartel with m members implements the same strategies as in the Stackelberg<br />
equilibrium with m leaders <strong>and</strong> endogenous entry, collusion is always sustainable.
120 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
3.6 State-Aids <strong>and</strong> Strategic Export Promotion<br />
Globalization leads to the intensification <strong>of</strong> competition on international markets<br />
<strong>and</strong> requires a deeper underst<strong>and</strong>ing <strong>of</strong> the effects <strong>of</strong> industrial policy<br />
at large in the international environment. In this section we will present a<br />
digression on the optimal state aid policy for exporting firms with particular<br />
reference to subsidies for exports, a topic on which there are contrasting<br />
views at both the policy <strong>and</strong> theoretical level.<br />
In the EU there are strong limitations to state aids distorting competition<br />
<strong>and</strong> affecting trade among member countries. Nevertheless, the EU<br />
heavily subsidizes exports <strong>of</strong> agricultural products <strong>and</strong> the aircraft industry<br />
(Airbus), France has a long tradition in supporting its “national champions”<br />
with public funding, Italy in supporting the Made in Italy. TheUShaveimplemented<br />
strong forms <strong>of</strong> export subsidization through tax exemptions for<br />
a fraction <strong>of</strong> export pr<strong>of</strong>its, foreign tax credits, export credit subsidies <strong>and</strong><br />
even an exemption from antitrust law for export cartels (the Webb-Pomerene<br />
Act exempts export associations from antitrust investigations as long as their<br />
actions do not restrain trade in the US <strong>and</strong> do not restrain the export trade<br />
<strong>of</strong> other domestic competitors). Nevertheless, at least in theory, the WTO<br />
forbids direct forms <strong>of</strong> export subsidization for industrial production.<br />
In front <strong>of</strong> such a complex <strong>and</strong> contradictory scenario, it is important to<br />
underst<strong>and</strong> whether state aids to exporting firms <strong>and</strong> export subsidies are<br />
beneficial (as unilateral policies) <strong>and</strong> what are their consequences for international<br />
markets. Economic theory is largely ambiguous on this point. In<br />
the neoclassical trade theory with perfect competition, for instance, export<br />
subsidies are not optimal because they deteriorate the terms <strong>of</strong> trade; more<br />
precisely, since export taxes are equivalent to import tariffs, their optimal<br />
value can be derived as 1/, where is the elasticity <strong>of</strong> dem<strong>and</strong> (see Helpman<br />
<strong>and</strong> Krugman, 1989). In case <strong>of</strong> imperfect competition, export promotion assumes<br />
a strategic dimension, so its main aim becomes shifting pr<strong>of</strong>its toward<br />
the domestic firms. A large body <strong>of</strong> literature has studied oligopolistic models<br />
with a fixed number <strong>of</strong> firms competing in a third market. In this case,<br />
the optimal unilateral policy is an export tax under price competition (or<br />
whenever SC holds; see Eaton <strong>and</strong> Grossman, 1986). Under quantity competition,<br />
an export subsidy can be optimal (Spencer <strong>and</strong> Br<strong>and</strong>er, 1983), but<br />
only under restrictive conditions. The ambiguity <strong>of</strong> these results represents a<br />
major problem to derive policy implications. 32<br />
When entry in the international market is free, however, the theory <strong>of</strong><br />
market leaders suggests that only a commitment able to turn the strategy <strong>of</strong><br />
the domestic firm into a more aggressive one is going to increase its pr<strong>of</strong>its.<br />
More precisely we can apply Prop. 2.3 <strong>and</strong> conclude that it is (unilaterally)<br />
optimal to implement any form <strong>of</strong> strategic export promotion that increases<br />
32 See Maggi (1996) for an important contribution which endogenizes the mode <strong>of</strong><br />
competition in the strategic trade literature.
3.6 State-Aids <strong>and</strong> Strategic Export Promotion 121<br />
the marginal pr<strong>of</strong>itability <strong>of</strong> the domestic firms: this may include direct or<br />
indirect state aids for exporting firms, policies that boost dem<strong>and</strong> or decrease<br />
transport costs, export subsidies, R&D subsidies for exporting firms or related<br />
strengthening <strong>of</strong> their IPRs (Etro, 2007,a). Here we will focus our attention<br />
on the optimal export subsidies following Etro (2006,f).<br />
To fix ideas with an example, imagine Harley & Davidson, Ducati <strong>and</strong><br />
Honda selling their motorbikes in a third unrelated market, say Australia.<br />
Consider the optimal unilateral policy <strong>of</strong> the US government toward H&D.<br />
According to the traditional view, the US government should tax exports<br />
<strong>of</strong> H&D. This would force H&D to increase its prices in Australia, which<br />
would lead Honda to increase its prices as well, <strong>and</strong> would generate higher<br />
American net pr<strong>of</strong>its from sales <strong>of</strong> H&D bikes in Australia, together with<br />
a tax revenue to be distributed between American citizens. The fallacy <strong>of</strong><br />
this argument relies on neglecting that other international companies, say<br />
Yamaha, Suzuki, Kawasaki, BMW or Aprilia, would be ready to enter in the<br />
Australian market for motorbikes whenever prices are high enough to promise<br />
positive pr<strong>of</strong>its. And when this is the case an export tax can only penalize<br />
H&D. When entry in the Australian motorbike market is endogenous, as we<br />
actually could expect, the optimal US trade policy is to subsidize Harley’s<br />
exports. Always.<br />
More formally, adopting the usual notation, it is immediate to verify that a<br />
(specific) export subsidy s increases the marginal pr<strong>of</strong>itability <strong>of</strong> the domestic<br />
firm, say firm H. For instance, under competition in quantities we have:<br />
Π(x H ,β H ,s)=[p(x H ,β H )+s] x H − c(x H ) (3.29)<br />
<strong>and</strong> Π 13 =1, while under competition in prices we have:<br />
Π(x H ,β H ,s)=(p H + s − c) D (p H ,β H ) with x H =1/p H (3.30)<br />
<strong>and</strong> Π 13 = −D 1 p 2 H > 0. Now, the optimal unilateral policy does not maximize<br />
the total pr<strong>of</strong>its <strong>of</strong> the domestic firm, but these pr<strong>of</strong>its net <strong>of</strong> the subsidy<br />
(notice that prices affect only foreign consumers). Therefore, the optimal<br />
policy must simply maximize the strategic impact on the domestic pr<strong>of</strong>its: it<br />
follows that, as long as entry in the international market is free, an export<br />
subsidy is always optimal.<br />
We can say something more than this: the optimal policy must implement<br />
nothing else than the Stackelberg equilibrium with endogenous entry in which<br />
the domestic firm is the leader, exactly the kind <strong>of</strong> equilibrium we have characterized<br />
in this chapter. Why this equilibrium? Simply because it is the best<br />
equilibrium that the domestic firm can aim for. It is now relatively simple<br />
to derive the subsidies that implement this equilibrium. For instance, with<br />
homogenous goods, increasing marginal costs <strong>and</strong> competition in quantities,<br />
the general expression for the optimal specific subsidy is (Etro, 2006,f): 33<br />
33 One can verify that the first order condition for the domestic subsidized firm:
122 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
s ∗ = p H<br />
> 0 (3.31)<br />
<br />
where p H is the equilibrium price <strong>of</strong> the domestic firm <strong>and</strong> = − (p H /x H )<br />
(dx H /dp H ) the corresponding elasticity <strong>of</strong> dem<strong>and</strong>. It is important to notice<br />
that this optimal subsidy rate is exactly the opposite <strong>of</strong> the optimal export<br />
tax rate in the neoclassical theory <strong>of</strong> trade policy.<br />
We can also derive the optimal specific subsidy under price competition.<br />
In our framework this is given by (Etro, 2006,f): 34<br />
s ∗ = (p H − c)D 2 (p H ,β H ) g 0 (p H )<br />
> 0 (3.32)<br />
[−D 1 (p H ,β H )]<br />
It is important to notice that the traditional optimal policy in the same model<br />
with exogenous entry would have required, according to the result <strong>of</strong> Eaton<br />
<strong>and</strong> Grossman (1986), a negative subsidy, that is an export tax.<br />
At this point, the intuition for the general optimality <strong>of</strong> export promoting<br />
policies should be straightforward. While firms are playing some kind <strong>of</strong> Nash<br />
competition in the foreign market, a government can give a strategic advantage<br />
to its domestic firm with an appropriate trade policy. When entry is<br />
free, an incentive to be accommodating is always counterproductive, because<br />
it just promotes entry by other foreign firms <strong>and</strong> shifts pr<strong>of</strong>its away from the<br />
domestic firm. It is instead optimal to provide an incentive to be aggressive,<br />
to exp<strong>and</strong> production or (equivalently) reduce the price, since this behavior<br />
limits entry increasing the market share <strong>of</strong> the domestic firm. This is only<br />
possible by subsidizing exports. 35 Of course, we need to remind the reader<br />
that we are here referring to the optimal unilateral policy: as well known,<br />
s + p(X)+x H p 0 (X) =c 0 (x H )<br />
satisfies the equilibrium condition (3.16) when the subsidy is the one in the<br />
text. As it should be clear after the discussion in this chapter, in the case <strong>of</strong><br />
constant or decreasing marginal costs, the optimal subsidy must implement an<br />
entry deterrence equilibrium.<br />
34 Again, one can verify that the first order condition for the domestic firm:<br />
(p H − c + s)D 1 (p H ,β H )+D(p H ,β H )=0<br />
corresponds to the pricing rule <strong>of</strong> a Stackelberg leader facing endogenous entry<br />
(3.20) when the subsidy is the one in the text.<br />
35 For related investigations on strategic trade policy see Kováč <strong>and</strong>Žigić (2006)<br />
<strong>and</strong> Boone et al. (2006). The first work analyzes strategic trade policy in markets<br />
where leaders choose the quality <strong>of</strong> their products before the followers. The<br />
second work shows that when domestic firmsareleadersinthedomesticmarket<br />
<strong>and</strong> invest in cost reducing innovations, but the protection <strong>of</strong> intellectual property<br />
rights on these innovation is limited abroad, positive tariffs can enhance<br />
consumer welfare (see also Žigić, 1998, 2000). The reason is that tariffs induce<br />
market leaders to be aggressive toward foreign imitators, whose entry is limited.
3.7 Privatizations 123<br />
if all countries were going to implement their optimal unilateral policies, an<br />
inefficient equilibrium would emerge. This may explain why international coordination<br />
tends to limit export subsidies.<br />
If we interpret globalization as the opening up <strong>of</strong> new markets to international<br />
competition we can restate the main result as follows: in a globalized<br />
word, there are strong strategic incentives to conquer market shares abroad<br />
by promoting exports.<br />
3.7 Privatizations<br />
A final application to privatizations deserves some comments. Recent decades<br />
have witnessed a huge sequence <strong>of</strong> privatizations, especially in Western European<br />
countries <strong>and</strong> in former communist countries. In many cases, public<br />
enterprises active in traditional markets were the subject <strong>of</strong> privatizations<br />
<strong>and</strong> an intense debate emerged on the conditions under which private or<br />
public property was better (see Boycko et al., 1997). In this section, following<br />
an important early contribution by Anderson et al. (1997) we provide an<br />
alternative way to approach this debate.<br />
Broadly speaking, a public firm is characterized by a different objective<br />
function, which we can (generously) identify with the welfare function, <strong>and</strong><br />
by likely inefficiencies associated with the lack <strong>of</strong> an optimal allocation <strong>of</strong><br />
incentives within public institutions. If this is the case, we can evaluate the<br />
behavior <strong>of</strong> the same firm when public <strong>and</strong> when privatized. A crucial issue<br />
in this case, is whether a process <strong>of</strong> liberalization, meaning <strong>of</strong> opening to<br />
endogenous entry <strong>of</strong> other private firms, occurs as well.<br />
As a benchmark case, consider the production <strong>of</strong> a homogenous good. A<br />
single public firm would maximize welfare by pricing at the marginal cost. If<br />
the same public firm faces a process <strong>of</strong> liberalization with entry <strong>of</strong> pr<strong>of</strong>it maximizing<br />
firms, it is immediate that the Marshall equilibrium will correspond<br />
exactly to the one under Stackelberg competition with endogenous entry. In<br />
such a case, the public firm would still price at marginal cost, 36 <strong>and</strong> the private<br />
firms will apply a markup to cover their fixed costs <strong>of</strong> production: the<br />
pr<strong>of</strong>its <strong>of</strong> the public firm would be positive only if its cost inefficiency is limited<br />
relative to the private firms. Consider privatization now. If the privatized<br />
firm is symmetric with respect to the other firms, it will end up obtaining<br />
zero pr<strong>of</strong>its as the others. If the privatized firm gains the role <strong>of</strong> the leader,<br />
it can keep pricing at its marginal cost while obtaining positive pr<strong>of</strong>its: the<br />
privatization does not affect the equilibrium price, but it increases pr<strong>of</strong>its<br />
for the privatized company. Overall, the privatization enhances welfare when<br />
36 The objective function <strong>of</strong> the public firm corresponds to π Publ = X<br />
<br />
p(j)dj −<br />
0<br />
i6=P c(x i) − c Publ (x P ) − nF where X is total production <strong>and</strong> the cost function<br />
<strong>of</strong> the public firm c Publ (·) can be different from that <strong>of</strong> the private firms because<br />
<strong>of</strong> some inefficiencies. Maximization <strong>of</strong> this function leads to p(X) =c 0 Publ(x P ).
124 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
the former public enterprise becomes the leader <strong>of</strong> a market with endogenous<br />
entry.<br />
Two remarks are in order. First, if products are differentiated or firms<br />
compete in prices (see Anderson et al., 1997) the gains from privatization<br />
may be larger because product variety flourishes. Second, if the privatization<br />
is not associated with a process <strong>of</strong> liberalization, it may lead to ambiguous<br />
results: for instance a privatized firm may increase its prices <strong>and</strong> induce other<br />
private firms to do the same. This cannot happen when entry is endogenous:<br />
liberalization is crucial to gain from privatizations.<br />
3.8 Conclusions<br />
In this chapter we analyzed different forms <strong>of</strong> competition in the market<br />
where leaders can exploit a strategic advantage to increase their pr<strong>of</strong>its. We<br />
noticed that their behavior depends on the entry conditions in a crucial way.<br />
The difference is quite evident under competition in prices. In markets where<br />
entry is limited exogenously leaders tend to behave in an accommodating<br />
way choosing high prices, which leads the followers to chose high prices as<br />
well. All firms obtain large pr<strong>of</strong>its but a second mover advantage emerges: the<br />
followers obtain larger pr<strong>of</strong>its than the leader. When entry is endogenous (<strong>and</strong><br />
determined by the opportunities to make pr<strong>of</strong>its in the market), new firms<br />
are attracted into the market from a similar accommodating strategy <strong>of</strong> both<br />
the leader <strong>and</strong> the followers. Since entry occurs until the net pr<strong>of</strong>its <strong>of</strong> the<br />
followers are driven to zero, the accommodating leader ends up with negative<br />
pr<strong>of</strong>its because <strong>of</strong> the second mover advantage (its pr<strong>of</strong>its must be lower than<br />
the pr<strong>of</strong>its <strong>of</strong> the followers, which in turn have been entirely dissipated by free<br />
entry). This implies that a leader can only gain from an aggressive pricing<br />
strategy: in equilibrium, the price <strong>of</strong> the leader is lower than the price <strong>of</strong> the<br />
followers <strong>and</strong> the first mover advantage is restored.<br />
With this chapter we have concluded our excursus on the different modes<br />
<strong>of</strong> competition in the market (in the choice <strong>of</strong> output or price variables). All<br />
sectors have such a component <strong>of</strong> competition. Nevertheless, in some sectors<br />
such a conponent plays a minor role in the interaction between firms <strong>and</strong><br />
in the entry process: these are sectors in which competition is mainly for<br />
the market <strong>and</strong> entry <strong>of</strong> new products or new firms derives from successful<br />
innovations. These sectors are the subject <strong>of</strong> the next chapter.
3.9 Appendix 125<br />
3.9 Appendix<br />
Pro<strong>of</strong><strong>of</strong>Prop3.2: The system (3.7)-(3.8) defines the impact on x <strong>and</strong> n<br />
to changes in x L . Totally differentiating the system we have:<br />
⎡<br />
⎣ dx ⎤ ⎡<br />
⎤ ⎡<br />
⎦ = − 1 Π 2 h(x)<br />
−Π 12 h(x)<br />
⎣<br />
⎦ ⎣ Π ⎤<br />
12h 0 (x L )dx L<br />
⎦<br />
∆<br />
dn −(n − 2)Π 2 h 0 (x) Π 11 +(n − 2)Π 12 h 0 (x) Π 2 h 0 (x L )dx L<br />
where ∆ = Π 11 Π 2 h(x) <strong>and</strong> Π 11 +(n − 2)Π 12 h 0 (x)+Π 2 h(x) < 0 (under the<br />
contraction condition in case <strong>of</strong> SC), which implies stability. It follows that:<br />
dx<br />
dx L<br />
=0<br />
dn<br />
= −h0 (s)<br />
< 0<br />
dx L h(x)<br />
dβ<br />
dx L<br />
=0<br />
dβ L<br />
dx L<br />
= −h 0 (x L ) < 0<br />
which shows that the strategy <strong>of</strong> the followers is independent from the one<br />
<strong>of</strong> the leader. Since this holds also for x L = x, which replicates the Marshall<br />
equilibrium, in a Stackelberg equilbrium with endogenous entry any active<br />
follower adopts the same strategy as in the Marshall equilibrium.<br />
At the entry stage, entry <strong>of</strong> at least one follower takes place for any<br />
x L < ¯x L ,where¯x L is such that:<br />
Π [x(h(¯x L )),h(¯x L )] = F<br />
<strong>and</strong> the pr<strong>of</strong>it <strong>of</strong> the leader is:<br />
π L = ΠL [x L , (n − 1)h(x)] − Fifx L < ¯x L<br />
Π L (x L , 0) − F if x L > ¯x L<br />
hence, the optimal strategy is x ∗ L that satisfies the first order condition:<br />
Π L 1 [x ∗ L, (n − 1)h(x)] = Π L 2 [x ∗ L, (n − 1)h(x)] h 0 (x ∗ L)<br />
if it is smaller than ¯x L <strong>and</strong>suchthat:<br />
Π L {x ∗ L, (n − 1) h(x)} >Π L (¯x L , 0)<br />
Otherwise the global optimum is the corner solution ¯x L . We will show that in<br />
equilibrium x L >xalways. In case <strong>of</strong> corner solution, this is trivial. Consider<br />
the case <strong>of</strong> an interior solution x ∗ L as defined above. Assume that x∗ L ≤ x;then<br />
it must be that β =(n − 2)h(x)+h(x ∗ L ) ≤ (n − 1)h(x) =β L, which implies<br />
Π(x ∗ L ,β L) ≤ Π(x ∗ L ,β) from the assumption Π 2 < 0. But the optimality <strong>of</strong><br />
x <strong>and</strong> the free entry condition imply Π(x ∗ L ,β)
126 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
Pro<strong>of</strong> <strong>of</strong> Proposition 3.3. Theeffect <strong>of</strong> a change in the fixed cost on<br />
the strategy <strong>and</strong> the number <strong>of</strong> firms are:<br />
dx<br />
dF = [−Π 12]<br />
[Π 11 Π 2 ]<br />
∙<br />
dn<br />
dF = Π11 +(n − 2)Π 12 h 0 ¸<br />
(x)<br />
+ ∂n ∂x L<br />
Π 11 Π 2 h(x) ∂x L ∂F<br />
The first derivative has the opposite sign <strong>of</strong> Π 12 . The second has a first negative<br />
term (under the contraction condition when Π 12 > 0) <strong>and</strong> a second ambiguous<br />
term. It follows that d[β + h(x)]/dF =[Π 11 − h 0 (x)Π 12 ]/Π 11 Π 2 h(x).<br />
Totally differentiating (3.12) we have:<br />
£ ¤<br />
∂x L Π<br />
L<br />
∂F = − 12 − h 0 (x L )Π22<br />
L [Π11 − Π 12 h 0 (x)]<br />
D L Π 11 Π 2 h(x)<br />
where D L < 0 from the assumption that the second order condition is satisfied.<br />
It follows that:<br />
∙<br />
dn<br />
dF ∝ Π 11 +(n − 2)Π 12 h 0 (x)+ h0 (x L )<br />
h(x)D L [Π 11 − Π 12 h 0 (x)] £ ¤¸<br />
Π12 L − h 0 (x L )Π22<br />
L<br />
The Proposition follows immediately after noticing from (3.12) that h 0 (x L )=<br />
Π L 1 /Π L 2 . Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.4: In a Marshall equilibrium, the number <strong>of</strong> firms is<br />
n m <strong>and</strong> each one produces x m ,withwelfare:<br />
W m =<br />
Z<br />
n m x m<br />
0<br />
p(j)dj − n m [c(x m )+F ]=<br />
Z<br />
n m x m<br />
0<br />
p(j)dj − p(n m x m )n m x m<br />
where we used the zero pr<strong>of</strong>it condition p(n m x m )x m = c(x m )+F . Under<br />
Stackelberg competition when there is endogenous entry by some followers,<br />
the strategy <strong>of</strong> each follower remains x m byProp.3.2,whilethenumber<strong>of</strong><br />
firms n s satisfies the zero pr<strong>of</strong>it condition:<br />
p [x L +(n s − 1)x m ] x m = c(x m )+F<br />
which implies the same total production in the two cases x L +(n s − 1)x m =<br />
n m x m . Hence the welfare will be:<br />
W s =<br />
=<br />
Z<br />
x L+(n s −1)x m<br />
Z<br />
0<br />
n m x m<br />
0<br />
p(j)dj − (n m − 1)c(x m ) − c(x L ) − n s F =<br />
p(j)dj − p(n m x m )n m x m +[x L +(n s − 1)x m ] p(n m x m )<br />
−(n s − 1)c(x m ) − c(x L ) − n s F<br />
= W m + x L p [x L +(n s − 1)x m ] − c(x L ) − F = W m + π L >W m
3.9 Appendix 127<br />
which proves the claim. Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.5: Adopt a generic cost function c(x) with c 00 (x) ≤ 0.<br />
Imagine an equilibrium without entry deterrence. The zero pr<strong>of</strong>it condition,<br />
stated in the pro<strong>of</strong> <strong>of</strong> Prop. 3.4, sets total production <strong>and</strong> hence the inverse<br />
dem<strong>and</strong> at the level:<br />
p[x m (n s − 1) + x L ]= F + c(xm )<br />
x m<br />
where x m is always the equilibrium production <strong>of</strong> the followers, which corresponds<br />
to the equilibrium production in the Marshall equilibrium. Then, the<br />
pr<strong>of</strong>it function <strong>of</strong> the leader becomes:<br />
with:<br />
Π L (x L )=x L p[x m (n s − 1) + x L ] − c(x L )=x L<br />
∙ F + c(x m )<br />
x m<br />
¸<br />
− c(x L )<br />
Π L0 (x L )= F + c(xm )<br />
x m − c 0 (x L ) > 0 Π L00 (x L )=−c 00 (x L ) ≥ 0<br />
since p(·) >c 0 (x m ) >c 0 (x L ) for any x L >x m . Hence, the leader always<br />
gains from increasing its production all the way to the level at which entry<br />
is deterred. This level satisfies the zero pr<strong>of</strong>it condition for n s =2,thatis<br />
p (x m +¯x L )=[F + c(x m )] /x m . Since the right h<strong>and</strong> side is also equal to<br />
p(n m x m ) by the zero pr<strong>of</strong>it condition in the Marshall equilibrium (see the<br />
pro<strong>of</strong> <strong>of</strong> Prop. 3.4), it follows that the entry deterrence strategy is exactly<br />
¯x L =(n m − 1)x m . Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.6: Total expenditure Ȳ for the representative agent is<br />
given by an exogenous part Y <strong>and</strong> the net pr<strong>of</strong>its <strong>of</strong> the firms P n<br />
i=1 π i,which<br />
is zero in the Marshall equilibria, but equal to the positive pr<strong>of</strong>its <strong>of</strong> the<br />
leader π L in the Stackelberg equilibrium with endogenous entry. The welfare<br />
comparison derives from the calculation <strong>of</strong> indirect utilities (2.22) for the<br />
Logit model <strong>and</strong> (2.23) for the Dixit-Stiglitz model in both cases. Labeling<br />
with W (Ȳ ) the indirect utility in function <strong>of</strong> total expenditure Ȳ , in the Logit<br />
case we have for both equilibria:<br />
W (Ȳ )=Ȳ + N µ<br />
λ ln 1+ N <br />
− N(1 + λc) − λF<br />
λF<br />
<strong>and</strong> in the Dixit-Stiglitz case we also have for both equilibria:<br />
W (Ȳ )= θ ¡ αȲ ¢ α £Ȳ<br />
− F (1 + α)<br />
¤<br />
c (1 + α) 1+α<br />
∙ (1 − θ) Ȳ<br />
(1 + α)F<br />
1−θ<br />
θ<br />
+ θ¸
128 3. Stackelberg <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
Since they are both increasing in total expenditure, the utility <strong>of</strong> the representative<br />
agent must be higher under Stackelberg competition with endogenous<br />
entry. Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.7: The analysis <strong>of</strong> the last stage is the same as before,<br />
<strong>and</strong>inparticulardx/dx L =0.Now,theleader’sfirst order condition becomes:<br />
Π L 1 [x L,β+ h(x) − h(x L ),k]=Π L 2 [x L,β+ h(x) − h(x L ),k] h 0 (x L )<br />
which defines a continuous function x L = x L (k). It follows that:<br />
x 0 L(k) ∝ Π L 13 [x L ,β+ h(x) − h(x L ),k]−Π L 23 [x L ,β+ h(x) − h(x L ),k] h 0 (x L )<br />
Clearly, when the condition in the proposition holds x 0 L (k) ≥ 0 <strong>and</strong> x L(k) ≥<br />
x L (0) >xbyProp.3.2.Otherwise,sincex L (0) >x, continuity implies that<br />
there is a neighborhood <strong>of</strong> x L (0) for k small enough where x L (0) >x L (k) >x.<br />
Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.8: The analysis is similar to the basic one, but now<br />
we have:<br />
with:<br />
⎡<br />
⎣ dx ⎤ ⎡<br />
⎦ = − 1<br />
dn<br />
∆ Ω ⎣ Π ⎤<br />
12h 0 (x L )dx L + [h(x L ) − h(x)]Π 12 dm<br />
⎦<br />
Π 2 h 0 (x L )dx L + [h(x L ) − h(x)]Π 2 dm<br />
⎡<br />
Π 2 h(x)<br />
−Π 12 h(x)<br />
⎤<br />
Ω ≡ ⎣<br />
⎦<br />
−(n − m − 1)Π 2 h 0 (x) Π 11 +(n − m − 1)Π 12 h 0 (x)<br />
which again implies dx/dx L =0<strong>and</strong> dn/dx L = −h 0 (x L )/h 0 (x). Moreoverwe<br />
have:<br />
dx dn<br />
dm<br />
=0,<br />
dm =1− h(x L)<br />
h(x)<br />
< 0<br />
The first order conditions for each one <strong>of</strong> the leaders become:<br />
Π L 1 (x L ,β L )=Π L 2 (x L ,β L ) h 0 (x L )<br />
where β L =(n−m)h(x)+(m−1)h(x L ). Totally differentiating this condition<br />
<strong>and</strong> using dn/dm it follows that dx L /dm =0.Thepr<strong>of</strong>it <strong>of</strong>eachleaderis<br />
not affected by the number <strong>of</strong> leaders since:<br />
dπ L<br />
dm = ΠL 2<br />
∙<br />
h(x L ) − h(x)+h(x) dn<br />
dm<br />
¸<br />
=0
3.9 Appendix 129<br />
which concludes the pro<strong>of</strong>. Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop3.9:Denotewithx i =[x i1 ,x i2 , ..., x iK ] the strategies <strong>of</strong><br />
a firm i. Assume again that a symmetric equilibrium in the strategies <strong>of</strong> the<br />
followers exist. The system <strong>of</strong> K +1 equilibrium conditions for the second<br />
stage:<br />
∂Π [x, (n − 2)h(x)+h(x L )]<br />
=0 for k =1, 2,..,K<br />
∂x k<br />
Π [x, (n − 2)h(x)+h(x L )] = F<br />
pins down the vector x <strong>and</strong> β =(n−2)h(x)+h(x L ). Consequently the pr<strong>of</strong>it<br />
<strong>of</strong> the leader is:<br />
π L = Π L [x L , (n − 1)h(x)] − F = Π L [x L ,β+ h(x) − h(x L )] − F<br />
which is maximized by the vector x L which satisfies the system <strong>of</strong> K first<br />
order conditions:<br />
∂Π L (x L ,β L )<br />
∂x Lk<br />
= ∂h(x L)<br />
∂x Lk<br />
∂Π L (x L ,β L )<br />
∂β L<br />
where clearly β L =(n − 1)h(x). Imagine that there is such an interior equilibrium<br />
with h(x L ) ≤ h(x). Then it must be that β ≤ β L , which implies<br />
Π(x L ,β L ) ≤ Π(x L ,β) from the assumption ∂Π/∂β < 0. But the optimality<br />
<strong>of</strong> x <strong>and</strong> the free entry condition imply Π(x L ,β)
4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous<br />
Entry<br />
Static analysis <strong>of</strong> market structures as those studied in the previous two<br />
chapters are not particularly relevant for fast-moving markets <strong>of</strong> high-tech<br />
<strong>and</strong> New Economy industries (computer hardware <strong>and</strong> s<strong>of</strong>tware, online businesses,<br />
mobile telephony <strong>and</strong> biotechnology). These industries are <strong>of</strong>ten characterized<br />
by massive R&D investments, strong reliance on IPRs <strong>and</strong> other<br />
intangible assets, network effects, high fixed sunk costs <strong>and</strong> low marginal<br />
costs. <strong>Competition</strong> in these markets is <strong>of</strong>ten dynamic in the sense that it<br />
takes place for the market in a winner-takes-all race. Leading firms in these<br />
markets might enjoy high market shares yet be subject to massive competitive<br />
pressure to constantly create better products at lower prices due to threats<br />
from innovative competitors <strong>and</strong> potential entrants. Companies that hold a<br />
significant share <strong>of</strong> the market at any given point <strong>of</strong> time may see this share<br />
decrease rapidly <strong>and</strong> significantly following the development <strong>and</strong> supply <strong>of</strong> a<br />
new <strong>and</strong> more attractive product by an actual or potential competitor (the<br />
launches <strong>of</strong> the iPod <strong>and</strong> the iPhone by Apple <strong>and</strong> their impact on the distribution<br />
<strong>of</strong> MP3 players <strong>and</strong> smart phones are good examples <strong>of</strong> such rapid<br />
<strong>and</strong> drastic market developments), or they may persist in their leading position<br />
thanks to heavy investments in R&D (think <strong>of</strong> Intel whose large <strong>and</strong><br />
increasing investments induced sequential innovations in the development <strong>of</strong><br />
chips). 1<br />
This chapter analyzes competition for the market through models where<br />
firms invest to obtain innovations <strong>and</strong> conquer a market. Since innovations<br />
<strong>of</strong>ten lead to patents or other forms <strong>of</strong> intellectual property rights that guarantee<br />
exploitation for a certain period, we <strong>of</strong>ten refer to this kind <strong>of</strong> competition<br />
as to a patent race. In Chapter 1 we analyzed a simple form <strong>of</strong><br />
competition for the market, but here we augment it with a number <strong>of</strong> realistic<br />
additions: we introduce a time dimension, so that firms discount pr<strong>of</strong>its<br />
from future innovations, we allow for explicit forms <strong>of</strong> dynamic investment,<br />
we consider sequential innovations <strong>and</strong> endogenize expected pr<strong>of</strong>its in partial<br />
equilibrium, <strong>and</strong> finally we evaluate the impact <strong>of</strong> alternative forms <strong>of</strong><br />
product market competition on the incentives to invest in R&D.<br />
1 The innovation process <strong>of</strong> Intel has been so systematic that the rule <strong>of</strong> thumb<br />
for which the number <strong>of</strong> transistors on an Intel chip doubles every two years has<br />
been labeled Moore’s Law from the intuition <strong>of</strong> Intel co-founder Gordon Moore.
132 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
A central focus <strong>of</strong> this chapter will be on the role <strong>of</strong> incumbents in innovative<br />
sectors, <strong>and</strong> we will show under what conditions these firms invest in<br />
R&D <strong>and</strong> when their technological leadership persists. The first economist to<br />
forcefully emphasize the fundamental role <strong>of</strong> established large firms in driving<br />
technological progress has probably been Schumpeter:<br />
“As soon as we go into details <strong>and</strong> inquire into the individual<br />
items in which progress was most conspicuous, the trail leads not to<br />
the doors <strong>of</strong> those firms that work under conditions <strong>of</strong> comparatively<br />
free competition but precisely to the doors <strong>of</strong> the large concerns -<br />
which, as in the case <strong>of</strong> agricultural machinery, also account for much<br />
<strong>of</strong> the progress in the competitive sector - <strong>and</strong> a shocking suspicion<br />
dawns upon us that big business may have had more to do with creating<br />
that st<strong>and</strong>ard <strong>of</strong> life than with keeping it down” (Schumpeter,<br />
1943).<br />
Related analysis <strong>of</strong> modern capitalism as driven by the innovative <strong>and</strong><br />
persistent leadership <strong>of</strong> large firms is also in the classic works <strong>of</strong> Galbraith<br />
(1952) <strong>and</strong> Ch<strong>and</strong>ler (1990).<br />
Recent evidence confirms that incumbents do a lot <strong>of</strong> research <strong>and</strong> their<br />
leadership persists through a number <strong>of</strong> innovations. One <strong>of</strong> the industry<br />
leaders investing more in innovation is Micros<strong>of</strong>t, the leading firm in operating<br />
systems: in 2000, its expenditure in R&D was $ 3.7 billion, corresponding<br />
to 16.4% <strong>of</strong> its total sales. High investments can also be found in many other<br />
major firms <strong>of</strong> high tech sectors. In the same year, the R&D/Sales ratio was<br />
15% for Pfizer <strong>and</strong> 5.8% for Merck, two leaders in the pharmaceutical sector,<br />
11.5% for Intel, leader in the chips market <strong>and</strong> 5.8% for IBM, <strong>and</strong> 5.4% for<br />
Hewlett-Packard, two leaders in computer technologies <strong>and</strong> services, 11.8%<br />
for Motorola <strong>and</strong> 8.5% for Nokia, leaders in wireless, broadb<strong>and</strong> <strong>and</strong> automotive<br />
communications technologies, 10% for Johnson & Johnson, the world’s<br />
most comprehensive manufacturer <strong>of</strong> health care products <strong>and</strong> services, 6.6%<br />
for 3M <strong>and</strong> 6.3% for Du Pont, which are active in many fields with a leading<br />
role, 5.6% for Xerox <strong>and</strong> for Kodak, leaders in the markets for printers<br />
<strong>and</strong> photographs. Things did not change much since then. Today American<br />
corporations spend around $ 200 billion on R&D every year, much <strong>of</strong> it<br />
on computing <strong>and</strong> communications: in 2006 Micros<strong>of</strong>t spent around $ 6.6 billion,<br />
IBM <strong>and</strong> Intel about $6 billion each, Cisco Systems <strong>and</strong> Hewlett-Packard<br />
around $4 billion each (The Economist, 2007, “Out <strong>of</strong> the dusty labs”, March<br />
1).<br />
More systematic evidence on the R&D activity by market leaders comes<br />
from patented innovations <strong>and</strong> expenditure on licenses. The comprehensive<br />
study by Malerba <strong>and</strong> Orsenigo (1999) on EU patents provides clear evidence<br />
on this point. 2 For instance, they show that the percentage <strong>of</strong> patents granted<br />
to firms that had already innovated within their sectors is 70 % in Germany,<br />
2 See also Malerba et al. (1997).
4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry 133<br />
68 % in US, 62 % in Japan, 60 % in France, 57 % in UK <strong>and</strong> 39 % in Italy;<br />
moreover, they conclude that:<br />
“a large fraction <strong>of</strong> new innovators is composed by occasional innovators<br />
that exit soon from the innovative scene [...] Only a fraction<br />
<strong>of</strong> entrants survives <strong>and</strong> grows larger (in terms <strong>of</strong> patents) as times<br />
goes by: they become persistent innovators. Older firms who survive<br />
<strong>and</strong> continue to patent are few in number but represent an important<br />
contribution to total patenting activities in any period. Here,<br />
cumulativeness <strong>of</strong> knowledge <strong>and</strong> competencies play a major role in<br />
affecting the continuity <strong>of</strong> innovative activity <strong>of</strong> these firms.”<br />
Czarnitzki <strong>and</strong> Kraft (2007a) is the first study looking at who purchases<br />
licenses on patents: on the basis <strong>of</strong> German data they show that incumbents<br />
invest more in licensing expenditures than effective <strong>and</strong> potential entrants,<br />
<strong>and</strong> that the investment <strong>of</strong> these incumbents is higher when the entry threats<br />
are stronger. 3<br />
The literature on patent races has studied equilibrium outcomes in the<br />
market for innovations starting with Loury (1979) <strong>and</strong> Dasgupta <strong>and</strong> Stiglitz<br />
(1980). 4 The st<strong>and</strong>ard hypotheses <strong>of</strong> this literature are decreasing returns<br />
to scale, fixed costs <strong>of</strong> innovations <strong>and</strong> Nash competition between firms. The<br />
participants <strong>of</strong> the patent race are the current monopolists <strong>of</strong> the market, who<br />
have a patent on the leading-edge product, <strong>and</strong> a number <strong>of</strong> entrant firms<br />
trying to replace the patentholder. A main result is that the patentholder<br />
does less research than any other entrant <strong>and</strong> zero research under free entry<br />
because its incentives to invest in R&D are lower due to the Arrow (1962)<br />
effect: the expected gain <strong>of</strong> the patentholder is just the difference between<br />
expected pr<strong>of</strong>its obtained with the next technology <strong>and</strong> the current one, while<br />
the expected gain for each outsider is given by all the expected pr<strong>of</strong>its obtained<br />
with the next technology. In the presence <strong>of</strong> sequential innovations, the<br />
fact that patentholders do not invest in R&D implies a continuous leapfrogging<br />
<strong>and</strong> no persistence <strong>of</strong> monopolistic positions between one innovation <strong>and</strong><br />
another, which is a quite counterintuitive picture <strong>of</strong> what is going on in the<br />
real world: that’s why the result is sometimes called the Arrow’s paradox.<br />
A number <strong>of</strong> solutions to the Arrow paradox have been proposed, most <strong>of</strong><br />
which are based on some technological advantage <strong>of</strong> the patentholder, <strong>of</strong>ten<br />
derived from a gradual accumulation <strong>of</strong> knowledge. 5 Despite the fact that<br />
3 The empirical research on the reaction <strong>of</strong> the investment <strong>of</strong> incumbents to entry<br />
is limited. Scherer <strong>and</strong> Keun (1992) look at the increase in high-tech imports<br />
in US <strong>and</strong> find that incumbents in sectors without barriers to entry react more<br />
aggressively to endogenous entry, increasing R&D/sales more than other firms.<br />
4 See also Lee <strong>and</strong> Wilde (1980), Reinganum (1982, 1983, 1985a,b), Harris <strong>and</strong><br />
Vickers (1985) <strong>and</strong> Beath et al. (1989).<br />
5 For a survey, see Tirole (1988, Ch.10). See also Reinganum (1982), Fudenberg et<br />
al. (1983), Harris <strong>and</strong> Vickers (1985) <strong>and</strong> Vickers (1986) for detailed analysis.
134 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
these are reasonable explanations for the puzzle, they do not seem to tell<br />
the whole story, as we see monopolists investing in R&D even if they do not<br />
have consistent technological advantage to the outsiders. Here we will study<br />
patent races where the patentholder has the opportunity to make a strategic<br />
precommitment to a flow <strong>of</strong> investment in R&D. This may happen through a<br />
specific investment in laboratories <strong>and</strong> related equipment for R&D, by hiring<br />
researchers or in a number <strong>of</strong> other ways. In the case <strong>of</strong> “contractual costs”<br />
<strong>of</strong> R&D, that is, when a fixed initial investment determines the arrival rate<br />
<strong>of</strong> the innovation, the interpretation <strong>of</strong> a strategic precommitment for the incumbent<br />
monopolist is very st<strong>and</strong>ard. The leader can choose to invest before<br />
the other firms, <strong>and</strong> since the leader is by definition the firm that has discovered<br />
the latest technology, it is reasonable to assume that such a discovery is<br />
associated with a first mover advantage in the following patent race. When<br />
the number <strong>of</strong> entrants is exogenous, the behavior <strong>of</strong> the incumbent is hard<br />
to predict: on one side the Arrow effect pushes toward a low investment, on<br />
the other side the strategic effect is ambiguous. Under reasonable conditions,<br />
however, a first mover advantage does not give strong incentives to invest for<br />
an incumbent monopolist, <strong>and</strong> the traditional view that monopolists stifle<br />
innovation is preserved: without competitive pressure, monopolists are not<br />
very innovative indeed. 6<br />
Under endogenous entry, the outcome is completely changed <strong>and</strong> generates<br />
a crucial result: the incumbent leader always invests in R&D <strong>and</strong> more<br />
so than any other firm, thus the Stackelberg assumption with endogenous<br />
entry delivers a new rationale for the persistence <strong>of</strong> a monopoly (Etro, 2004).<br />
The rationale for endogenous innovation by leaders is similar to the general<br />
rationale for aggressive strategies by leaders facing endogenous entry: competitive<br />
pressure determines the aggregate rate <strong>of</strong> innovation <strong>and</strong> the investment<br />
<strong>of</strong> the leader cannot affect this or the expected length <strong>of</strong> the current rent.<br />
Since the expected pr<strong>of</strong>its from the current technology are not affected by<br />
the leader’s strategy, the Arrow effect disappears <strong>and</strong>, as we also know from<br />
our general analysis in Chapter 3, the optimal behavior for a Stackelberg<br />
leader facing endogenous entry is always aggressive. The empirical results <strong>of</strong><br />
Blundell et al. (1999) “are in line with models where high market share firms<br />
6 In the pre-industrial world, barriers to entry in the innovative sectors, monopolized<br />
for centuries by guilds, have represented a substantial limit to innovation.<br />
Dutch guilds opposed progress in shipbuilding, Swiss printers obtained laws to<br />
avoid improvements in printing press <strong>and</strong> French paper producers sabotaged<br />
machines that could speed up pulp production. Interesting historical evidence<br />
is described by Ogilvie (2004a,b) in a study on merchant guilds between the<br />
XVI <strong>and</strong> the XVIII century. These kinds <strong>of</strong> guilds, spread for centuries around<br />
Europe, were strongly restricting entry in many sectors <strong>and</strong> were detrimental to<br />
innovation activity. Finally, the English Luddites, organized in trade unions, had<br />
a similar role at the beginning <strong>of</strong> the Industrial Revolution.
4.1 A Simple Patent Race with Contractual Costs <strong>of</strong> R&D 135<br />
have greater incentives to preemptively innovate”. Their conclusion is rather<br />
explicit:<br />
“It is <strong>of</strong>ten asserted that the superior performance <strong>of</strong> large firms<br />
in innovating is because they have higher cash flows from which to<br />
finance investment in R&D. Our findings suggest that this is not the<br />
whole story - dominant firms innovate because they have a relatively<br />
greater incentive to do so. Firm with high market shares who innovate<br />
get a higher valuation on the stock market than those who do not.”<br />
However, notice that, contrary to purely-preemptive models <strong>of</strong> innovation,<br />
in our environment incumbents do not necessarily deter entry, but they typically<br />
invest more than other firms, so that their leadership is only partially<br />
persistent.<br />
When innovations are sequential, not only incumbent monopolists keep<br />
investing under the pressure <strong>of</strong> endogenous entry, but the same value <strong>of</strong> their<br />
leadership is enhanced, which in turn increases the aggregate incentives to<br />
invest in R&D. Contrary to a common belief for which monopolies would stifle<br />
innovation, the persistence <strong>of</strong> monopoly can be caused by innovative pressure<br />
<strong>and</strong> can enhance technological progress. This result appears in line with the<br />
original ideas <strong>of</strong> Schumpeter (1943) on the role <strong>of</strong> large established firms in<br />
fostering innovation, <strong>and</strong> we will use it to sketch a model <strong>of</strong> technological<br />
progress driven by market leaders.<br />
The chapter is organized as follows. Section 4.1 presents a simple model<br />
<strong>of</strong> patent races, Section 4.2 extends it in more realistic ways <strong>and</strong> Sections<br />
4.3 considers sequential innovations. Finally, Section 4.4. discusses the relation<br />
between competition in the market <strong>and</strong> competition for the market <strong>and</strong><br />
Section 4.5 concludes.<br />
4.1 A Simple Patent Race with Contractual Costs <strong>of</strong><br />
R&D<br />
In this chapter we will develop models <strong>of</strong> competition for the market. We<br />
already developed an example in Chapter 1, but in that case we assumed a<br />
very simple technology <strong>of</strong> investment in innovations. Investment could be just<br />
successful or not (<strong>and</strong> by investing enough a firm could even innovate with<br />
certainty), while in the real world it takes time <strong>and</strong> risk to innovate, <strong>and</strong> future<br />
gains are properly discounted taking into account alternative investment<br />
opportunities. In this section we will introduce a time dimension developing<br />
a simple patent race in which investment can only increase the chances <strong>of</strong><br />
innovating early on. Of course this is crucial in a competition where the first<br />
to innovate wins a patent <strong>and</strong> the associated pr<strong>of</strong>its, while all the others get<br />
nothing. Nevertheless, we still assume that an initial investment determines<br />
the future chances to innovate, therefore we are still dealing with a form <strong>of</strong>
136 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
competition which is partially static (in the next section we will augment the<br />
model with a genuinely dynamic investment).<br />
Following the pathbreaking contribution <strong>of</strong> Loury (1979) <strong>and</strong> Dasgupta<br />
<strong>and</strong> Stiglitz (1980) we will adopt a particular R&D technology, assuming<br />
that, given the investment choices <strong>of</strong> the firms, innovations arrive according<br />
to a stochastic Poisson process in the continuum. According to this process,<br />
the probability that a single firm i will obtain the innovation before a certain<br />
amount <strong>of</strong> time t ∈ [0, ∞) is independent across firms, memoryless <strong>and</strong> given<br />
by:<br />
G(t, i) =1− e −h it<br />
where h i is a firm specific parameter. Notice that this probability does not<br />
depend on the corresponding probability <strong>of</strong> other firms <strong>and</strong> does not depend<br />
on the probability <strong>of</strong> innovation <strong>of</strong> the same firm i before time t. The density<br />
function is g(t, i) =h i e −hit . Another property <strong>of</strong> a Poisson process is that<br />
the so-called hazard rate, the instantaneous probability <strong>of</strong> innovation in t<br />
conditioned to previous failure, corresponds to the firm specific parameter<br />
h i > 0. Indeed, we have:<br />
Pr(i innovates in t) =<br />
g(t, i)<br />
1 − G(t, i) = h i<br />
The simplest kind <strong>of</strong> investment we can consider is a fixed investments,<br />
usually called a contractual cost <strong>of</strong> innovation. In this case, at the beginning<br />
<strong>of</strong> the race, each firm i invests a fixed amount F to participate to the contest,<br />
<strong>and</strong> decides a variable amount, x i , so that the arrival rate <strong>of</strong> an innovation<br />
is:<br />
h i = h(x i ) with h(0) = 0, h 0 (x) > 0 <strong>and</strong> h 00 (x) R 0 for x S ˆx<br />
If we look at h(x) as to a stochastic production function <strong>of</strong> innovation, loosely<br />
speaking we are allowing for increasing returns to scale for low investment,<br />
but we assume decreasing returns for investment greater than a cut <strong>of</strong>f ˆx ≥ 0.<br />
Using basic properties <strong>of</strong> probability theory, we can calculate the probability<br />
that firm i winstheraceattimet as: 7<br />
Pr(i wins in t) =g(t, i) Y j6=i<br />
[1 − G(t, j)] = h i e − n<br />
j=1 hj<br />
The exogenous value <strong>of</strong> the innovation is V . In most <strong>of</strong> our discussion, for<br />
simplicity, we will refer to this as to the value <strong>of</strong> a patent. More generally,<br />
we may think <strong>of</strong> this as the expected value <strong>of</strong> the pr<strong>of</strong>its obtained by the<br />
innovation. For instance, the innovation could be kept secret <strong>and</strong> exploited<br />
7 Since we work in the continuum, the probability that two firms innovate at the<br />
same time is zero: there will always be a unique winner in these contests.
4.1 A Simple Patent Race with Contractual Costs <strong>of</strong> R&D 137<br />
until other innovations will replace it, or it could be disclosed with the innovator<br />
enjoying a first mover advantage in the marketing <strong>of</strong> the invention<br />
(even when facing free entry <strong>of</strong> imitators, as we have seen in the models <strong>of</strong><br />
the previous chapters). Nevertheless, it should be clear that a strengthening<br />
<strong>of</strong> the protection <strong>of</strong> IPRs will increase the value <strong>of</strong> the innovation V .Since<br />
we have introduced a time dimension, we need to take in consideration the<br />
present discounted value <strong>of</strong> the expected pr<strong>of</strong>its. Given the exogenous interest<br />
rate r, expectedpr<strong>of</strong>its from the patent race are:<br />
π i =<br />
Z ∞<br />
t=0<br />
e −rt V Pr(i wins in t)dt − x i − F =<br />
= h(x i)V<br />
r + p<br />
− x i − F<br />
wherewedefined with p = P n<br />
j=1 h(x j) the aggregate instantaneous probability<br />
<strong>of</strong> innovation. Also this pr<strong>of</strong>it function is nested in the general version<br />
(2.1) employed in the previous chapters. Rearranging <strong>and</strong> defining<br />
β i = P n<br />
k=1,k6=i h(x k), wehave:<br />
Π (x i ,β i )=<br />
h(x i )V<br />
r + h(x i )+β i<br />
− x i<br />
Here it can be verified that expected pr<strong>of</strong>its for firm i areaninvertedU<br />
function <strong>of</strong> the investment <strong>of</strong> the same firm x i <strong>and</strong> are decreasing in the investment<br />
<strong>of</strong> each other firm, since the relative probability <strong>of</strong> winning the race<br />
is what matters. However, in this case the cross derivative (Π 12 ) has an ambiguous<br />
sign. When another competitor invests more, the relative probability<br />
<strong>of</strong> winning is reduced, which makes a marginal investment less pr<strong>of</strong>itable,<br />
but at the same time the aggregate probability <strong>of</strong> innovation in the market<br />
is increased <strong>and</strong> this creates an effect in the opposite direction. If the first<br />
effect prevails R&D investments are strategic substitutes, as in our simple<br />
example <strong>of</strong> Chapter 1. In such a case, we would expect that a firm with a<br />
first mover advantage over a rival would invest more because <strong>of</strong> what we<br />
called the Stackelberg effect: a higher investment reduces the incentives <strong>of</strong><br />
the competitor to invest <strong>and</strong> increases the relative probability <strong>of</strong> winning the<br />
contest.<br />
We can also easily incorporate an asymmetric position for the incumbent<br />
monopolist. Assume that this monopolist has a flow <strong>of</strong> pr<strong>of</strong>its K from its<br />
own leading edge technology. Assume also that the innovation is drastic, so<br />
the incumbent obtains nothing in case <strong>of</strong> innovation by another firm: this<br />
characterizes a situation where the “winner takes all”. The expected pr<strong>of</strong>its<br />
<strong>of</strong> the monopolist are now:<br />
Π (x M ,β M ,K)=<br />
h(x M)V + K<br />
− x M<br />
r + h(x M )+β M<br />
What in Chapter 1 we called the Arrow’s effect is again at work: this<br />
effect tells us that current pr<strong>of</strong>its reduce the marginal pr<strong>of</strong>itability <strong>of</strong> R&D
138 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
investment (Π 13 < 0), <strong>and</strong> consequently they reduce the incentives <strong>of</strong> the<br />
incumbent monopolist to invest in innovation. Therefore, in any Nash equilibrium<br />
with an exogenous number <strong>of</strong> firms, the incumbent monopolist will<br />
invest less than any other firm. 8 The behavior <strong>of</strong> the incumbent monopolist<br />
acting as a leader in this model is complex because the Arrow effect <strong>and</strong> the<br />
Stackelberg effect may work in opposite directions. If the fixed costs <strong>of</strong> entry<br />
are high enough, an entry deterring strategy can be optimal, but when this<br />
is not the case, the optimal strategy for the incumbent monopolist will be<br />
biased toward a lower investment if the Arrow effect prevails.<br />
4.1.1 Endogenous Entry<br />
As shown in Etro (2004, 2008), any ambiguity <strong>of</strong> the results disappears in<br />
equilibria with endogenous entry. Consider first a Marshall equilibrium, where<br />
all firms compete in Nash strategies <strong>and</strong> entry takes place as long as there are<br />
pr<strong>of</strong>itable opportunities. In this environment, as we noticed, the incumbent<br />
monopolist is always investing less than the rivals because <strong>of</strong> the Arrow effect.<br />
When entry has dissipated all pr<strong>of</strong>itable opportunities for the other firms, the<br />
optimality condition for the outsiders <strong>and</strong> the free entry condition are:<br />
h 0 (x)V<br />
r + p<br />
= h0 (x)h(x)V<br />
(r + p) 2 +1,<br />
h(x)V<br />
r + p = x + F (4.1)<br />
These conditions determine the equilibrium investment <strong>of</strong> each outsider <strong>and</strong><br />
the aggregate probability <strong>of</strong> innovation independently from the equilibrium<br />
strategy <strong>of</strong> the monopolist. In particular, the investment <strong>of</strong> each outsider can<br />
be implicitly expressed as:<br />
h 0 (x)<br />
µ<br />
1 − x + F<br />
V<br />
<br />
= h(x)<br />
x + F<br />
(4.2)<br />
<strong>and</strong> it can be verified to increase in the value <strong>of</strong> innovation. 9<br />
Let us now look at the equilibrium behavior <strong>of</strong> the incumbent monopolist,<br />
<strong>and</strong> in particular at its incentives to invest in this competition. First <strong>of</strong> all,<br />
notice that the aggregate probability <strong>of</strong> innovation is going to be independent<br />
from the investment <strong>of</strong> the incumbent monopolist. Therefore, the expected<br />
pr<strong>of</strong>its from the leading edge technology will be the same whether the monopolist<br />
invests or not to innovate. Consider now its expected pr<strong>of</strong>its from the<br />
8 This can be easily seen comparing the respective first order conditions in a Nash<br />
equilibrium where the fixed costs are assumed low enough that all firms invest:<br />
the marginal cost <strong>of</strong> investment is higher for the monopolist because an increase<br />
in the aggregate probability <strong>of</strong> innovation reduces the expected lenght <strong>of</strong> exploitation<br />
<strong>of</strong> the current technology.<br />
9 A simple example with linear technology, h(x) =x, can be solved analytically.<br />
In this case the Marshall equilibrium implies x = √ VF − F .
4.1 A Simple Patent Race with Contractual Costs <strong>of</strong> R&D 139<br />
actual patent race. Because <strong>of</strong> the Arrow effect, the monopolist is going to<br />
invest less than the outsiders. On the other side, the outsiders are investing to<br />
maximize the expected pr<strong>of</strong>its from the actual patent race. Nevertheless, endogenous<br />
entry reduces to zero these expected pr<strong>of</strong>its. Consequently, it must<br />
be that the alternative strategy <strong>of</strong> the monopolist can only reach negative<br />
expected pr<strong>of</strong>its in the actual patent race. In conclusion, it is better for the<br />
monopolist to withdraw from the competition <strong>and</strong> retain the current flow <strong>of</strong><br />
pr<strong>of</strong>its until some other firms will innovate.<br />
Finally, we will study the case in which the incumbent monopolist is<br />
the leader <strong>of</strong> the patent race. In a Stackelberg equilibrium with endogenous<br />
entry, as long as entry takes place, the first order condition <strong>and</strong> the free entry<br />
condition at the second stage are the same as before, <strong>and</strong> they generate the<br />
same investment for the outsiders (4.2), <strong>and</strong> the same aggregate probability<br />
<strong>of</strong> innovation implicit in the free entry condition in (4.1). As a consequence<br />
<strong>of</strong> the usual neutrality result emerging under endogenous entry, the strategy<br />
<strong>of</strong> the leader is not going to affect the strategy <strong>of</strong> the active followers, but<br />
just their number. Using (4.1), we can now re-express the expected pr<strong>of</strong>its <strong>of</strong><br />
the incumbent monopolist as:<br />
π M = h(x M)V + K<br />
− x M − F =<br />
r + p<br />
= h(x M)(x + F )<br />
h(x)<br />
+ K(x + F )<br />
h(x)V<br />
− x M − F<br />
where the investment <strong>of</strong> the outsiders x is now taken as given according to<br />
the equilibrium condition (4.2). The incumbent monopolist can now exploit<br />
its first mover advantage choosing its investment according to the optimality<br />
condition:<br />
h 0 (x M )= h(x)<br />
(4.3)<br />
x + F<br />
which defines a local maximum when h 00 (x M ) < 0, aswewillassume,<strong>and</strong>it<br />
is associated with a higher investment than the one <strong>of</strong> the outsiders defined<br />
in (4.2). Since the monopolist could still invest as much as the outsiders<br />
<strong>and</strong> obtain zero expected pr<strong>of</strong>its from the actual patent race, the optimality<br />
condition above, which differs from that <strong>of</strong> the outsiders, implies that the<br />
monopolist can do even better <strong>and</strong> obtain positive pr<strong>of</strong>its from the patent<br />
race. This also implies that the strategy defined by (4.3) is always preferred<br />
to the corner strategy <strong>of</strong> not participating to the race. However, it may not<br />
be preferred to the corner strategy that deters entry. Such an entry deterring<br />
strategy would require an investment high enough to deter entry, that is<br />
h(x M )=(V − F − x)h(x)/(x + F ) − r. 10 The possibility <strong>of</strong> entry deterrence<br />
10 For instance, this is what happens in the case <strong>of</strong> a linear technology, h(x) =<br />
x. Given the expected behavior <strong>of</strong> the outsiders, the expected pr<strong>of</strong>its <strong>of</strong> the
140 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
by the monopolist was pointed out by Gilbert <strong>and</strong> Newbery (1982) in a<br />
different duopolistic framework. 11 .<br />
Notice that the level <strong>of</strong> current pr<strong>of</strong>its does not affect the equilibrium<br />
outcome, 12 which confirms two results. First, the Arrow’s paradox disappears:<br />
the monopolist that is leader in a patent race with free entry takes as given<br />
the expected value <strong>of</strong> the current monopoly <strong>and</strong> simply exploits its strategic<br />
advantage to increase the relative probability <strong>of</strong> success in the patent race.<br />
Second, the escape competition effect associated with Aghion <strong>and</strong> Griffith<br />
(2005) disappears: an increase in the intensity <strong>of</strong> product market competition<br />
associated with a decrease in the pr<strong>of</strong>its before a drastic innovation does not<br />
affect the aggregate level <strong>of</strong> innovation. In this model, effective competition<br />
for the market leads the incentives to innovate, <strong>and</strong> competition in the market<br />
cannot enhance further these incentives.<br />
In conclusion, our extension <strong>of</strong> the simple model <strong>of</strong> competition for the<br />
market analyzed in Chapter 1 allows to generalize the result obtained in that<br />
simpler environment: incumbent monopolists facing a competitive pressure<br />
in the competition for future markets behave in an aggressive way <strong>and</strong> invest<br />
more than each other rival, but they do not necessarily deter entry. We can<br />
summarize our findings as follows:<br />
Proposition 4.1. In a competition for the market with contractual<br />
costs <strong>of</strong> R&D, the incumbent monopolist invests more than<br />
any other firm <strong>and</strong> independently from its current pr<strong>of</strong>its when<br />
has a leadership <strong>and</strong> entry is endogenous.<br />
As intuitive, entry deterrence can be optimal when investment is not too<br />
costly or its marginal productivity is constant (or not too much decreasing).<br />
However, when the marginal productivity <strong>of</strong> investment diminishes strongly<br />
with the same investment, entry deterrence requires a very large <strong>and</strong> costly<br />
monopolist turn out to be linearly increasing in its investment. The monopolist<br />
is better <strong>of</strong>f deterring entry with the limit investment ¯x M = V + F − 2 √ VF− r.<br />
11 Gilbert <strong>and</strong> Newbery (1982) obtained entry deterrence by the monopolist in a<br />
deterministic contest, where investment reduces the waiting time for innovation<br />
in a deterministic way. They also suggested that a similar result could occur in<br />
stochastic patent races, providing an early insight for our result (see also Gilbert<br />
<strong>and</strong> Newbery, 1984). However, they did not move one step further <strong>and</strong> show that<br />
even when entry deterrence is not optimal, the monopolist with a first mover<br />
advantage invests more than any outsider as long as entry is free. For this reason,<br />
their result was forced to suggest a rationale for “sleeping patents” without<br />
innovative purposes <strong>and</strong> used by monopolists to preempt entry. Our point here is<br />
the exact opposite: under competitive pressure incumbent monopolists are led to<br />
invest a lot in R&D to conquer useful patents <strong>and</strong> generally without exclusionary<br />
purposes.<br />
12 This is a consequence <strong>of</strong> Prop. 3.7 since in equilibrium we have Π13 L = −(r +<br />
p) −2 = Π23h L 0 (x M).
4.1 A Simple Patent Race with Contractual Costs <strong>of</strong> R&D 141<br />
investment <strong>and</strong> becomes suboptimal. As noticed by Kortum (1993), Griliches<br />
(1994), Cohen <strong>and</strong> Klepper (1996) <strong>and</strong> other empirical works, investments in<br />
R&D are characterized by decreasing marginal productivity at the firm level.<br />
Cohen <strong>and</strong> Klepper (1996) show that “the assumption <strong>of</strong> diminishing returns<br />
to R&D is well grounded empirically” for a broad sample <strong>of</strong> industries. 13 Even<br />
Aghion <strong>and</strong> Howitt (1998, Ch.12) accept this as a stylized fact. Therefore,<br />
it is reasonable to focus on this case where the marginal productivity <strong>of</strong><br />
investment is decreasing <strong>and</strong>, accordingly, both the monopolist <strong>and</strong> some<br />
outsiders invest in R&D.<br />
4.1.2 Welfare Analysis<br />
Before, moving on in our discussion, we want to analyze our equilibria from<br />
a welfare point <strong>of</strong> view. Assuming that V ∗ is the social value <strong>of</strong> innovations,<br />
potentially higher than its private value, a social planner would maximize a<br />
welfare function based on the discounted expected social value <strong>of</strong> the innovation<br />
net <strong>of</strong> the total investment costs:<br />
P n<br />
i=1<br />
W =<br />
h(x i)V ∗ n<br />
r + P n<br />
i=1 h(x i) − X<br />
(x i + F )<br />
i=1<br />
The social planner problem amounts to choosing n ∗ firms <strong>and</strong> an investment<br />
x ∗ for each firm to solve:<br />
max W = nh(x)V ∗<br />
x,n r + nh(x) − n(x + F )<br />
Combining the optimality conditions, one obtains the optimal investment as<br />
satisfying:<br />
h(x ∗ )<br />
x ∗ + F = h0 (x ∗ )<br />
which implies that the investment <strong>of</strong> each firm is too low in Marshall equilibrium.<br />
Moreover, the number <strong>of</strong> firms is too high when the social value <strong>of</strong><br />
the innovation is small, for instance when it coincides with its private value<br />
(W n < 0 at the number <strong>of</strong> firmswhichmakesnetpr<strong>of</strong>its equal to zero), <strong>and</strong><br />
it is too low when the social value <strong>of</strong> the innovation is large enough. In Stackelberg<br />
equilibrium with endogenous entry the incumbent monopolist invests<br />
more than the outsiders, reducing the number <strong>of</strong> firms but not the aggregate<br />
13 From a theoretical point <strong>of</strong> view, notice that, while in most <strong>of</strong> the productive<br />
sectors there are good reasons to believe that doubling the amount <strong>of</strong> input total<br />
production will double (constant returns to scale hold), there are no reasons to<br />
believe that doubling the amount <strong>of</strong> inputs in the R&D activity will double the<br />
expected amount <strong>of</strong> innovations.
142 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
probability <strong>of</strong> innovation, which remains the same. This leads to a simple<br />
welfare comparison (Etro, 2008):<br />
Proposition 4.2. In the competition for the market with contractual<br />
costs <strong>of</strong> R&D <strong>and</strong> endogenous entry, the allocation <strong>of</strong> resources<br />
in the Stackelberg equilibrium with endogenous entry is<br />
Pareto superior compared to the Marshall equilibrium.<br />
4.2 Dynamic <strong>Competition</strong> for the <strong>Market</strong><br />
<strong>Competition</strong> for the market is an intrinsically dynamic phenomenon <strong>and</strong> not<br />
a static one, as we <strong>of</strong>ten remarked. Nevertheless, until now we considered<br />
simple forms <strong>of</strong> this competition where an initial investment by each firm<br />
was exhausting the research activity. In reality, firms invest over time <strong>and</strong><br />
keep investing until one <strong>of</strong> them innovates: just at that point the race is over<br />
<strong>and</strong> all firms stop spending for that innovation. In the rest <strong>of</strong> this chapter we<br />
will study patent races where firms continuously invest a flow <strong>of</strong> resources in<br />
R&D <strong>and</strong> their probability <strong>of</strong> innovation depends on this flow.<br />
Following Lee <strong>and</strong> Wilde (1980), if x i is now the flow <strong>of</strong> investment <strong>of</strong><br />
firm i determining an instantaneous probability <strong>of</strong> innovation h(x i ) assumed<br />
positive, increasing <strong>and</strong> strictly concave, the expected pr<strong>of</strong>its <strong>of</strong> a generic<br />
outsider are given by:<br />
π i =<br />
Z ∞<br />
t=0<br />
e −rt [V Pr(i wins in t)dt − x i Pr(no one wins in t)] − F =<br />
= h(x i)V − x i<br />
− F<br />
r + p<br />
which again can be rewritten as a particular case <strong>of</strong> our general formulation<br />
(2.1) employed in the previous chapters, with:<br />
Π(x i ,β i )=<br />
h(x i)V − x i<br />
[r + h(x i )+β i ]<br />
(4.4)<br />
An interesting feature <strong>of</strong> this model is that now we can determine unambiguously<br />
the sign <strong>of</strong> the cross derivative. In particular, when firm i maximizes<br />
its expected pr<strong>of</strong>its, the impact <strong>of</strong> a change in the strategy <strong>of</strong> the other firms<br />
on its marginal pr<strong>of</strong>it is:<br />
Π 12 ≡ [h0 (x i )V − 1]<br />
[r + h(x i )+β i ] 2 > 0<br />
Contrary to the simple example <strong>of</strong> Chapter 1, where investments <strong>of</strong> the firms<br />
were strategic substitutes, <strong>and</strong> to the ambiguous case <strong>of</strong> the previous section,<br />
we now realize that under more realistic conditions, investment strategies
4.2 Dynamic <strong>Competition</strong> for the <strong>Market</strong> 143<br />
are strategic complements. When a firm invests more in R&D, the aggregate<br />
probability <strong>of</strong> innovation in the sector increases <strong>and</strong> this reduces expected<br />
pr<strong>of</strong>its <strong>of</strong> the other firms, but it also increases their expected marginal pr<strong>of</strong>its,<br />
<strong>and</strong> therefore their incentives to invest.<br />
Finally, we can derive the objective function <strong>of</strong> the incumbent monopolist<br />
with a flow <strong>of</strong> current pr<strong>of</strong>its K as follows:<br />
Π (x M ,β M ,K)= h(x M)V + K − x M<br />
r + h(x M )+β M<br />
(4.5)<br />
which is again characterized by Π 13 < 0: an increase in current pr<strong>of</strong>its reduces<br />
the marginal pr<strong>of</strong>itability <strong>of</strong> investment. In what follows, we will describe in<br />
detail the equilibrium <strong>of</strong> the competition for the market under alternative<br />
forms <strong>of</strong> strategic interaction.<br />
4.2.1 Nash Equilibrium<br />
Under Nash competition the equilibrium symmetric optimality condition for<br />
the investment <strong>of</strong> each entrant is:<br />
[h 0 (x)V − 1] (r + p) =h 0 (x)[h(x)V − x] (4.6)<br />
where p = h(x M )+(n − 1)h(x) is the aggregate probability <strong>of</strong> innovation.<br />
Straightforward differentiation shows that the investment <strong>of</strong> each entrant is<br />
increasing in the expected value <strong>of</strong> innovation, in the interest rate <strong>and</strong> in the<br />
number <strong>of</strong> firms (since SC holds). If the incumbent invests, its choice x M<br />
satisfies the first order condition:<br />
[h 0 (x M )V − 1] (r + p) =h 0 (x M )[h(x M )V + K − x M ] (4.7)<br />
which differs from the previous one just because the flow <strong>of</strong> current pr<strong>of</strong>its<br />
increases the marginal cost <strong>of</strong> investment: this is a consequence <strong>of</strong> the Arrow<br />
effect <strong>and</strong> it implies that, ceteris paribus, the incumbent invests less than<br />
each entrant <strong>and</strong> has lower expected pr<strong>of</strong>its from participating to the patent<br />
race (Reinganum, 1983). Because <strong>of</strong> SC, a change in K affects all firms in the<br />
same way: for instance if we interpret an increase in the intensity <strong>of</strong> product<br />
market competition as a reduction in current pr<strong>of</strong>its K, allfirms invest more<br />
in R&D according to the escape competition effect. Summarizing we have:<br />
Proposition 4.3. A Nash equilibrium in the competition for the<br />
market implies a lower investment by the incumbent monopolist<br />
than any other firm <strong>and</strong> an investment for each firm which is decreasing<br />
in the current pr<strong>of</strong>its.
144 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
4.2.2 Marshall Equilibrium<br />
Let us assume free entry now. Since the expected pr<strong>of</strong>it functions <strong>of</strong> all firms<br />
derived in the Nash equilibrium are decreasing in the number <strong>of</strong> firms, <strong>and</strong> the<br />
incumbent expects lower pr<strong>of</strong>its from the R&D investment than the others,<br />
we can conclude that the incumbent will stop researching if the number <strong>of</strong><br />
firms is great enough: the Arrow effect induces the incumbent to withdraw<br />
from the competition for the market. Moreover, the entrants will break even<br />
if the number <strong>of</strong> firms achieves a still higher bound. This bound is defined by<br />
the free entry condition:<br />
r + p = h(x)V − x<br />
(4.8)<br />
F<br />
Rearranging the equilibrium first order condition for the outsiders <strong>and</strong> this<br />
free entry condition, we can re-express the equilibrium flow <strong>of</strong> investment in<br />
the following implicit way:<br />
h 0 (x) = 1<br />
(4.9)<br />
V − F<br />
which is increasing in the difference between the expected value <strong>of</strong> the innovation<br />
<strong>and</strong> the fixed cost, but independent from the interest rate. Moreover,<br />
the equilibrium number <strong>of</strong> firms is increasing in the value <strong>of</strong> innovation <strong>and</strong><br />
decreasing in the fixed cost <strong>of</strong> entry <strong>and</strong> in the interest rate, while it is independent<br />
from the current pr<strong>of</strong>its <strong>of</strong> the incumbent monopolist. Summing up,<br />
we have:<br />
Proposition 4.4. A Marshall equilibrium in the competition for<br />
the market implies that the incumbent monopolist does not invest<br />
<strong>and</strong> the investment <strong>of</strong> the outsiders <strong>and</strong> the aggregate probability<br />
<strong>of</strong> innovation do not depend on the current pr<strong>of</strong>its.<br />
In general, if the social value <strong>of</strong> innovation is higher enough than its<br />
private value, equilibrium investment is too low <strong>and</strong> there are too few firms.<br />
Nevertheless, if the social value <strong>of</strong> innovation is close enough to its private<br />
value, the equilibrium number <strong>of</strong> firms can be excessive.<br />
4.2.3 Stackelberg Equilibrium<br />
We will now assume that the patentholder has the opportunity to make a<br />
strategic precommitment to a level <strong>of</strong> investment in R&D. This may happen<br />
through a specific investment in R&D laboratories, by hiring researchers or<br />
in a number <strong>of</strong> other ways. Our strategic assumption seems a natural one<br />
since the patentholder can be easily seen in a different perspective from all<br />
the other entrants in the patent race.<br />
Assume that the fixed costs are low enough that the entry deterrence<br />
strategy is not optimal. Then, the incumbent leader will commit to a low
4.2 Dynamic <strong>Competition</strong> for the <strong>Market</strong> 145<br />
level <strong>of</strong> investment because such a strategy will induce a reduction in the<br />
investment <strong>of</strong> the other firms <strong>and</strong> a longer expected lifespan <strong>of</strong> the current<br />
patent (Reinganum, 1985,b). The reason <strong>of</strong> this unambiguous result is that<br />
now the Stackelberg effect <strong>and</strong> the Arrow effect work in the same direction.<br />
The first pushes toward a low investment by the monopolist because it reduces<br />
the incentives <strong>of</strong> the followers to invest as well. The second pushes in the same<br />
direction because a lower investment by the monopolist reduces the aggregate<br />
probability <strong>of</strong> innovation so as to increase the length <strong>of</strong> time in which the<br />
monopolist enjoys the pr<strong>of</strong>it flow from the current patent.<br />
More formally, each entrant chooses its own investment according to the<br />
optimality condition (4.6). In the initial stage, the choice <strong>of</strong> the leader x M<br />
satisfies the optimality condition:<br />
[h 0 (x M )V − 1] (r + p) =<br />
∙<br />
h 0 (x M )+(n − 1) ∂h(x) ¸<br />
[h(x M )V + K − x M ]<br />
∂x M<br />
(4.10)<br />
unless current pr<strong>of</strong>its are so high that the incumbent leader prefers to withdraw<br />
from the race. The system (4.6)-(4.10) defines the interior equilibrium.<br />
The effect <strong>of</strong> SC is now strengthened by the Arrow effect <strong>and</strong> leads to a low<br />
investment <strong>of</strong> the incumbent monopolist compared to the entrants. In the<br />
Appendix we show that the investment by each firm is increasing in the interest<br />
rate r <strong>and</strong> decreasing in the flow <strong>of</strong> current pr<strong>of</strong>its, but ambiguously<br />
dependent on the value <strong>of</strong> the innovation V <strong>and</strong> the number <strong>of</strong> firms n. Summarizing:<br />
Proposition 4.5. A Stackelberg equilibrium in the competition<br />
for the market implies a lower investment for the incumbent monopolist<br />
than for the other firmsaslongasentryisaccommodated;<br />
investment by each firm is decreasing in the current pr<strong>of</strong>its.<br />
An immediate corollary <strong>of</strong> this result is that a Stackelberg equilibrium<br />
implies an aggregate investment in R&D which is increasing in the interest<br />
rate <strong>and</strong> decreasing in the current pr<strong>of</strong>its <strong>of</strong> the incumbent, <strong>and</strong> an expected<br />
lifespan <strong>of</strong> the current patent which is affected in the opposite way. Compared<br />
to the Nash equilibrium, both the incumbent <strong>and</strong> each entrant invest less,<br />
<strong>and</strong>, since the number <strong>of</strong> firms is exogenous, the aggregate investment must<br />
be lower. In conclusion, a Stackelberg leadership with a fixed number <strong>of</strong> firms<br />
does not give a rationale for incumbents’ investment in R&D.<br />
Finally, notice that the escape competition effect is now working: if we<br />
imagine that an increase in product market competition decreases current<br />
pr<strong>of</strong>its K but not the value <strong>of</strong> the innovation (because this is a drastic innovation),<br />
then a more intense competition increases individual <strong>and</strong> aggregate<br />
investment in R&D.
146 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
4.2.4 Stackelberg Equilibrium with Endogenous Entry<br />
Let us now consider the endogenous entry case, in which the leader has to<br />
foresee the effects <strong>of</strong> its investment choice on the equilibrium number <strong>of</strong> entrants.<br />
In this case, as shown by Etro (2004), the results <strong>of</strong> the previous three<br />
market structures are radically modified: the incumbent monopolist has incentives<br />
to invest more than any other firm, the Arrow’s paradox disappears<br />
<strong>and</strong> the escape competition effect disappears as well.<br />
Once again, we focus on the realistic case in which entry <strong>of</strong> followers<br />
occurs in equilibrium. In the last stage all the entrants choose the same flow<br />
<strong>of</strong> investment x determined by the symmetric optimality condition:<br />
[h 0 (x)V − 1] [r +(n − 1)h(x)+h(x M )] = h 0 (x)[h(x)V − x] (4.11)<br />
Using symmetry, the zero pr<strong>of</strong>it condition becomes:<br />
h(x)V − x<br />
r +(n − 1)h(x)+h(x M ) = F (4.12)<br />
Substituting this in (4.11) we obtain the same implicit expression for the<br />
entrant’s investment as under Marshall competition (4.9):<br />
h 0 (x) = 1<br />
V − F<br />
which does not depend on the leader’s decision. However, the equilibrium<br />
number <strong>of</strong> firms does depend on the leader’s choice as predicted by the free<br />
entry condition. Totally differentiating the latter, using the fact that x does<br />
not depend on x M , delivers the expected change <strong>of</strong> investment in R&D <strong>of</strong><br />
each entrant for a change in the leader’s investment:<br />
∂ [(n − 1)h(x)]<br />
= −h 0 (x M )<br />
∂x M<br />
which shows that a higher investment <strong>of</strong> the incumbent reduces the aggregate<br />
investment <strong>of</strong> the other firms through a reduction in the number <strong>of</strong> entrants.<br />
In the initial stage, the incumbent monopolist maximizes pr<strong>of</strong>its according<br />
to the optimality condition:<br />
[h 0 (x M )V − 1] (r + p) =<br />
∙<br />
h 0 (x M )+<br />
¸<br />
∂ [(n − 1)h(x)]<br />
[h(x M )V + K − x M ]<br />
∂x M<br />
<strong>and</strong>, substituting our expression for the indirect impact ∂ [(n − 1)h(x)] /∂x M<br />
we obtain a simple equilibrium expression:<br />
h 0 (x M )= 1 V<br />
(4.13)
4.2 Dynamic <strong>Competition</strong> for the <strong>Market</strong> 147<br />
which shows a larger investment than the one <strong>of</strong> the entrants. This also<br />
implies that the equilibrium number <strong>of</strong> firms is lower than in the Marshall<br />
equilibrium. 14 Summarizing, we have:<br />
Proposition 4.6. A Stackelberg equilibrium with endogenous<br />
entry in the competition for the market implies a) the same investment<br />
as in Marshall equilibrium for the entrants with a lower<br />
number <strong>of</strong> entrants, b) a higher investment for the incumbent monopolist<br />
than for each <strong>of</strong> the other firms, <strong>and</strong> c) a higher total<br />
investment than in Marshall equilibrium.<br />
Once again, Stackelberg competition with endogenous entry induces the<br />
aggressive behavior <strong>of</strong> the incumbent. The intuition is related to the perception<br />
the leader has <strong>of</strong> the entry process. It is understood that any pr<strong>of</strong>itable<br />
opportunity for doing R&D left open by the leader will be seized by new<br />
entrants until their expected pr<strong>of</strong>its are zero. The aggregate probability <strong>of</strong><br />
innovation is determined by the free entry constraint independently from the<br />
investment <strong>of</strong> the leader <strong>and</strong> is thus taken as given by the latter. So, the monopolist<br />
looses the strategic incentive to keep its investment low: the latter<br />
is not going to affect the expected lifespan <strong>of</strong> the current patent. The Arrow<br />
effect disappears. Therefore, the only purpose <strong>of</strong> investing in R&D for the<br />
leader is to actually win the patent race, <strong>and</strong> the incentives to do it are now<br />
higher than those <strong>of</strong> any other entrant.<br />
An intuitive way to see this asymmetry relies on the fact that the leader<br />
maximizes its pr<strong>of</strong>its taking as given the aggregate probability <strong>of</strong> innovation,<br />
which is equivalent to maximize h(x M )V − x M , without taking into account<br />
the impact on the aggregate arrival rate <strong>of</strong> innovation. This impact, instead,<br />
is taken into account by each entrant <strong>and</strong> reduces the marginal pr<strong>of</strong>its <strong>of</strong> each<br />
entrant, explaining why the entrants invest less than the leader. 15 We finally<br />
derive some comparative statics in the following proposition:<br />
Proposition 4.7. A Stackelberg equilibrium with endogenous entry<br />
in the competition for the market implies an investment for each<br />
entrant firm which is increasing in the value <strong>of</strong> the innovation <strong>and</strong><br />
decreasing in the fixedcost,<strong>and</strong>aninvestmentfortheincumbent<br />
monopolist which is increasing in the value <strong>of</strong> innovation while none<br />
<strong>of</strong> them is affected by changes in the current pr<strong>of</strong>its.<br />
An immediate corollary <strong>of</strong> this result is that a Stackelberg equilibrium<br />
with endogenous entry implies an aggregate investment in R&D which is decreasing<br />
in the interest rate <strong>and</strong> independent from current pr<strong>of</strong>its. We confirm<br />
14 Also in this model we have entry deterrence when the marginal productivity <strong>of</strong><br />
investment is not too decreasing. In this case, the equilibrium investment <strong>of</strong> the<br />
monopolist satisfies h(¯x M)=(V −F )h(x)/F −x/F −r. In the rest <strong>of</strong> the chapter<br />
we will focus on the case in which there is entry <strong>of</strong> outsiders in equilibrium.<br />
15 The result holds even when the leader has a lower gain from innovation than the<br />
outsidersaslongthisgainishigherthanV − F (Lee <strong>and</strong> Sung, 2004).
148 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
that the escape competition effectemphasizedbyAghion<strong>and</strong>Griffith (2005)<br />
disappears when there is endogenous innovation by leaders: here competition<br />
for the market eliminates the impact <strong>of</strong> product market competition on the<br />
incentives to innovate. We will discuss later on the implications <strong>of</strong> this result.<br />
Finally, from a welfare point <strong>of</strong> view, a leadership reduces the number<br />
<strong>of</strong> firms <strong>and</strong> hence the expenditure in fixed costs, but it increases the total<br />
flow <strong>of</strong> investment, maintaining the aggregate probability <strong>of</strong> innovation at the<br />
same level. This makes ambiguous a welfare comparison between the Marshall<br />
outcome <strong>and</strong> the Stackelberg outcome with endogenous entry.<br />
4.2.5 Non-drastic <strong>Innovation</strong>s<br />
Until now we confined our analysis to drastic innovations.Oftentimes,once<br />
an outsider has introduced an innovation, the previous leader is not completely<br />
replaced, <strong>and</strong> both firms can still obtain positive pr<strong>of</strong>its; in these<br />
cases we have non-drastic innovations. Imagine that if the incumbent loses<br />
the patent race a duopoly between the winner <strong>and</strong> the incumbent sets in. Let<br />
us denote the value <strong>of</strong> winning the patent race for the incumbent with V W .<br />
When an outsider wins, the previous incumbent obtains V L <strong>and</strong> the entrant<br />
obtains V ≤ V W . The st<strong>and</strong>ard assumption is that, even if the innovation<br />
is drastic <strong>and</strong> the duopoly is characterized by perfect collusion, the sum <strong>of</strong><br />
the discounted pr<strong>of</strong>its obtained by the two duopolists cannot be greater than<br />
the discounted pr<strong>of</strong>its obtained by the incumbent who wins the patent race<br />
V W ≥ V + V L . Notice that the case <strong>of</strong> drastic innovations is a particular<br />
case for V W = V <strong>and</strong> V L =0. Using the properties <strong>of</strong> Poisson processes<br />
in a st<strong>and</strong>ard fashion, the objective function <strong>of</strong> each outsider is the same as<br />
before, (4.4), with a value <strong>of</strong> innovation V , while the gross expected pr<strong>of</strong>its<br />
<strong>of</strong> the incumbent monopolist are now:<br />
Π (x M ,β M ,K)= h(x M)V W + K + β M V L − x M<br />
r + h(x M )+β M<br />
(4.14)<br />
In Nash <strong>and</strong> Stackelberg equilibria the comparison between the incentives<br />
<strong>of</strong> the incumbent monopolist <strong>and</strong> the outsiders to invest are ambiguous because,<br />
beyond the usual Arrow <strong>and</strong> Stackelberg effects, we now have two new<br />
effects. On one side the gain from innovating for the incumbent is larger than<br />
for an outsider (V W >V), which increases the relative marginal benefit <strong>of</strong><br />
innovating for the incumbent. On the other side the gain from the duopolistic<br />
pr<strong>of</strong>its <strong>of</strong> the incumbent in the case in which another firm innovates (V L > 0)<br />
increases the marginal cost <strong>of</strong> innovating for the incumbent. Of course, if the<br />
first effect is strong enough, the incumbent may be the only firm to invest. 16<br />
16 Gilbert <strong>and</strong> Newbery (1982) studied an auction for a non drastic innovation between<br />
an incumbent <strong>and</strong> an entrant <strong>and</strong> noticed that the incumbent is willing<br />
to pay more for the innovation than an outsider. In theory, their deterministic
4.2 Dynamic <strong>Competition</strong> for the <strong>Market</strong> 149<br />
However, in what follows we will not deal with entry deterring strategies,<br />
but we will focus on the more realistic case where both the leader <strong>and</strong> the<br />
followers invest in R&D.<br />
Consider equilibria with endogenous entry. In a Marshall equilibrium,<br />
as long as entry <strong>of</strong> outsiders drives expected pr<strong>of</strong>its to zero, we can obtain<br />
the same equilibrium condition for the investment <strong>of</strong> each outsider (4.9) as<br />
obtained earlier, h 0 (x) =1/ (V − F ). In a Stackelberg equilibrium with endogenous<br />
entry, when the incumbent monopolist is the leader, the equilibrium<br />
is characterized by this same investment for the outsiders <strong>and</strong> by an aggregate<br />
probability <strong>of</strong> innovation p = h(x M )+β M which is again independent<br />
from the strategy <strong>of</strong> the incumbent. Accordingly, the incumbent monopolist<br />
maximizes:<br />
π M = h(x M)V W + K +[p − h(x M )] V L − x M<br />
r + p<br />
− F<br />
which is equivalent to maximize h(x M )V W − h(x M )V L − x M , <strong>and</strong> implies<br />
the optimal investment:<br />
h 0 1<br />
(x M )=<br />
V W − V L (4.15)<br />
Clearly, condition V W ≥ V + V L always implies that that the monopolist<br />
invests more than each outsider. The investment <strong>of</strong> the leader is directly<br />
related to the net perspective value <strong>of</strong> innovating V W − V L ,which<br />
is strictly higher than the one <strong>of</strong> the entrant V E . Assuming for simplicity<br />
that a symmetric duopoly takes place in case<strong>of</strong>innovationbyanoutsider,<br />
V = V L ∈ (F, V W /2), wecanconcludewith:<br />
Proposition 4.8. With non-drastic innovations, a Stackelberg<br />
equilibrium with endogenous entry in the competition for the market<br />
implies that the incumbent monopolist invests more than any<br />
other firm, all investments are not affected by changes in current<br />
pr<strong>of</strong>its, <strong>and</strong> the investment <strong>of</strong> the monopolist (outsiders) is decreasing<br />
(increasing) in the value <strong>of</strong> the duopolistic competition.<br />
Also in this case, the basic escape competition effect disappears: an increase<br />
in product market competition leading to lower current pr<strong>of</strong>its for<br />
the incumbent does not affect the investment in R&D <strong>of</strong> any firm, including<br />
the same incumbent. However, in this case, we can extend our analysis<br />
to another interesting experiment. When tougher product market competition<br />
reduces the duopolistic pr<strong>of</strong>its expected by an innovative outsider <strong>and</strong><br />
model would apply to cases in which firms can license existing innovations, however<br />
Salant (1984) has shown that the result collapses if any firm can license<br />
the patented innovation, <strong>and</strong> Czarnitzki <strong>and</strong> Kraft (2007b) have extended the<br />
model to entry <strong>of</strong> challengers (endogenizing the number <strong>of</strong> licenses) obtaining<br />
ambiguous results.
150 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
the incumbent (V ), the investment <strong>of</strong> the former is always reduced <strong>and</strong> the<br />
one <strong>of</strong> the latter is always increased: a st<strong>and</strong>ard Schumpeterian effect impacts<br />
on the outsider, <strong>and</strong> an escape competition effect àlaAghion et al.<br />
(2005) impacts on the incumbent monopolist. As we have seen, this happens<br />
also when entry in the competition for the market is endogenous. However,<br />
the aggregate impact <strong>of</strong> a higher intensity <strong>of</strong> product market competition is<br />
unambiguously in favor <strong>of</strong> the Schumpeterian effect. Formally, remembering<br />
that p = h(x M )+(n − 1)h(x), wehave:<br />
∂p<br />
∂V = h(x)<br />
F − h00 (x)(V − F ) 2 > 0<br />
Once again, we realize that when competition for the market is free, product<br />
market competition cannot increase the aggregate incentives to innovate<br />
through the escape competition effect. In a sense, when leaders are endogenously<br />
innovating to escape from the innovative pressure <strong>of</strong> the outsiders,<br />
they cannot escape also from product market competition.<br />
4.2.6 Strategic Commitments<br />
The model can also be extended to the case in which the size <strong>of</strong> innovations<br />
is actually endogenous. A widespread view claims that the innovations <strong>of</strong><br />
the outsiders are more radical since patentholders may have a technological<br />
advantage in obtaining small improvements on their technologies, so as to<br />
induce entrants to try replacing the patentholder with radical innovations.<br />
Etro (2004) questions such a view showing that in this model the incumbent<br />
monopolist invests also in more radical innovations than the other firms as<br />
long as it is the leader in the competition for the market.<br />
Before moving on, we should notice that in this chapter we focus on a<br />
purely strategic advantage for the incumbent monopolist. As we know from<br />
Chapter 2, however, similar results would emerge if we allowed the incumbent<br />
to engage in preliminary investments that could induce an aggressive<br />
behavior. For instance, the incumbent could commit to invest in R&D more<br />
than its rivals through a strategic investment that reduces the variable costs<br />
<strong>of</strong> R&D (Section 2.6), or one that increases the value <strong>of</strong> innovation (Section<br />
2.7): examples include research efforts aimed at obtaining more radical innovations,<br />
entry in related sectors where the same innovation could be fruitfully<br />
exploited in the future, or expansion <strong>of</strong> the market for the future innovation.<br />
According to our general analysis <strong>of</strong> debt financing in Section 2.8, competition<br />
for the market is the typical case in which a bias toward debt financing<br />
in the financial structure (for instance through venture capital financing)<br />
would lead to aggressive investment in a risky activity as R&D: this would<br />
endogenously reduce the cost <strong>of</strong> innovation, since in case <strong>of</strong> failure, debtholders<br />
would bear those costs.<br />
Finally, a recent interesting work by Erkal <strong>and</strong> Piccinin (2007,b) has studied<br />
R&D cartels, which are aimed at coordinating R&D investments, <strong>and</strong>
4.3 Sequential <strong>Innovation</strong>s 151<br />
research joint ventures (RJV) cartels which are aimed at sharing the results<br />
<strong>of</strong> cooperative R&D investment, in the presence <strong>of</strong> endogenous entry. 17 As<br />
we have seen in the more general case <strong>of</strong> Section 2.13, R&D cartels are ineffective<br />
as any other form <strong>of</strong> horizontal collusion, because they induce less<br />
investment for the members <strong>of</strong> the cartel than for the outsiders, which leads<br />
to lower pr<strong>of</strong>its under endogenous entry. On the contrary, RJV cartels between<br />
a small number <strong>of</strong> members can manage to increase their pr<strong>of</strong>its by<br />
coordinating on a larger investment level than the other firms. This happens<br />
because RJV cartels increase the expected value <strong>of</strong> innovation: as long as<br />
one <strong>of</strong> the members wins the race, the right <strong>of</strong> exploiting the innovation is<br />
awarded to all <strong>of</strong> them. Under endogenous entry, these cartels do not affect<br />
the aggregate arrival rate <strong>of</strong> innovations: therefore, when RJV cartels take<br />
place, they can increase welfare if an increase <strong>of</strong> the number <strong>of</strong> firms with the<br />
new technology is expected to create gains for the consumers. In other words,<br />
antitrust authorities evaluating RJV cartels should focus their attention on<br />
the foreseen impact on the product market.<br />
4.3 Sequential <strong>Innovation</strong>s<br />
Many innovative markets are characterized by a continuous development<br />
through sequential innovations. It has been sometimes argued that, in the<br />
presence <strong>of</strong> sequential technological advances, patents may stifle innovation<br />
because they may refrain outsiders from improving the existing technologies<br />
leaving the burden <strong>of</strong> innovation to slacker monopolists. 18 On the contrary, we<br />
will show that in an environment where innovations are sequential, patents<br />
<strong>and</strong> intellectual property rights play a crucial role in fostering innovation<br />
because they can start a virtuous circle <strong>of</strong> incentives to innovate, <strong>and</strong> this<br />
happens exactly when incumbent monopolists are the leaders in the patent<br />
races. The idea, fully developed in Etro (2001, 2007,a), is quite simple.<br />
In a one shot patent race the value <strong>of</strong> the expected monopolistic pr<strong>of</strong>its<br />
provides the incentives to invest in R&D, <strong>and</strong>, when entry is endogenous,<br />
the aggregate incentives are unchanged when the outsiders or the incumbent<br />
monopolist invest. However, in a sequential patent race, the value <strong>of</strong> becoming<br />
a monopolist patentholder is what provides the incentives to invest, <strong>and</strong><br />
that value is crucially affected by the role <strong>of</strong> the incumbent monopolist. If<br />
17 In a related work, De Bondt <strong>and</strong> V<strong>and</strong>ekerckhove (2007) have extended the<br />
model <strong>of</strong> Etro (2004) to the case where the players may commit to share their<br />
rewards. The larger investment by the leaders is confirmed when sharing may<br />
occur among all entrants, but not necessarily when the leader shares with all the<br />
entrants (“winner does not take all”).<br />
18 For instance, see Bessen <strong>and</strong> Maskin (2002). On this issue, see also Erkal (2005),<br />
Etro (2005d), Denicolò (2007), <strong>and</strong> Scotchmer (2004, Ch. 5) for a survey.
152 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
the latter does not invest, the market is characterized by systematic replacement<br />
<strong>of</strong> the monopolist with a new one <strong>and</strong> the value <strong>of</strong> being a patentholder<br />
coincides with the expected pr<strong>of</strong>it flow <strong>of</strong> a single patent. If the incumbent<br />
monopolist is the leader in the patent race <strong>and</strong> hence, as we know by now,<br />
invests in R&D more than any other firm, there is a chance that its monopolistic<br />
position will be preserved at the time <strong>of</strong> the new innovation, <strong>and</strong> then<br />
at the time <strong>of</strong> the following one, <strong>and</strong> so on. This possibility <strong>of</strong> a persistent<br />
innovation dramatically increases the value <strong>of</strong> being a patentholder, which in<br />
turn enhances the incentives to invest by all firms, the incumbent <strong>and</strong> the<br />
outsiders. In this case, we will have to associate a (partial) persistence <strong>of</strong><br />
monopolies with stronger incentives to invest in R&D <strong>and</strong> therefore with a<br />
faster technological progress.<br />
This process is at the source <strong>of</strong> technologically driven growth in the global<br />
economy. In this section we will examine this mechanism, dividing it in two<br />
separate steps: the first is to endogenize the value <strong>of</strong> a patent as function<br />
<strong>of</strong> the related innovation <strong>and</strong> all the subsequent innovations, <strong>and</strong> the second<br />
is to endogenize the value <strong>of</strong> technological progress in a partial equilibrium<br />
production economy. One could also take a third step <strong>and</strong> endogenize the<br />
interest rate in a general equilibrium framework, but this is beyond the scope<br />
<strong>of</strong> this book, whose analysis is limited to a partial equilibrium context.<br />
4.3.1 Endogenous Value <strong>of</strong> <strong>Innovation</strong>s<br />
Consider a sequence <strong>of</strong> drastic innovations τ =1, 2, ...T − 1,T, each one associated<br />
with the exogenous pr<strong>of</strong>it flow K τ . Every innovation can be obtained<br />
after winning a patent race as the one we studied in the previous section.<br />
Participation to the patent race for the innovation τ requires a fixed cost F τ<br />
<strong>and</strong>aninvestmentx τ , which induces an instantaneous probability <strong>of</strong> innovation<br />
h τ (x τ ) with the same properties as before, but potentially changing for<br />
different innovations. The interest rate is always exogenous <strong>and</strong> constant at<br />
the level r. The value <strong>of</strong> conquering the patent on innovation τ is defined V τ<br />
<strong>and</strong> for now will be taken as given. This is natural since it does not depend<br />
on investment choices during the regime <strong>of</strong> innovation τ − 1, <strong>and</strong>allfirms<br />
will consider it as exogenous while choosing their investments to conquer it.<br />
Accordingly, the expected pr<strong>of</strong>it <strong>of</strong>anoutsiderfirm i participating to the<br />
patent race for the innovation τ is:<br />
π iτ = h τ (x iτ )V τ − x iτ<br />
r + p τ<br />
− F τ (4.16)<br />
where p τ = P h τ (x jτ ) is the aggregate probability <strong>of</strong> innovation in this patent<br />
race. Of course, while this patent race takes place, the current monopolist<br />
has a patent on the previous innovation τ − 1, which is associated with a<br />
flow <strong>of</strong> pr<strong>of</strong>its K τ−1 . The expected pr<strong>of</strong>it <strong>of</strong> this incumbent monopolist can<br />
be expressed analogously, taking into account the flow <strong>of</strong> pr<strong>of</strong>its from the<br />
current patent:
4.3 Sequential <strong>Innovation</strong>s 153<br />
π Mτ−1 = h τ (x Mτ )V τ + K τ−1 − x Mτ<br />
− F τ · I[x Mτ > 0] (4.17)<br />
r + p τ<br />
where I[x Mτ > 0] is an indicator function with value 1 if x Mτ > 0 <strong>and</strong> 0<br />
otherwise.<br />
While the value <strong>of</strong> the innovation, the current flow <strong>of</strong> pr<strong>of</strong>its <strong>and</strong> the fixed<br />
cost <strong>of</strong> production may change over time, each patent race can be characterized<br />
exactly as in our previous analysis. In equilibrium, the investment <strong>of</strong> each<br />
firm <strong>and</strong>, in case <strong>of</strong> endogenous entry, the number <strong>of</strong> firms investing in R&D<br />
will depend (positively) on the value <strong>of</strong> the innovation in ways that we have<br />
examined earlier <strong>and</strong> that change with the kind <strong>of</strong> competition. In particular,<br />
the incumbent monopolist will not invest in a Marshall equilibrium, but will<br />
invest more than any other outsider when is leader in the patent race, as we<br />
have seen for the Stackelberg equilibrium with endogenous entry.<br />
However, following Reinganum (1985a) <strong>and</strong> Etro (2004), we can now endogenize<br />
the value <strong>of</strong> these innovations, because the value <strong>of</strong> holding patent τ<br />
must correspond to the equilibrium expected pr<strong>of</strong>it <strong>of</strong> the incumbent monopolist<br />
with the patent on the innovation τ, <strong>and</strong> the value <strong>of</strong> patent τ − 1 must<br />
correspond to the equilibrium expected pr<strong>of</strong>it <strong>of</strong> the incumbent monopolist<br />
with the patent on innovation τ − 1, <strong>and</strong> so on. Accordingly, V s = π Ms for<br />
any s = τ − 1,τ,...<br />
For instance, if Marshall competition takes place in every patent race,<br />
we know that the incumbent monopolist will not participate, each outsider<br />
will invest in the patent race for innovation τ an amount x τ (V τ ) satisfying<br />
the condition h 0 τ (x τ )(V τ − F τ )=1, <strong>and</strong> the aggregate probability <strong>of</strong> innovation<br />
will be determined by the zero pr<strong>of</strong>it condition for the outsiders. Using<br />
these equilibrium conditions, the dynamic relation that links the value <strong>of</strong><br />
subsequent innovations becomes simply:<br />
K τ−1 F τ<br />
V τ−1 =<br />
(4.18)<br />
h τ [x τ (V τ )]V τ − x τ (V τ )<br />
whose right h<strong>and</strong> side is decreasing in V τ . Given the value <strong>of</strong> the last innovation<br />
(say V T = K T /r at time T ), one can recursively obtain the value <strong>of</strong> all<br />
the previous innovations. 19<br />
Something analogous emerges with Stackelberg competition <strong>and</strong> endogenous<br />
entry. In this case the incumbent monopolist participates always to<br />
the patent race <strong>and</strong>, assuming that entry deterrence is not optimal (which<br />
19 Notice that this implies a negative relation between the value <strong>of</strong> subsequent innovations.<br />
The intuition is straightforward: if the value <strong>of</strong> innovation τ is expected<br />
to be large, there will be more investment in the patent race to obtain this innovation,<br />
which reduces the expected length <strong>of</strong> the monopoly associated with the<br />
previous patent τ − 1, whose value will be smaller as a consequence. This may<br />
lead to innovation cycles (see Etro, 2004).
154 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
requires the h τ function to be concave enough), 20 its investment x Mτ (V τ )<br />
satisfies h 0 τ (x Mτ ) V τ =1, while the equilibrium investment <strong>of</strong> the outsiders<br />
<strong>and</strong> the aggregate probability <strong>of</strong> innovation are given by the same conditions<br />
as before. The relation between subsequent values <strong>of</strong> innovations becomes:<br />
½ ¾<br />
hτ [x Mτ (V τ )]V τ + K τ−1 − x Mτ (V τ )<br />
V τ−1 =<br />
− 1 F τ (4.19)<br />
h τ [x τ (V τ )]V τ − x τ (V τ )<br />
which implies an important consequence. Since the first mover advantage<br />
represents a strategic advantage for the incumbent monopolist <strong>and</strong> increases<br />
its expected pr<strong>of</strong>its compared to the outcome without such an advantage, the<br />
value <strong>of</strong> being the incumbent monopolist is endogenously increased. 21 But,<br />
since the value <strong>of</strong> being the current monopolist is what provides the incentives<br />
to invest in R&D, also the total investment <strong>and</strong> the aggregate probability<br />
<strong>of</strong> innovation must endogenously increase. More precisely, for every innovation<br />
except the last one, the value <strong>of</strong> becoming the incumbent monopolist is<br />
higher under Stackelberg competition with endogenous entry rather than under<br />
Marshallian competition. This induces a larger investment by each firm<br />
<strong>and</strong> a larger aggregate investment when the incumbent monopolist has a first<br />
mover advantage. Summarizing, we have:<br />
Proposition 4.9. With sequential innovations, competition for<br />
the market with endogenous entry implies that the aggregate probability<br />
<strong>of</strong> innovation is higher when the incumbent monopolist has<br />
a leadership in the patent races.<br />
The bottom line is that, far from stifling innovation, incumbent monopolists<br />
facing endogenous entry <strong>of</strong> competitors enhance aggregate investment<br />
in R&D. Of course, the first mover advantage <strong>of</strong> these monopolists is a precondition<br />
for both a larger investment in R&D <strong>and</strong> a more likely persistence<br />
<strong>of</strong> technological leadership. Therefore, we obtain the paradoxical result for<br />
which endogenous entry in the competition for the market is associated with<br />
persistent monopolies.<br />
Notice that our theory suggests a way to discriminate between different<br />
degrees <strong>of</strong> persistence <strong>of</strong> leadership in innovative sectors. As we have seen,<br />
when entry <strong>of</strong> firms in the competition for the market is endogenous we should<br />
expect that technological leaders invest a lot <strong>and</strong> their persistence is more<br />
likely. Of course, when there is no competition for the market we would also<br />
expect that the leadership is persistent. However, when the degree <strong>of</strong> competition<br />
for the market is intermediate, we expect that the incumbent does not<br />
20 The analysis <strong>of</strong> sequential patent races in case <strong>of</strong> entry deterrence can be found in<br />
Denicolò (2001) in a related framework with linear technology, <strong>and</strong> an additional<br />
externality from aggregate investment, <strong>and</strong> in Etro (2001) within our framework.<br />
See also Cozzi (2007) for further discussion.<br />
21 The right h<strong>and</strong> side <strong>of</strong> (4.19) is always larger than the right h<strong>and</strong> side <strong>of</strong> (4.18).
4.3 Sequential <strong>Innovation</strong>s 155<br />
invest much in R&D <strong>and</strong> its leadership is more likely to be replaced. This suggests<br />
an inverted U curve between the degree <strong>of</strong> persistence <strong>of</strong> technological<br />
leadership <strong>and</strong> the degree <strong>of</strong> competition for the market. This may explain<br />
why it is so difficult to find empirical support for the dynamic view <strong>of</strong> competition<br />
which suggests that a leadership position should rapidly vanish. 22 In<br />
the last part <strong>of</strong> the chapter, we will discuss the relation between competition<br />
in the market <strong>and</strong> for the market, <strong>and</strong> draw some policy implications.<br />
4.3.2 Endogenous Technological Progress<br />
In all our static <strong>and</strong> dynamic description <strong>of</strong> patent races we have kept exogenous<br />
the flow <strong>of</strong> pr<strong>of</strong>its obtained by the incumbent monopolists. It is now<br />
time to endogenize it <strong>and</strong>, for this purpose, we need to describe explicitly the<br />
market through which firms exploit their innovations, employ their patents<br />
<strong>and</strong> derive their pr<strong>of</strong>its. We will do it in a framework where innovations improve<br />
the productivity <strong>of</strong> intermediate goods that are used in the production<br />
<strong>of</strong> final goods. This implies that the incentives to invest to improve the quality<br />
<strong>of</strong> these intermediate goods derive from the pr<strong>of</strong>its obtained from sales to<br />
the market for final goods.<br />
Following the pathbreaking analysis <strong>of</strong> Romer (1990), Segerstrom et al.<br />
(1990) <strong>and</strong> Aghion <strong>and</strong> Howitt (1992, 1998), consider a competitive market<br />
for final goods with a production function as: 23<br />
Z<br />
Y = (q τ j<br />
X j ) α dj (4.20)<br />
j∈J<br />
where output Y is produced employing intermediate goods <strong>of</strong> different kinds<br />
(from a set J). Each one <strong>of</strong> these intermediate goods is produced by a monopolist<br />
with a patent on its leading technology at a constant <strong>and</strong> unitary<br />
marginal cost. An infinite sequence <strong>of</strong> product innovations characterizes these<br />
intermediate goods: an innovation τ j for the intermediate good j implies that<br />
X j units <strong>of</strong> this input are equivalent to qX j units produced with the preexisting<br />
technology τ j − 1, withq>1/α, which guarantees that the innovation<br />
is drastic. Dem<strong>and</strong> for an input sold at a price 1+µ j , that is with a mark up<br />
µ j > 0, canbederivedasD τj = £ αq ατ j /(1 + µ j ) ¤ 1/(1−α)<br />
.Thisimpliesthat<br />
the pr<strong>of</strong>it maximizing price <strong>of</strong> a monopolist producing this input would be<br />
1+µ j =1/α, however, we will maintain a general expression for the equilibrium<br />
price to encompass alternative assumptions. 24 Since each sector works<br />
22 See Cable <strong>and</strong> Mueller (2006).<br />
23 Other inputs are held constant <strong>and</strong> normalized to unity for simplicity. As long<br />
as their markets are perfectly competitive the analysis is not affected by them.<br />
See Barro <strong>and</strong> Sala i Martin (1995) for a discussion.<br />
24 Our result generalizes to non-drastic innovations if Bertr<strong>and</strong> competition with<br />
free entry takes place. In such a case, the equilibrium implies limit pricing by
156 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
in the same way, in what follows we will disregard the sector index j. Hence,<br />
each patent τ for any intermediate good gives the right to a flow <strong>of</strong> pr<strong>of</strong>its:<br />
µ 1<br />
αq<br />
ατ 1−α<br />
K τ = µD τ = µ<br />
1+µ<br />
(4.21)<br />
Suppose that the probability <strong>of</strong> innovation is given by:<br />
h τ (x τ )=(φ τ x τ ) (4.22)<br />
where ∈ (0, 1). To have an idea <strong>of</strong> the realistic shape <strong>of</strong> this function, notice<br />
that the first estimate <strong>of</strong> the elasticity <strong>of</strong> the number <strong>of</strong> innovations with<br />
respect to investment in R&D by Pakes an Griliches (1980) was 0.6, while<br />
the time series study <strong>of</strong> Hausman et al. (1984) estimated an elasticity <strong>of</strong> 0.87<br />
using the Poisson distribution, decreased to 0.5 with the larger sample used<br />
by Hall et al. (1986). More recently, Kortum (1993) suggests a range between<br />
0.1 <strong>and</strong> 0.6 <strong>and</strong> Blundell et al. (2002) find a long-run elasticity close to 0.5.<br />
Most <strong>of</strong> these estimates are based on the relation between investment <strong>and</strong><br />
the number <strong>of</strong> patented innovations, which is not necessarily a good measure<br />
<strong>of</strong> innovation (since only a small percentage <strong>of</strong> patents are really valuable). 25<br />
Acemoglu <strong>and</strong> Linn (2004) have focused on the new drugs obtained in the<br />
pharmaceutical industry (rather than the new patents) obtaining an implicit<br />
estimate <strong>of</strong> the elasticity <strong>of</strong> the innovations with respect to R&D investment<br />
around 0.8.<br />
Finally, assume that new ideas are more difficult to obtain when there<br />
is an increase in the scale <strong>of</strong> the sector, as represented by expected production<br />
with the new technology. Furthermore, assume that the fixed cost is a<br />
constant fraction <strong>of</strong> the expected cost <strong>of</strong> production with the new technology.<br />
Summarizing, assume φ τ =1/ζD τ <strong>and</strong> F τ = ηD τ /(r + p τ+1 ),where<br />
ζ>0 <strong>and</strong> η ∈ (0,µ) parametrize how costly are innovations. With these last<br />
assumptions we want to capture the idea that the larger is the scale <strong>of</strong> expected<br />
production <strong>of</strong> a firm, the larger are the costs necessary to discover <strong>and</strong><br />
the last innovator (1+µ j = q for any j) <strong>and</strong> no other firms active in the market.<br />
Cournot competition with free entry would imply that more than one firm would<br />
produce intermediate goods, but Stackelberg competition in quantities with free<br />
entry would result again in having only the last innovator producing for the<br />
market <strong>and</strong> obtaining positive pr<strong>of</strong>its (something quite similar to the idea <strong>of</strong> the<br />
first mover avantage <strong>of</strong> the innovators in a world without patents advanced by<br />
Boldrin <strong>and</strong> Levine, 2005).<br />
25 Moreover, as Scotchmer (2004) notices, “these estimates should be interpreted<br />
with caution, due to the noisiness <strong>of</strong> the data. It is not clear that the estimated<br />
coefficients address the experiment <strong>of</strong> increasing the R&D spending in firms,<br />
since other circumstances <strong>of</strong> the invention environment change.” See also the<br />
discussion in Denicolò (2007). Notice that Segerstrom (2007) assumes =0.3 in<br />
his model.
4.3 Sequential <strong>Innovation</strong>s 157<br />
develop the associated technology (construction <strong>of</strong> prototypes <strong>and</strong> samples,<br />
new assembly lines <strong>and</strong> training <strong>of</strong> workers). 26 These ingredients allow us to<br />
fully characterize the equilibria <strong>of</strong> the sequential patent races in function <strong>of</strong><br />
the interest rate r. 27<br />
Under Marshall competition in the patent races the incumbent monopolist<br />
never invests in R&D <strong>and</strong> is systematically replaced by a new firm when the<br />
subsequent innovation is obtained: this process <strong>of</strong> continuous “leapfrogging”<br />
between firms implies that monopolies are not persistent <strong>and</strong> technological<br />
progress is driven by outsider firms. This is the st<strong>and</strong>ard result in the literature<br />
on Schumpeterian growth (Barro <strong>and</strong> Sala-i-Martin, 1995; Aghion <strong>and</strong><br />
Howitt, 1998), even if it has little to do with the original ideas <strong>of</strong> the late<br />
Schumpeter (1943), for which large established firms are the main drivers <strong>of</strong><br />
innovation <strong>and</strong> technological progress. The original Schumpeterian characterization<br />
<strong>of</strong> the innovation process emerges when Stackelberg competition with<br />
endogenous entry takes place in the competition for the market: when the<br />
incumbent monopolist has a first mover advantage in the patent races <strong>and</strong><br />
invests in R&D more than any other firm, its leadership is partially persistent<br />
<strong>and</strong> technological progress is driven by both the outsiders <strong>and</strong> the incumbent<br />
monopolists. 28 Moreover, as we have seen in the previous section, the partial<br />
persistence <strong>of</strong> monopoly associated with this leadership must increase the<br />
incentives to invest for all firms.<br />
As long as entry in the competition for the market is free, under both<br />
forms <strong>of</strong> competition, the aggregate probability <strong>of</strong> innovation is positively<br />
correlated to the mark up <strong>and</strong> negatively correlated to the interest rate.<br />
In particular, as shown in the Appendix, in steady state the probability <strong>of</strong><br />
innovation for each patent race is:<br />
∙ (µ ∗ ¸ ∙<br />
− η) (1 − )(µ ∗ − η)<br />
p =<br />
+1¸1−<br />
− r (4.23)<br />
ζ<br />
η<br />
where µ ∗ can be interpreted as the effective gross return on a patent. Under<br />
Marshall competition this is simply equal to the mark up µ, since this is the<br />
only gain expected by a patentholder. Under Stackelberg competition, µ ∗ is<br />
26 See Peretto <strong>and</strong> Connolly (2005) on the role <strong>of</strong> these kinds <strong>of</strong> fixed costs in<br />
endogenous growth models, <strong>and</strong> Peretto (2007) for further applications.<br />
27 Full fledged patent races with decreasing marginal productivity have been introduced<br />
in the Schumpeterian growth model in Etro (2004). The previous literature,<br />
starting with the pathbreaking contribution <strong>of</strong> Aghion <strong>and</strong> Howitt (1992)<br />
assumed linear technology <strong>of</strong> innovation so that a no-arbitrage condition was<br />
able to pin down the aggregate investment in R&D without any insights on the<br />
industrial organization <strong>of</strong> the patent races. For a related treatment <strong>of</strong> patent<br />
races in growth models see Zeira (2004).<br />
28 Here we focus on the case where is realistically low. When is high enough,<br />
the incumbent monopolist deters entry.
158 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
higher <strong>and</strong> includes the value <strong>of</strong> a partially persistent leadership, which is<br />
also increasing in the size <strong>of</strong> innovations. Summarizing, we have:<br />
Proposition 4.10. With sequential innovations, competition for<br />
the market with endogenous entry implies a steady state aggregate<br />
probability <strong>of</strong> innovation that is increasing in the mark up on<br />
patented products <strong>and</strong> that is higher when the incumbent monopolist<br />
has a leadership in the patent races.<br />
The relation (4.23) provides an implicit equilibrium relation between the<br />
interest rate <strong>and</strong> the investment in innovation, which is expressed in terms<br />
<strong>of</strong> the aggregate probability <strong>of</strong> innovation that the firms can support. Of<br />
course, a higher interest rate reduces the incentives to invest in R&D since it<br />
increases the return on alternative investments. To evaluate the consequences<br />
for growth, one could endogenize savings <strong>of</strong> the consumers as a (decreasing)<br />
function <strong>of</strong> the interest rate, <strong>and</strong> determine the equilibrium interest rate that<br />
clears the credit market (equating investments <strong>and</strong> savings) <strong>and</strong> consequently<br />
the growth rate <strong>of</strong> the economy. 29<br />
This framework can be used for a number <strong>of</strong> macroeconomic experiments,<br />
that are however beyond the scope <strong>of</strong> this book. 30 Here,wewillsummarize<br />
a few results that are relevant for our purposes. First, one can show that the<br />
decentralized equilibrium is always characterized by dynamic inefficiency because<strong>of</strong>abiasintheR&Dsectortowardfirms<br />
investing too little - essentially<br />
because, for a given total investment in R&D, too many firms do research,<br />
since they do not consider the negative externality induced by their entry on<br />
the expected pr<strong>of</strong>its <strong>of</strong> the other firms. The presence <strong>of</strong> incumbent monopolists<br />
doing a lot <strong>of</strong> research limits this inefficiency, but does not eliminate it.<br />
Dynamic inefficiency means that a reallocation <strong>of</strong> resources in the innovation<br />
sector (inducing larger research units) could increase both current <strong>and</strong> future<br />
consumption, <strong>and</strong> a consequence <strong>of</strong> this is that the optimal innovation policy<br />
29 If the final good is consumed by a representative agent with logarithmic utility,<br />
the Euler condition for utility maximization implies the growth rate <strong>of</strong> consumption<br />
g C = r − ρ, whereρ is the time preference rate. Since the equilibrium<br />
α<br />
α 1−α<br />
production <strong>of</strong> the final good must amount to Y =<br />
1+µ<br />
j∈J q κ j α<br />
1−α dj,its<br />
growth rate can be approximated as g Y =(pα ln q) /(1 − α). Equating these two<br />
expressions for the unique steady state growth rate, one obtains an implicit expression<br />
for the savings that the agent is willing to provide at a given interest<br />
rate, expressed in terms <strong>of</strong> the aggregate probability <strong>of</strong> innovation that these<br />
savings can support p =(1− α)(r − ρ) /α ln q. Equating this with (4.23) one<br />
obtains the equilibrium interest rate, <strong>and</strong> consequently the general equilibrium<br />
growth rate <strong>of</strong> the economy.<br />
30 On macroeconomic policy <strong>and</strong> the effect <strong>of</strong> aggregate dem<strong>and</strong> shocks in this<br />
framework see Etro (2001).
4.4 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> <strong>Competition</strong> for the <strong>Market</strong> 159<br />
requires always R&D subsidies. 31 Nevertheless, the equilibrium growth rate<br />
may well be below its socially optimal level (essentially because the private<br />
value <strong>of</strong> innovations can be lower than their social value), therefore the optimal<br />
innovation policy may require also subsidies to entry in the competition<br />
for the market.<br />
Segerstrom (2007) has introduced the possibility <strong>of</strong> imitation by the followers<br />
(which drives industry pr<strong>of</strong>its to zero), showing that an increase in<br />
the probability <strong>of</strong> imitation can increase the incentives to invest <strong>of</strong> the leader<br />
whose innovation has been copied (through a sort <strong>of</strong> escape competition effect),<br />
but it reduces the value <strong>of</strong> the endogenous leadership <strong>and</strong> hence the<br />
aggregate incentives to invest (that are always determined by the free entry<br />
condition for the outsiders in the competition for the market).<br />
One can also explore in more details the markets for inputs, which we assumed<br />
to be perfectly competitive in our discussion, 32 <strong>and</strong> introduce other<br />
forms <strong>of</strong> productivity growth to study their impact on the innovation activity<br />
in general equilibrium. 33 Finally, one could also extend the analysis to a<br />
multicountry framework to study global growth <strong>and</strong> the difference between<br />
strategic (unilateral) innovation policy <strong>and</strong> optimal international coordination<br />
<strong>of</strong> the same policy (in terms <strong>of</strong> R&D subsidies <strong>and</strong> protection <strong>of</strong> IPRs<br />
as well). 34<br />
4.4 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> <strong>Competition</strong> for the<br />
<strong>Market</strong><br />
The basic theories <strong>of</strong> innovation, as those described until now, suggest that<br />
competition in the patent races increases investment in R&D, but also the<br />
31 See Etro (2007a). The interesting work <strong>of</strong> Minniti (2006) has introduced the<br />
first complete analysis <strong>of</strong> multiproduct firms in the Schumpeterian framework,<br />
showing that the equilibrium is characterized by too many firms (too much interfirm<br />
diversity) <strong>and</strong> too few products per firm (too little intra-firm diversity). On<br />
the effectiveness <strong>of</strong> R&D subsidies in promoting investment in innovation see the<br />
empirical work <strong>of</strong> Aerts <strong>and</strong> Schmidt (2007).<br />
32 See Koulovatianos (2005) <strong>and</strong> Grieben (2005).<br />
33 In general, an increase in an exogenous growth rate <strong>of</strong> total factor productivity<br />
has a positive direct effect (since directly enhances the value <strong>of</strong> innovations) <strong>and</strong> a<br />
negative general equilibrium effect due to the increase in the interest rate (needed<br />
to increase savings to sustain a higher growth). This implies that an increase in<br />
total factor productivity growth increases the growth rate <strong>of</strong> the economy, but<br />
has an ambiguous impact on the percentage <strong>of</strong> income spent in R&D activity.<br />
This may explain the lack <strong>of</strong> a clear correlation between R&D per capita <strong>and</strong><br />
growth over time <strong>and</strong> across countries (see Scotchmer, 2004, Ch. 9). For related<br />
investigations see Kornprobsty (2006).<br />
34 See Etro (2007a), <strong>and</strong> Impullitti (2006 a,b, 2007).
160 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
market power <strong>of</strong> the innovators in the product market is positively related<br />
with investment in R&D. While the firstresultisconsistentwiththeevidence,<br />
the second one is, to some extent, at odd with empirical evidence. This shows<br />
a positive relation between competition <strong>and</strong> technological progress (Blundell<br />
et al., 1999), or at most a non monotone relation, positive for low levels <strong>of</strong><br />
competition <strong>and</strong> negative for high levels (Aghion et al., 2005).<br />
Aghion <strong>and</strong> Griffith (2005) have provided a possible explanation for this<br />
relation in a model <strong>of</strong> Schumpeterian growth with exogenous innovation by<br />
leaders. They consider step by step innovations, that is they assume that<br />
frontier technologies can be used by their developers while other firms have<br />
to develop them before trying to exp<strong>and</strong> the frontier. In this set up, tougher<br />
competition may increase the incentives <strong>of</strong> the leaders to innovate with the<br />
aim <strong>of</strong> escaping competition. The intuition <strong>of</strong> this “escape competition effect”<br />
is simple because, as usual, the incentives to invest for the leaders depend on<br />
the difference between the pr<strong>of</strong>its with innovation <strong>and</strong> those without innovation:<br />
competition reduces both, but tends to reduce more the pr<strong>of</strong>its <strong>of</strong> a<br />
leader that does not innovate, since a leader that obtains a drastic innovation<br />
is less constrained by competition. 35<br />
While this theory is fascinating, it is not entirely convincing. In particular,<br />
Aghion <strong>and</strong> Griffith (2005) do not derive innovation by leaders endogenously,<br />
but assume that the technological leaders invest in innovation <strong>and</strong><br />
there is not entry <strong>of</strong> outsiders in the competition for the market. 36 Since we<br />
have seen that innovation by incumbent monopolists emerges endogenously<br />
exactly when there is free entry in the competition for the market <strong>and</strong> the incumbents<br />
are leaders in this competition, leaving entry aside does not appear<br />
neutral: the escape competition effect heavily depends on the hypothesis that<br />
the leaders undertake the research activity, since st<strong>and</strong>ard incentives would<br />
drive the investment <strong>of</strong> the outsiders (namely less investment when competition<br />
is tougher). As we have noticed in a number <strong>of</strong> models, the escape<br />
competition effect works when competition for the market is exogenously<br />
limited, but when competition for the market is free we noticed that the behavior<br />
<strong>of</strong> outsiders determines the rate <strong>of</strong> innovation (constraining in a way<br />
or another the strategy <strong>of</strong> the leaders), <strong>and</strong> the escape competition effect<br />
vanishes. Finally, Aghion <strong>and</strong> Griffith (2005) do not associate the intensity<br />
<strong>of</strong> competition with more competitive structures in the product market, but<br />
with a lower price <strong>of</strong> the competitive fringe <strong>of</strong> firms, with a higher probability<br />
<strong>of</strong> entry (see also Aghion et al., 2006) or with other exogenous elements. The<br />
crucial interaction between competition in the market <strong>and</strong> for the market<br />
35 This does not happen always but just when firms are neck-<strong>and</strong>-neck, that is<br />
when the technology <strong>of</strong> the leader is similar to that <strong>of</strong> the other firms <strong>and</strong> the<br />
leader has strong incentives to escape competition. The result is strengthened<br />
when competition increases the fraction <strong>of</strong> neck-<strong>and</strong>-neck sectors.<br />
36 Aghion et al. (2005) augment the model with a single follower, but still without<br />
free entry in the competition for the market.
4.4 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> <strong>Competition</strong> for the <strong>Market</strong> 161<br />
remains to be studied for the escape competition effecttobeconvincingfrom<br />
a theoretical point <strong>of</strong> view.<br />
Denicolò <strong>and</strong> Zanchettin (2006) adopt an alternative approach <strong>and</strong> compare<br />
alternative forms <strong>of</strong> competition in the market for intermediate goods<br />
when innovations are non drastic. They describe a sort <strong>of</strong> “Darwinian selection<br />
effect” induced by competition. When this is weak many inefficient firms<br />
can be active in the product market, while tough competition is consistent<br />
with just few efficient firms. In other words, when the intensity <strong>of</strong> competition<br />
increases, inefficient firms have to exit the market leaving the most efficient<br />
ones in it. Moreover, this process gradually shifts pr<strong>of</strong>its from less efficient<br />
to more efficient firms (“front-loading effect”), that are the most recent innovators.<br />
As a result <strong>of</strong> these effects, industry pr<strong>of</strong>its for the efficient firms<br />
may increase in such a way that also the incentives to invest in R&D are<br />
strengthened.<br />
More formally, let us extend our model <strong>of</strong> section 4.3.2 with different<br />
forms <strong>of</strong> competition in the market for intermediate products. In case <strong>of</strong><br />
innovations <strong>of</strong> limited size (q 1, <strong>and</strong> it is easy to verify that with our dem<strong>and</strong> function this leads to the<br />
equilibrium price 1+µ =(1+q)/(1+α). This price is always higher than the<br />
limit price under Bertr<strong>and</strong> competition 1+µ = q, but may generate lower<br />
industry pr<strong>of</strong>its <strong>and</strong> lower pr<strong>of</strong>its for the technological leader, because part<br />
<strong>of</strong> the production <strong>of</strong> the latest innovator is replaced by the production <strong>of</strong> a<br />
less efficient firm.Thisalwayshappenswhenq is close to the monopolistic<br />
price 1/α, since industry pr<strong>of</strong>its under Bertr<strong>and</strong> competition remain close to<br />
their monopolistic level, while industry pr<strong>of</strong>its (<strong>and</strong> the pr<strong>of</strong>its <strong>of</strong> the latest<br />
37 As we know by now, a leadership for the latest innovator also in the product<br />
market competition would lead to limit pricing as well, leaving our analysis<br />
unchanged again.
162 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
innovator) under Cournot competition have a first order reduction due to the<br />
entry <strong>of</strong> a less efficient firm. 38<br />
More in general, whenever industry pr<strong>of</strong>its are lower with Cournot competition<br />
than with Bertr<strong>and</strong> competition, <strong>and</strong> they are shifted toward the less<br />
efficient firms, the incentives to innovate are lower as well - even if Cournot<br />
competition generates higher prices than Bertr<strong>and</strong> competition. In particular,<br />
in our duopolistic example, the value <strong>of</strong> becoming the last innovator is a<br />
weighted discounted average <strong>of</strong> the pr<strong>of</strong>its expected as a producer with the<br />
leading technology <strong>and</strong> with the second best technology (once a better one is<br />
invented), <strong>and</strong> under Marshall competition in the patent races, this value is<br />
what drives the investment <strong>of</strong> the outsiders (current producers do not invest<br />
because <strong>of</strong> the Arrow effect once again).<br />
Denicolò <strong>and</strong> Zanchettin (2006) show that the same positive relation between<br />
the intensity <strong>of</strong> competition <strong>and</strong> growth can emerge when there is<br />
endogenous persistence <strong>of</strong> technological leadership due to Stackelberg competition<br />
with endogenous entry in the patent races. As we have seen before,<br />
non drastic innovations that give raise to duopolies between the last two<br />
innovators do not affect the general principle for which the leader invests<br />
more than any other firm. Denicolò <strong>and</strong> Zanchettin focus on the extreme<br />
case where only the last innovator invests ( =1) <strong>and</strong> the persistence <strong>of</strong><br />
technological leadership is complete. Notice that the incentives to invest <strong>of</strong><br />
the outsiders determine the entry deterrence investment <strong>of</strong> the technological<br />
leader <strong>and</strong> those incentives depend again on a weighted discounted average<br />
<strong>of</strong> the expected pr<strong>of</strong>its in the potential duopoly. This implies that, under the<br />
same circumstances as before, Cournot competition in the product market<br />
leads to lower industry pr<strong>of</strong>its<strong>and</strong>lowerinvestmentsinR&DthanBertr<strong>and</strong><br />
competition. Nevertheless, in this case duopolistic competition in the market<br />
for intermediate goods does not take place in equilibrium since all innovations<br />
are due to a single leading firm with eternal leadership. Similar results<br />
are likely to emerge in the more realistic case where investment by outsiders<br />
takes place <strong>and</strong> the persistence <strong>of</strong> technological leadership is only partial.<br />
4.5 Conclusions<br />
In their Epilogue, Aghion <strong>and</strong> Griffith (2005) address some policy issues <strong>and</strong><br />
emphasize two contrasting views:<br />
“some commentators have argued there is a specificity <strong>of</strong> innovative<br />
markets with respect to competition. They see the role <strong>of</strong><br />
antitrust action in innovative sectors as one <strong>of</strong> counteracting incumbent<br />
firms that try to prevent innovation by new entrants by issuing<br />
38 Denicolò <strong>and</strong> Zanchettin (2006) prove that this outcome emerges under more<br />
general conditions.
4.5 Conclusions 163<br />
<strong>and</strong> accumulating (unjustified) patents. In other words, antitrust action<br />
should focus on fostering competition for the market, but not<br />
so much on increasing competition in the market, since this would<br />
reduce innovation incentives by reducing rents. In innovative markets<br />
where incumbents innovate, antitrust action should be restrained so<br />
as not to stamp out monopoly power in such markets. Instead, our<br />
analysis suggests that stimulating competition in the market, especially<br />
in sectors that are close to the corresponding world frontier<br />
<strong>and</strong>/or where incumbent innovators are neck-<strong>and</strong>-neck, can also foster<br />
competition for the market through the escape competition effect.<br />
Incumbent firms innovate precisely as a response to increased product<br />
market competition or to increased entry threat, at least up to<br />
some level.” 39<br />
We are not sure that this distinction is properly motivated. First, we<br />
do believe that there is a specificity <strong>of</strong> innovative markets with respect to<br />
competition, because firms in high-tech markets compete mainly with investments<br />
to create better products rather than with st<strong>and</strong>ard price strategies,<br />
<strong>and</strong> this should be taken into account. Second, we do not see any contradiction<br />
between the claim <strong>of</strong> the theory <strong>of</strong> market leaders <strong>and</strong> endogenous entry<br />
for which strong competition for the market enhances technological progress<br />
<strong>and</strong> the fact that competition in the market may enhance it as well under<br />
certain conditions: when this is the case, antitrust policy should be aimed at<br />
promoting both forms <strong>of</strong> competition in innovative markets. Nevertheless, we<br />
have shown that when competition for the market is characterized by endogenous<br />
entry (<strong>and</strong> by a leadership position), the incentives to invest in R&D<br />
are maximized <strong>and</strong> there is a limited space for competition in the market to<br />
enhance investment: to a large extent, competition for the market is a good<br />
substitute for competition in the market in dynamic sectors.<br />
An interesting exception to this principle derives from the Darwinian selection<br />
effect, which implies that tougher product market competition can endogenously<br />
exclude inefficient firms from production <strong>and</strong> constrain the price<br />
<strong>of</strong> the efficient ones, while still promoting innovation (<strong>of</strong> the most efficient<br />
firms) through the gains in production efficiency. 40<br />
Finally, it is clear, <strong>and</strong> in no way contradicted by the results <strong>of</strong> Aghion <strong>and</strong><br />
Griffith (2005), that the ultimate engine <strong>of</strong> market-driven innovations is associated<br />
with the possibility <strong>of</strong> exploiting the fruits <strong>of</strong> uncertain investments<br />
through intellectual property rights. Therefore, we believe that a main policy<br />
implication <strong>of</strong> this research is that antitrust policy should promote competi-<br />
39 Aghion <strong>and</strong> Griffith (2005, p. 91) associate the two positions respectively with<br />
Etro (2004) <strong>and</strong> Vickers (2001).<br />
40 This is another case in which competition leads to exit <strong>of</strong> the competitors <strong>of</strong> the<br />
leader, but it enhances consumer welfare as well.
164 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
tion both for <strong>and</strong> in the market, 41 but should never interfere with the legal<br />
protection <strong>of</strong> patents <strong>and</strong> trade secrets, which drive the private incentives to<br />
invest in R&D.<br />
With this chapter we have concluded the theoretical part <strong>of</strong> the book.<br />
In the following chapters we will move on to the policy implications <strong>of</strong> the<br />
theories we have examined.<br />
41 For a policy analysis on the benefits <strong>of</strong> product market reform taking in considerations<br />
the effects on innovation see Faini et al. (2006), Parasc<strong>and</strong>olo <strong>and</strong><br />
Sgarra (2006), Barone <strong>and</strong> Cingano (2007) <strong>and</strong> Leiner-Killinger et al. (2007).
4.6 Appendix 165<br />
4.6 Appendix<br />
Pro<strong>of</strong><strong>of</strong>Prop.4.2.Imagine that the social value <strong>of</strong> the innovation is V ∗ .<br />
Under Marshall competition with n firms investing x each, welfare is:<br />
W N = nh(x)V ∗<br />
− nx − nF<br />
r + nh(x)<br />
Under Stackelberg competition with a leader investing x M <strong>and</strong> n s −1 followers<br />
investing x, usingthefactthatnh(x) =h(x M )+(n s − 1)h(x), wehavean<br />
increase in welfare:<br />
W S = [h(x M)+(n s − 1)h(x)] V ∗<br />
r + h(x M )+(n s − 1)h(x) − x M +(n s − 1)x − n s F<br />
= W N + (x + F )(x ∙<br />
M + F ) h(xM )<br />
h(x) x M + F − h(x) ¸<br />
>W N<br />
x + F<br />
since the second term is positive because x M >x. Notice that this second<br />
term corresponds to the expected pr<strong>of</strong>it <strong>of</strong> the leader from the patent race.<br />
Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop4.5.Symmetry between the entrants in the second stage<br />
implies the equilibrium system:<br />
f(·) ≡ [h 0 (x)V − 1] [r +(n − 1)h(x)+h(x M )] − h 0 (x)[h(x)V − x] =0<br />
g(·) ≡ [h 0 (x M )V − 1] [r +(n − 1)h(x)+h(x M )]+<br />
∙<br />
− h 0 (x M )+ ∂nh(x) ¸<br />
[h(x M )V + K − x M ]=0<br />
∂x M<br />
with ∂nh(x)/∂x M = nh 0 (x)φ 0 (x M ) where x = φ(x M ) is the common reaction<br />
function for x as a function <strong>of</strong> x M <strong>and</strong> increasing in it:<br />
φ 0 (x M )=<br />
− [h 0 (x M )V − 1] h 0 (x M )<br />
h ” (x) {V [r +(n − 1)h(x)+h(x M )] + x}<br />
Since ∂φ 0 (x M )/∂r < 0, ∂φ 0 (x M )/∂K =0<strong>and</strong> ∂φ 0 (x M )/∂n > 0 , while<br />
the sign <strong>of</strong> ∂φ 0 (x M )/∂V is ambiguous, by totally differentiating the system<br />
above we obtain the comparative statics for y = r, n, K, V :<br />
" #<br />
dx<br />
dy<br />
dx M<br />
= − 1 ∙<br />
gxM<br />
dy<br />
∆<br />
¸ ∙ ¸<br />
−f xM fy<br />
−g x f x g y<br />
where ∆ ≡ f x g xM − f xM g x > 0 by assumption <strong>of</strong> stability, <strong>and</strong> assuming<br />
f x < 0 <strong>and</strong> noting that f xM > 0, f r > 0, f K =0, f n > 0, f V > 0, g x > 0,<br />
g xM < 0, g r > 0, g K < 0 while g n <strong>and</strong> g V have the only ambiguous signs. It<br />
follows that comparative statics for n <strong>and</strong> V is ambiguous, but dx M /dr > 0,<br />
dx/dr > 0, dx M /dK < 0, <strong>and</strong>dx/dK < 0. Q.E.D.
166 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
Pro<strong>of</strong><strong>of</strong>Prop.4.6: To complete the pro<strong>of</strong> we need to rigorously show<br />
that the choice <strong>of</strong> the leader is indeed a global maximum, or, in other words,<br />
that the option <strong>of</strong> zero investment is dominated by that choice. If we use the<br />
equilibrium free entry condition <strong>of</strong> the second stage to rewrite the objective<br />
function <strong>of</strong> the leader as:<br />
Π L h(x M )V + K − x M<br />
=<br />
[r +(n − 1) h(x)+h(x M )] − F = h(x M)V + K − x M<br />
F − F<br />
h(x)V − x<br />
we notice that the local maximum satisfying the first order equilibrium condition<br />
h 0 (x M ) V =1is a global maximum if:<br />
h(x M )V +K−x M<br />
h(x)V −x<br />
F − F><br />
K<br />
h(x)V −x F ⇔<br />
h(x M )V −x M<br />
h(x)V −x<br />
> 1<br />
but this is always true since we know that h(x M )V − x M >h(x)V − x. The<br />
last part follows noticing that nh(x) =h(x M )+(n s − 1)h(x) implies:<br />
nx − [x M +(n s − 1)x] = h(x M)x<br />
− x M = x ∙<br />
Mx h(xM )<br />
− h(x) ¸<br />
< 0<br />
h(x)<br />
h(x) x M x<br />
since h 00 (x) < 0. Q.E.D.<br />
Pro<strong>of</strong><strong>of</strong>Prop.4.9.Consider first the case <strong>of</strong> Marshall competition<br />
in each patent race, so that incumbent monopolists do not invest <strong>and</strong> are<br />
replaced at each innovation. Under our functional form assumptions every<br />
patent race will be characterized by an investment for each outsider:<br />
x τ = 1<br />
τ (V τ − F τ ) 1−<br />
1<br />
<strong>and</strong> by a zero pr<strong>of</strong>it condition:<br />
(φ τ x τ ) V τ − x τ<br />
= F τ<br />
r + p τ<br />
The value <strong>of</strong> innovation is simply:<br />
V τ =<br />
µD τ<br />
r + p τ+1<br />
<br />
1− φ<br />
1−<br />
where D τ is the dem<strong>and</strong> <strong>of</strong> the corresponding intermediate good sold to the<br />
final good sector as in (4.21). Solving the endogenous entry condition we<br />
have:<br />
r + p τ = (φ τ x τ ) V τ − x τ<br />
F τ<br />
=<br />
= [(φ <br />
1−<br />
τ )(V τ − F τ )] V τ − 1 1−<br />
τ (V τ − F τ ) 1<br />
1−<br />
=<br />
F τ<br />
1− φ<br />
<br />
= (/ζ) <br />
1−<br />
(µ − η) <br />
1− [µ − (µ − η)]<br />
η (r + p τ+1 ) <br />
1−
4.6 Appendix 167<br />
which shows a negative relation between probabilities <strong>of</strong> innovation <strong>of</strong> subsequent<br />
patent races, exactly as in the model <strong>of</strong> sequential patent races with<br />
an exogenous flow <strong>of</strong> pr<strong>of</strong>its for each innovation. Focusing on the steady state<br />
with a constant probability <strong>of</strong> innovation p, wecansolvetheaboverelation<br />
for the effective discount factor in steady state, r + p, thatsatisfies:<br />
(r + p) = (/ζ) <br />
1−<br />
(µ − η) <br />
1− [µ − (µ − η)]<br />
η (r + p) <br />
1−<br />
from which we obtain:<br />
∙ ¸ ∙ (µ − η) µ(1 − )+η<br />
r + p =<br />
ζ<br />
η<br />
¸1−<br />
This is increasing in the mark up, <strong>and</strong> it allows to derive explicitly the investment<br />
for each firm:<br />
µ ∙ ¸ 1<br />
x τ = 1<br />
1−<br />
1<br />
1− 1−<br />
(µ − η) D τ =<br />
ζD τ r + p τ+1<br />
= η (µ − η) D τ<br />
µ − (µ − η)<br />
that is increasing in the mark up <strong>and</strong> also in the size <strong>of</strong> dem<strong>and</strong> for the<br />
corresponding product. The explicit equilibrium expression for the value <strong>of</strong><br />
innovations is:<br />
V τ =<br />
µ (ζ/) η 1− D τ<br />
(µ − η) [µ(1 − )+η] 1−<br />
Consider now the case <strong>of</strong> Stackelberg competition with endogenous entry.<br />
In each patent race we still have the same general rules for the investment <strong>of</strong><br />
the outsiders in function <strong>of</strong> the value <strong>of</strong> innovation, <strong>and</strong> the same free entry<br />
condition as before. However, also the incumbent monopolist participates to<br />
the patent race, investing according to the following rule:<br />
x Mτ = 1<br />
<br />
1<br />
1− φ<br />
1−<br />
τ V τ<br />
1−<br />
<strong>and</strong> the value <strong>of</strong> being a monopolist with the patent τ − 1 is now given by<br />
the following recursive relation:<br />
V τ−1 = (φ τx Mτ ) V τ + K τ−1 − x Mτ<br />
r + p τ<br />
− F τ<br />
While this is general a complex relation, our modeling assumptions on technological<br />
progress allow us to derive a complete solution. First <strong>of</strong> all, notice<br />
that free entry by the outsiders determines the aggregate probability <strong>of</strong> innovation<br />
in each patent race independently from the behavior <strong>of</strong> the incumbent,<br />
while the value <strong>of</strong> innovation depends on the behavior <strong>of</strong> the leader as well.
168 4. Dynamic <strong>Competition</strong> <strong>and</strong> Endogenous Entry<br />
The pr<strong>of</strong>it function <strong>of</strong> the producers <strong>of</strong> intermediate goods implies that<br />
dem<strong>and</strong> increases in a deterministic way through subsequent innovations:<br />
D τ = q α/(1−α) D τ−1 . Because <strong>of</strong> this, it turns out that also the value <strong>of</strong><br />
innovation increases at the same rate between subsequent innovations. We<br />
can solve for this using the method <strong>of</strong> undetermined coefficients. Guessing a<br />
functional form V τ = ψD τ /(r + p τ+1 ) we have also:<br />
1−α α<br />
Dτ<br />
V τ−1 = ψq−<br />
r + p τ<br />
Substituting the equilibrium investment <strong>of</strong> the incumbent monopolist in the<br />
recursive relation above, we obtain:<br />
i h(φ τ ) 1 1 <br />
1−<br />
V τ<br />
1− Vτ + K τ−1 − 1 <br />
1<br />
1− φ<br />
1−<br />
τ V τ<br />
1−<br />
V τ−1 =<br />
− ηD τ<br />
r + p τ r + p τ+1<br />
Equating the two expressions for V τ−1 <strong>and</strong> solving in steady state we have:<br />
µ " <br />
p =<br />
ζ<br />
(1 − ) q α<br />
1−α<br />
ψ − µ + ηq α<br />
1−α<br />
# 1−<br />
ψ − r<br />
which provides a negative relation between the aggregate probability <strong>of</strong> innovation<br />
p <strong>and</strong> the rate <strong>of</strong> return from the leadership ψ (for ψ small enough):<br />
the higher is the probability <strong>of</strong> innovation, the shorter is the lifetime <strong>of</strong> an<br />
innovation, <strong>and</strong>, consequently, the lower is the value <strong>of</strong> being a leader.<br />
Moreover, using again our guess, we can solve for the free entry condition<br />
fortheoutsidersasbefore:<br />
∙ ¸ ∙ ¸1− (ψ − η) ψ(1 − )+η<br />
p =<br />
− r<br />
ζ<br />
η<br />
This is a positive relation between the aggregate probability <strong>of</strong> innovation p<br />
<strong>and</strong> the rate <strong>of</strong> return from leadership ψ: the higher is the value <strong>of</strong> being a<br />
leader, the larger will be the investment in R&D <strong>and</strong> hence the probability<br />
<strong>of</strong> innovation.<br />
Finally, putting together these last two relations we can derive an implicit<br />
expression for the equilibrium value <strong>of</strong> ψ:<br />
ψ(µ) =µ +<br />
(1 − ) ηq 1−α α 1<br />
ψ<br />
1−<br />
(ψ − η) <br />
1−<br />
[ψ(1 − )+η] − ηq α<br />
which must be larger than µ for our guess to be consistent (otherwise the<br />
incumbent monopolist would not find it convenient to invest), which can be<br />
verified to be the case for a wide set <strong>of</strong> parameter values.<br />
To close our equilibrium description, the investments <strong>of</strong> the incumbent<br />
monopolists <strong>and</strong> <strong>of</strong> each outsider for any patent race are:<br />
1−α
4.6 Appendix 169<br />
x Mτ =<br />
ηψD τ<br />
ψ − (ψ − η)<br />
x τ = η (ψ − η) D τ<br />
ψ − (ψ − η)<br />
where the first is 1/(1 − η/ψ) times the second. The expression for the aggregate<br />
probability <strong>of</strong> innovation can be obtained with µ ∗ = µ in case <strong>of</strong><br />
Marshall equilibrium <strong>and</strong> µ ∗ = ψ in case <strong>of</strong> Stackelberg equilibrium with<br />
endogenous entry. Q.E.D.
5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
The scope <strong>of</strong> antitrust policy is to avoid distortions <strong>of</strong> competition that may<br />
negatively affect consumers, like collusive arrangements aimed at fixing prices<br />
above their competitive level, mergers aimed at creating a dominant position,<br />
<strong>and</strong> abuse <strong>of</strong> dominance by market leaders against the interests <strong>of</strong> consumers.<br />
Given the particular focus on market leaders in this book, our attention in<br />
this chapter will be mainly on the last aspect <strong>of</strong> antitrust policy: abuse <strong>of</strong><br />
dominance with anticompetitive purposes. 1<br />
In the United States the main federal antitrust statute is the Sherman Act<br />
<strong>of</strong> 1890, which was developed in reaction to the widespread growth <strong>of</strong> large<br />
scale business cartels <strong>and</strong> trusts. Section 1 <strong>of</strong> the Sherman Act prohibits restraints<br />
<strong>of</strong> trade in general, while Section 2 deals with monopolization stating<br />
that:<br />
“Every person who shall monopolize, or attempt to monopolize, or<br />
combine or conspire with any other person or persons, to monopolize<br />
any part <strong>of</strong> trade or commerce among the several States, or with<br />
foreign nations, shall be deemed guilty <strong>of</strong> a felony”.<br />
Enforcement at the federal level is shared by the <strong>Antitrust</strong> Division <strong>of</strong> the<br />
Department <strong>of</strong> Justice <strong>and</strong> by the Federal Trade Commission. The current<br />
interpretation <strong>of</strong> US antitrust law associates abusive conduct with predatory<br />
or anticompetitive actions having the specific intent to acquire, preserve or<br />
enhance monopoly power distinguished from acquisition through a superior<br />
product, business acumen or historical accident (hence monopoly per se is<br />
not illegal). It is generally accepted that an action is anticompetitive when it<br />
harms consumers.<br />
In Europe, competition policy has a more recent history which is mostly<br />
associated with the creation <strong>of</strong> the European Union <strong>and</strong> its coordination <strong>of</strong><br />
policies for the promotion <strong>of</strong> free competition in the internal market. The<br />
main provisions <strong>of</strong> European <strong>Competition</strong> Law concerning abuse <strong>of</strong> domi-<br />
1 On the first two aspects <strong>of</strong> antitrust, see Motta (2004, Ch. 4-5) for a wide survey<br />
<strong>of</strong> the economic literature, <strong>and</strong> Sections 2.13 <strong>and</strong> 3.5 for some implications <strong>of</strong><br />
the endogenous entry approach. Recently, Rhee (2006) has applied aspects <strong>of</strong> the<br />
theory <strong>of</strong> market leaders to merger policy for the New Economy.
172 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
nance are contained in the Article 82 <strong>of</strong> the Treaty <strong>of</strong> the European Communities<br />
which states that:<br />
“Any abuse by one or more undertakings <strong>of</strong> a dominant position<br />
withinthecommonmarketorinasubstantialpart<strong>of</strong>itshallbeprohibited<br />
as incompatible with the common market in so far as it may<br />
affect trade between Member States. Such abuse may, in particular,<br />
consist in: (a) directly or indirectly imposing unfair purchase or selling<br />
prices or other unfair trading conditions; (b) limiting production,<br />
markets or technical development to the prejudice <strong>of</strong> consumers; (c)<br />
applying dissimilar conditions to equivalent transactions with other<br />
trading parties, thereby placing them at competitive disadvantage; (d)<br />
making the conclusion <strong>of</strong> contracts subject to acceptance by other parties<br />
<strong>of</strong> supplementary obligations which, by their nature or according<br />
to commercial usage, have no connection with the subject <strong>of</strong> such<br />
contracts.”<br />
This article on abuse <strong>of</strong> dominance is part <strong>of</strong> the law <strong>of</strong> each member<br />
state <strong>and</strong> is enforced by the European Commission <strong>and</strong> by all the National<br />
<strong>Competition</strong> Authorities (as Article 81 on horizontal <strong>and</strong> vertical agreements<br />
<strong>and</strong> the Merger Regulation). 2 The application <strong>of</strong> EU competition law on<br />
abuse <strong>of</strong> dominance involves the finding <strong>of</strong> a dominant position <strong>and</strong> <strong>of</strong> an<br />
abusive behavior <strong>of</strong> the dominant firm, usually associated with exploitative<br />
practices such as excessive pricing, 3 <strong>and</strong> with exclusionary practices such<br />
as predatory pricing, rebates, tying or bundling, exclusive dealing or refusal<br />
to supply. However, the analysis <strong>of</strong> both dominance <strong>and</strong> abusive behaviors<br />
entails complex economic considerations <strong>and</strong> its reform in the EU is the<br />
subject <strong>of</strong> an ongoing debate.<br />
Many economists have pointed out the necessity <strong>of</strong> a closer focus on consumer<br />
welfare in the implementation <strong>of</strong> competition policy with specific reference<br />
to abuses <strong>of</strong> dominance. While antitrust legislation was written with this<br />
objective in mind, its concrete application has sometimes been biased against<br />
market leaders <strong>and</strong> in defense <strong>of</strong> their competitors rather than toward the<br />
defense <strong>of</strong> competition <strong>and</strong> <strong>of</strong> the interests <strong>of</strong> consumers. The two objectives<br />
do not necessarily overlap. The development <strong>of</strong> the New Economy, characterized<br />
by very dynamic <strong>and</strong> innovative markets, has increased the pressure<br />
for a new approach, already somewhat developed in the United States, <strong>and</strong><br />
in progress in Europe. An important EU Report by Rey et al. (2005), has<br />
2 TheCommissionactsbothasaprosecutor<strong>and</strong>judgeatafirst level. The Court<br />
<strong>of</strong> First Instance has jurisdiction in all actions brought against the decisions <strong>of</strong><br />
the Commission, while the European Court <strong>of</strong> Justice decides on appeal actions<br />
brought against the judgments <strong>of</strong> the Court <strong>of</strong> First Instance. Motta (2004)<br />
provides a careful treatment <strong>of</strong> competition policy in the EU. For a non-technical<br />
treatment see Riela (2005).<br />
3 On this point see Katsoulacos (2006).
5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance 173<br />
recently argued in favor <strong>of</strong> an effects-based approach to competition policy,<br />
which associates abuses <strong>of</strong> dominant positions with anti-competitive strategies<br />
that harm consumers.<br />
In line with this proposal, we believe that a new approach to competition<br />
policy should be based on rigorous economic analysis, from both a theoretical<br />
<strong>and</strong> an empirical point <strong>of</strong> view. Rey et al. (2005) emphasize this element in<br />
the antitrust procedure:<br />
“a natural process would consist <strong>of</strong> asking the competition authority<br />
to first identify a consistent story <strong>of</strong> competitive harm, identifying<br />
the economic theory or theories on which the story is based, as<br />
well as the facts which support the theory as opposed to competing<br />
theories. Next, the firm should have the opportunity to present its<br />
defence, presumably to provide a counter-story indicating that the<br />
practice in question is not anticompetitive, but is in fact a legitimate,<br />
perhaps even pro-competitive business practice.”<br />
Moreover, any theory <strong>of</strong> the market structure able to provide guidance<br />
in detecting abuses <strong>of</strong> dominant positions should take into account the role<br />
<strong>and</strong> the strategies <strong>of</strong> market leaders, describe the equilibrium outcomes as a<br />
function <strong>of</strong> the entry conditions <strong>and</strong> <strong>of</strong> the dem<strong>and</strong> <strong>and</strong> supply conditions,<br />
<strong>and</strong> provide welfare comparisons under alternative set-ups.<br />
In this chapter we will try to argue that, while the Chicago school <strong>and</strong><br />
the post-Chicago approach had problems in providing a unified framework<br />
which matches these requirements, the theory <strong>of</strong> market leaders formalized<br />
in the previous chapters has provided alternative insights that may be useful<br />
for this purpose. The general principle derived until now is that market<br />
leaders may behave in an anti-competitive way, accommodating or predatory,<br />
in markets where the number <strong>of</strong> firms is exogenous (meaning that outsiders<br />
cannot overcome barriers to entry even when there are pr<strong>of</strong>itable opportunities),<br />
while they always behave in an aggressive way when entry into the<br />
market is endogenous (meaning that it depends on the pr<strong>of</strong>it opportunities).<br />
In the first situation a large market share <strong>of</strong> the leader can be the fruit <strong>of</strong><br />
anti-competitive strategies, but in the second situation a large market share<br />
<strong>of</strong> the leader is a consequence <strong>of</strong> its aggressive strategies <strong>and</strong> <strong>of</strong> the entry conditions,<br />
<strong>and</strong> not <strong>of</strong> market power. Therefore, there should be no presumption<br />
<strong>of</strong> a positive association between market shares <strong>and</strong> market power unless the<br />
lack <strong>of</strong> free entry conditions has been established.<br />
This has a main implication: while the old approach to abuses <strong>of</strong> dominant<br />
positions needs to verify dominance through structural indicators <strong>and</strong><br />
the existence <strong>of</strong> a certain abusive behavior, a new economic approach would<br />
simply need to verify the existence <strong>of</strong> harm to consumers. As Rey et al. (2005)<br />
correctly point out, “the case law tradition <strong>of</strong> having separate assessments<br />
<strong>of</strong> dominance <strong>and</strong> <strong>of</strong> abusiveness <strong>of</strong> behavior simplifies procedures, but this<br />
simplification involves a loss <strong>of</strong> precision in the implementation <strong>of</strong> the legal
174 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
norm. The structural indicators which traditionally serve as proxies for ‘dominance’<br />
provide an appropriate measure <strong>of</strong> power in some markets, but not<br />
in others”, in particular not in markets where entry is an important factor<br />
(a concentration index is uniquely concerned with actual competition <strong>and</strong><br />
ignores potential competition) <strong>and</strong> when innovation is important (a concentration<br />
index can deal with competition in the market, not for the market).<br />
In this chapter we review the traditional approaches to antitrust analysis<br />
in Section 5.1 <strong>and</strong> the market leaders approach in Section 5.2, while Section<br />
5.3 contains a digression on the protection <strong>of</strong> IPRs. We apply our results to<br />
a pr<strong>of</strong>ound policy oriented discussion in Section 5.4 <strong>and</strong> conclude in Section<br />
5.5.<br />
5.1 The Traditional Approaches to Abuse <strong>of</strong> Dominance<br />
In this section, we review some aspects <strong>of</strong> the traditional approaches to antitrust<br />
policy on abuse <strong>of</strong> dominance <strong>and</strong> start comparing them with the<br />
insights <strong>of</strong> the recent theoretical attempts to build a comprehensive theory<br />
<strong>of</strong> market leadership <strong>and</strong> competition policy.<br />
5.1.1 The Chicago School<br />
The so-called pre-Chicago approach was mostly based on the simplistic insights<br />
<strong>of</strong> the early studies on imperfect competition, which associated monopolistic<br />
behavior <strong>and</strong> abusive conduct with firms having large market shares.<br />
Such a naïve view has been challenged since the 50s- 60s by what we now call<br />
the “Chicago school”, led by Aaron Director <strong>and</strong> other exponents <strong>of</strong> the Law<br />
School <strong>of</strong> the University <strong>of</strong> Chicago, whose main merit has been to introduce a<br />
systematic economic approach to antitrust - as opposed to what Posner (2001)<br />
calls the “populist” approach <strong>and</strong> Bork (1993) associates with a “farrago <strong>of</strong><br />
amorphous <strong>and</strong> leftist political <strong>and</strong> sociological propositions”. 4 While the<br />
Chicago school was seriously attacking collusive agreements as conducive to<br />
large welfare losses, it was less critical <strong>of</strong> mergers <strong>and</strong> exclusionary practices.<br />
Many scholars were (<strong>and</strong> still are) convinced that, when there are potential<br />
entrants in a given sector, mergers are mostly aimed at creating beneficial<br />
4 Bork (1993) cites the following two primary characteristics <strong>of</strong> the early Chicago<br />
school. “The first is the insistence that the exclusive goal <strong>of</strong> antitrust adjudication,<br />
the sole consideration the judge must bear in mind, is the maximization <strong>of</strong><br />
consumer welfare. The judge must not weigh against consumer welfare any other<br />
goal, such as the supposed social benefits <strong>of</strong> preserving small businesses against<br />
superior efficiency. Second, the Chicagoans applied economic analysis more rigorously<br />
than was common at the time to test the propositions <strong>of</strong> the law <strong>and</strong> to<br />
underst<strong>and</strong> the impact <strong>of</strong> business behavior on consumer welfare” (p. xi).
5.1 The Traditional Approaches to Abuse <strong>of</strong> Dominance 175<br />
cost efficiencies, while aggressive strategies such as bundling, price discrimination<br />
<strong>and</strong> exclusive dealing, are not necessarily anti-competitive but may<br />
instead have a strong efficiency rationale behind them. For instance, bundling<br />
is typically used for price discrimination purposes <strong>and</strong> not for exclusionary<br />
purposes. Moreover, according to a widespread view in the Chicago school,<br />
there is no such a thing as predatory pricing, which is a reduction <strong>of</strong> the<br />
price below cost to induce exit by the competitors in order to compensate for<br />
the initial losses with future monopolistic pr<strong>of</strong>its. The main reason is that, if<br />
the predator can sustain such initial losses, any other prey can also sustain<br />
the induced losses (which are smaller since its output is lower) as long as<br />
credit markets are properly working, therefore predatory pricing would not<br />
be effective to start with. 5<br />
More recently, Posner (2001) has taken a less extreme position, claiming<br />
that:<br />
“there is an economic basis for concern with at least some exclusionary<br />
practices, in at least some circumstances; <strong>and</strong> a few practices<br />
that are not exclusionary (though so classified in the law), like<br />
persistent price discrimination, may still be undesirable on strictly<br />
economic grounds” (Posner, 2001, p. 4)<br />
Accordingly, Posner proposes a moderate st<strong>and</strong>ard for judging practices<br />
claimed to be exclusionary:<br />
“in every case in which such a practice is alleged, the plaintiff<br />
must prove first that the defendant has monopoly power <strong>and</strong> second<br />
that the challenged practice is likely in the circumstances to exclude<br />
from the defendant’s market an equally or more efficient competitor.<br />
The defendant can rebut by proving that although it is a monopolist<br />
<strong>and</strong> the challenged practice exclusionary, the practice is, on balance,<br />
efficient” (ibidem, pp. 194-5).<br />
This efficiency defense is at the basis <strong>of</strong> the rule <strong>of</strong> reason approach, for<br />
which a business practice is not per se illegal, but can be justified if it does<br />
not harm consumers or creates efficiencies.<br />
In the modern economic debate, the Chicago school has been criticized for<br />
failing to provide results that were robust enough to withst<strong>and</strong> full-fledged<br />
game theoretic analysis <strong>of</strong> dynamic competition between incumbents <strong>and</strong><br />
5 See McGee (1958) on the St<strong>and</strong>ard Oil Trust (1911), a famous US case <strong>of</strong> predatory<br />
pricing which led to the break up <strong>of</strong> the Rockefeller’s oil refining company<br />
into thirty four small companies. Beyond what pointed out in the text, another<br />
reason why firms should not engage in predatory pricing is that a merger would<br />
be a better solution. Of course this alternative is not viable if the merger is prohibited<br />
by the same antitrust law (but in the US, mergers capable <strong>of</strong> reducing<br />
competition became the subject <strong>of</strong> antitrust investigations only after the Clayton<br />
Act <strong>of</strong> 1914).
176 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
entrants. The so-called “post-Chicago” approach has shown that in the presence<br />
<strong>of</strong> strategic asymmetries between incumbents <strong>and</strong> entrants <strong>and</strong> pervasive<br />
market imperfections, strategies such as price-cuts, bundling or vertical restraints<br />
can be anti-competitive because they can successfully deter entry<br />
in the short run <strong>and</strong> protect monopolistic rents in the long run. Broadly<br />
speaking, US antitrust authorities have been highly influenced by all <strong>of</strong> these<br />
approaches over time, while it is hard to claim that the same is true <strong>of</strong><br />
the EU antitrust authorities. It has recently been pointed out by Ahlborn<br />
et al. (2004) that “in Europe it has taken longer for new developments in<br />
economic theory to affect competition policy. While U.S. antitrust has been<br />
influenced by Chicago school <strong>and</strong> post-Chicago school theories, pre-Chicago<br />
school considerations still play a role in Europe, albeit at times dressed up<br />
in post-Chicago clothing”. 6<br />
We believe that the Chicago school provided fundamental insights into<br />
many antitrust issues, but it failed to provide a complete underst<strong>and</strong>ing <strong>of</strong><br />
the behavior <strong>of</strong> market leaders. In particular, it limited most <strong>of</strong> its analysis<br />
to the underst<strong>and</strong>ing <strong>of</strong> how monopolistic <strong>and</strong> perfectly competitive markets<br />
work, <strong>and</strong> in a few cases it focused on markets characterized by a monopolist<br />
facing a competitive fringe <strong>of</strong> potential entrants. 7 Dismissing the useful<br />
progress in the applications <strong>of</strong> game theory, the Chicago school ignored the<br />
important role <strong>of</strong> the strategic interactions between incumbents <strong>and</strong> entrants.<br />
Consequently, its approach to exclusionary practices has <strong>of</strong>ten been biased<br />
toward a competitive role <strong>of</strong> the incumbents without an updated theoretical<br />
support.<br />
5.1.2 The Post-Chicago Approach<br />
In the 80s, while the Chicago school was succeeding in reducing the enforcement<br />
attitudes <strong>of</strong> US antitrust law, especially under the Reagan Administration,<br />
a new school <strong>of</strong> thought started to exp<strong>and</strong> its influence between economists<br />
<strong>and</strong>, in the following decade, also between antitrust scholars. The socalled<br />
post-Chicago approach introduced new game theoretic tools to study<br />
complex market structures <strong>and</strong> derive sound normative implications, which<br />
6 A symptom <strong>of</strong> the pervasive European approach, which is more against market<br />
leaders than in favor <strong>of</strong> free entry, emerges in basic business <strong>and</strong> industrial<br />
regulation. For instance, in many European countries the development <strong>of</strong> large<br />
chains <strong>of</strong> supermarkets is condemned as an unfair threat to small retail businesses.<br />
Similarly, it is hard to liberalize entry in the markets for taxicabs even if<br />
the efficiency gains for the consumers would be quite clear.<br />
7 Somewhatrelatedtothisliteratureisthe theory <strong>of</strong> contestable markets by Baumol<br />
et al. (1982), which, however, was mostly limited to simple forms <strong>of</strong> price<br />
competition with homogenous goods. The theory <strong>of</strong> Stackelberg competition with<br />
endogenous entry generalizes that theory to product differentiation <strong>and</strong> other<br />
forms <strong>of</strong> competition.
5.1 The Traditional Approaches to Abuse <strong>of</strong> Dominance 177<br />
represents one <strong>of</strong> the main contributions <strong>of</strong> this line <strong>of</strong> research. With reference<br />
to exclusionary practices, the post-Chicago approach has shown that in<br />
thepresence<strong>of</strong>strategiccommitmentsto undertake preliminary investments,<br />
<strong>of</strong> asymmetric information between firms, <strong>of</strong> credit market imperfections or<br />
in the presence <strong>of</strong> limited forms <strong>of</strong> irrationality, predatory pricing can be<br />
an equilibrium strategy for the incumbent, can deter entry <strong>and</strong> it can harm<br />
consumers. Similarly, it has shown that bundling can be used to strengthen<br />
price competition <strong>and</strong> exclude a rival from a secondary market. Analogously,<br />
many other strategies can have an exclusionary purpose.<br />
One should keep in mind that many <strong>of</strong> the results <strong>of</strong> the post-Chicago<br />
approach (summarized in the early but still unsurpassed work <strong>of</strong> Tirole,<br />
1988) are quite weak, <strong>and</strong> they largely depend on a number <strong>of</strong> restrictive<br />
assumptions. For example, predatory pricing has been shown to be exclusionary<br />
under extreme circumstances, including forms <strong>of</strong> irrational behavior<br />
(in reputation models) or pervasive market imperfections, <strong>and</strong>, even when<br />
exclusion emerges under more plausible conditions, it is not necessarily associated<br />
with a pricing below cost or even with reductions in consumer welfare<br />
(in signalling models), which is what should matter in drawing antitrust implications.<br />
Nevertheless, the intellectual achievements <strong>of</strong> the post-Chicago<br />
approach, especially the introduction <strong>of</strong> game theory as the ultimate tool <strong>of</strong><br />
industrial organization <strong>and</strong> the pro<strong>of</strong> <strong>of</strong> the possibility <strong>of</strong> pr<strong>of</strong>itable exclusionary<br />
strategies, are remarkable.<br />
Our critique <strong>of</strong> the post-Chicago approach is not focused on its game theoretic<br />
foundation or on its specific results, but on the general applicability <strong>of</strong><br />
these results for policy purposes. In most cases, the modern game theoretic<br />
literature in industrial organization has studied the behavior <strong>of</strong> incumbent<br />
monopolists facing a single potential entrant. To cite the most known theoretical<br />
works with strong relevance for antitrust issues, this was the case <strong>of</strong> the<br />
Dixit (1980) model <strong>of</strong> entry deterrence, <strong>of</strong> the models by Kreps <strong>and</strong> Wilson<br />
(1982) <strong>and</strong> Milgrom <strong>and</strong> Roberts (1982) <strong>of</strong> predatory pricing, by Fudenberg<br />
<strong>and</strong> Tirole (1984) <strong>and</strong> by Bulow et al. (1985) on strategic investment, by<br />
Br<strong>and</strong>er <strong>and</strong> Lewis (1986) on strategic debt financing, by Rey <strong>and</strong> Stiglitz<br />
(1988) <strong>and</strong> Bonanno <strong>and</strong> Vickers (1988) on vertical restraints, by Whinston<br />
(1990) on bundling for entry deterrence purposes, <strong>and</strong> many other subsequent<br />
works based on the analysis <strong>of</strong> duopolies with an incumbent <strong>and</strong> an<br />
entrant. 8 Most <strong>of</strong> the st<strong>and</strong>ard results on the behavior <strong>of</strong> incumbents in terms<br />
<strong>of</strong> pricing, R&D investments, mergers, quality choices <strong>and</strong> vertical <strong>and</strong> horizontal<br />
differentiation are derived in duopolistic models, where the incumbent<br />
chooses its own strategies in competition with a single entrant. While this<br />
analysis simplifies the interaction between incumbents <strong>and</strong> competitors, it<br />
can be highly misleading, since it assumes away the possibility <strong>of</strong> endogenous<br />
entry, <strong>and</strong> hence limits its relevance to situations where the incumbent<br />
already has an exogenous amount <strong>of</strong> market power.<br />
8 See Motta (2004) <strong>and</strong> Whinston (2006) on the post-Chicago approach.
178 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
It is not surprising that the results <strong>of</strong> the post-Chicago approach are<br />
<strong>of</strong>ten biased toward an anti-competitive role <strong>of</strong> the incumbents: these incumbents<br />
engage in predatory pricing, threaten or undertake overinvestments in<br />
complementary markets <strong>and</strong> patent new technologies only to preempt entry,<br />
impose exclusive dealing contracts, or bundle their goods with the sole<br />
purpose <strong>of</strong> deterring the entry <strong>of</strong> the competitor. Otherwise they are accommodating,<br />
engaging in excessive pricing or in anticompetitive mergers aimed<br />
at increasing prices, or stifling innovation to preserve their power. In such a<br />
simple scenario, what antitrust authorities should do is unambiguously fight<br />
against incumbents: punish their aggressive pricing strategies as predatory,<br />
<strong>and</strong> their accommodating pricing strategies as exploitative, punish investments<br />
in complementary markets as attempts to monopolize them, weaken<br />
their intellectual property rights, forbid bundling strategies, prohibit mergers<br />
<strong>and</strong> so on. The bottom line is that, according to this view, antitrust authorities<br />
should sanction virtually any behavior <strong>of</strong> the incumbents which does not<br />
conform to that <strong>of</strong> their competitors.<br />
The fallacy <strong>of</strong> this line <strong>of</strong> thought, in our view, derives from a simple fact:<br />
it is based on a partial theory <strong>of</strong> oligopoly limited to the analysis <strong>of</strong> duopolies<br />
with an incumbent <strong>and</strong> an entrant which does not take into account that,<br />
at least in most cases, entry by competitors is not an exogenous fact, but<br />
an endogenous choice. Whether entry is more or less costly, it is typically<br />
the fruit <strong>of</strong> an endogenous decision by the potential competitors. Of course,<br />
entry can be regarded as an exogenous phenomenon in the case <strong>of</strong> a natural<br />
monopoly or when there are legal barriers to entry, but these cases should not<br />
be a subject <strong>of</strong> antitrust analysis, but <strong>of</strong> regulatory analysis. When entry can<br />
be regarded as an endogenous element which depends on the technological<br />
conditions that constrain the pr<strong>of</strong>itability <strong>of</strong> the firms, we need a complete<br />
underst<strong>and</strong>ing <strong>of</strong> the behavior <strong>of</strong> leaders facing endogenous entry.<br />
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous<br />
Entry<br />
The theory <strong>of</strong> market leaders studied in the previous chapters clarifies the role<br />
<strong>of</strong> market leaders <strong>and</strong> <strong>of</strong> the entry conditions in a game theoretic framework<br />
that is more general than most analysis within the post-Chicago approach.<br />
In this section we will review its results <strong>and</strong> compare its implications for<br />
antitrust with those <strong>of</strong> the traditional approaches, but before doing that,<br />
we need to clarify a few concepts concerning the determinants <strong>of</strong> entry in a<br />
market. 9<br />
The industrial organization literature has emphasized different kinds <strong>of</strong><br />
constraints on entry. The definition <strong>of</strong> barriers to entry has been quite debated<br />
in the literature. Bain (1956) associated them with the situation in<br />
9 This section is partly derived from Etro (2006b).
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous Entry 179<br />
which established firms can elevate their selling prices above minimal average<br />
costs <strong>of</strong> production without inducing entry in the long run. Broadly<br />
speaking, such a situation corresponds to what we define as competition between<br />
an exogenous number <strong>of</strong> firms: even if pr<strong>of</strong>its can be obtained in the<br />
market, entry is not possible. Stigler (1968) has proposed a different definition<br />
<strong>of</strong> barriers to entry, associating them with costs <strong>of</strong> production which<br />
must be borne by firms seeking to enter an industry but not borne by the<br />
incumbents; a similar approach has prevailed more recently (Baumol et al.,<br />
1982), so that we can talk <strong>of</strong> barriers to entry as sunk costs <strong>of</strong> entry for the<br />
competitors which are above the corresponding costs <strong>of</strong> the incumbent (or<br />
have been already paid by the incumbent). According to this definition, sunk<br />
costs can be binding on the entry decision <strong>of</strong> the followers, therefore, they<br />
can be a crucial determinant <strong>of</strong> the endogeneity <strong>of</strong> entry in a market. 10 A<br />
final category is that <strong>of</strong> the fixed costs <strong>of</strong> entry: these are equally faced by<br />
the incumbent <strong>and</strong> the followers to produce in the market, but they can also<br />
represent a binding constraint on entry. While there is a fundamental difference<br />
in the concepts <strong>of</strong> sunk costs <strong>and</strong> fixed costs <strong>of</strong> entry, their role in<br />
endogenizing entry is virtually the same, <strong>and</strong> we will not stress the difference<br />
in what follows. 11<br />
5.2.1 <strong>Competition</strong> in the <strong>Market</strong> <strong>and</strong> Policy Implications<br />
The main point emerging from our analysis <strong>of</strong> the behavior <strong>of</strong> market leaders<br />
facing or not facing endogenous entry is that st<strong>and</strong>ard measures <strong>of</strong> the concentration<br />
<strong>of</strong> a market have no relation to the market power <strong>of</strong> the leaders <strong>of</strong><br />
a market, <strong>and</strong> may lead to misleading welfare comparisons.<br />
10 There is a relation between the theory <strong>of</strong> market leaders <strong>and</strong> the “bounds approach”<br />
by Sutton (1998, 2005). His approach is largely based on the concept <strong>of</strong><br />
endogenous sunk costs as strategic investments - see Etro (2006,b) <strong>and</strong> Chapter 1<br />
for an attempt to endogenize sunk costs in the theory <strong>of</strong> market leaders. However,<br />
his focus is more on explaining market concentration rather than the strategies<br />
<strong>of</strong> market leaders. The two approaches could be seen as complementary.<br />
11 Another important aspect is about the source <strong>of</strong> these barriers <strong>and</strong> costs. As<br />
we noticed before, they can constitute a source <strong>of</strong> antitrust examination if they<br />
have been artificially created or enlarged by the incumbent; they cannot if their<br />
source is purely technological. Nevertheless, it is hard to imagine how artificial<br />
barriers could be erected under normal circumstances. The Chicago school is<br />
quite clear on this point, as we can conclude from the following position <strong>of</strong> Bork<br />
(1993): “If everything that makes entry more difficult is viewed as a barrier, <strong>and</strong><br />
if barriers are bad, then efficiency is evil. That conclusion is inconsistent with<br />
consumer-oriented policy. What must be proved to exist, therefore, is a class <strong>of</strong><br />
barriers that do not reflect superior efficiency <strong>and</strong> can be erected by firms to<br />
inhibit rivals. I think it clear that no such class <strong>of</strong> artificial barriers exists.”
180 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
<strong>Competition</strong> in Quantities. Theirrelevance<strong>of</strong>marketsharesfortheevaluation<br />
<strong>of</strong> the market power <strong>of</strong> leaders emerges quite clearly in the simplest<br />
environment we studied, that <strong>of</strong> competition in quantities with homogenous<br />
goods, constant marginal costs <strong>and</strong> a fixed cost <strong>of</strong> production. Such a simple<br />
structure approximates the situation <strong>of</strong> many sectors where product differentiation<br />
is not very important but there are high costs to starting production<br />
(as in many high-tech sectors). In such markets the characterization <strong>of</strong> the<br />
equilibrium structure is drastically different when entry conditions change.<br />
First <strong>of</strong> all, as long as the number <strong>of</strong> firms is exogenously given <strong>and</strong> the fixed<br />
costs <strong>of</strong> production are not too high, the leader is aggressive but leaves space<br />
for the followers to be active in the market. As external observers, we would<br />
look at this as a market characterized by an incumbent with a market share<br />
typically larger than its rivals, but with a certain number <strong>of</strong> competitors<br />
whose supply <strong>of</strong> goods reduces the equilibrium market price. The higher the<br />
number <strong>of</strong> these competitors, the lower the price will be: in such a case, lower<br />
concentration would be correctly associated with higher welfare.<br />
Radical changes occur when entry in the market is endogenous, <strong>and</strong> is determined<br />
by the existence <strong>of</strong> pr<strong>of</strong>itableopportunitiesinthesamemarket.In<br />
such a case (as we have seen in Section 1.1) the leader would exp<strong>and</strong> production<br />
until no one <strong>of</strong> the potential entrants has incentives to supply its goods<br />
in the market. The intuition for this extremely aggressive behavior <strong>of</strong> the<br />
market leader is simple. When entry is endogenous, the leader underst<strong>and</strong>s<br />
that a low production creates a large space for entry in the market while a<br />
high production reduces entry opportunities. More precisely, knowing how<br />
technological constraints govern the incentives to enter in the industry, the<br />
leader is aware that its output exactly crowds out the output <strong>of</strong> the competitors<br />
leaving unchanged the aggregate supply <strong>and</strong> hence the equilibrium price.<br />
However, taking this equilibrium price for the market as given, the leader can<br />
increase its pr<strong>of</strong>its by increasing its output <strong>and</strong> reducing the average costs<br />
<strong>of</strong> production. Here the fixed costs <strong>of</strong> production (associated with constant<br />
marginal costs) are crucial: on one side they constrain the pr<strong>of</strong>itability <strong>of</strong><br />
entry, while on the other side they create scale economies in the production<br />
process that can be exploited by the leader through an expansion <strong>of</strong> its output.<br />
Actually, it is always optimal for the leader to produce enough to crowd<br />
out all output by the competitors: exploiting the economies <strong>of</strong> scale over the<br />
entire market allows the leader to enjoy positive pr<strong>of</strong>its even if no entrant<br />
couldobtainpositivepr<strong>of</strong>its in this market. As external observers, in this<br />
case, we would simply see a single firm obtaining positive pr<strong>of</strong>its in a market<br />
where no one else enters, <strong>and</strong>, following traditional paradigms, we would<br />
associate this situation with a monopolistic environment, or at least with a<br />
dominant position derived by some barriers to entry. But this association is<br />
not correct, since entry is indeed free in this market: it is the competitive<br />
pressure <strong>of</strong> the potential entrants that induces the leader to produce so much<br />
to drive down the equilibrium price until no other firm can enter. We are
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous Entry 181<br />
referring to firms that are as efficient as the leader (assuming identical cost<br />
technologies). Finally, in Chapter 1, we even noticed that this equilibrium<br />
with only the leader in the market is associated with a higher welfare than<br />
the free entry equilibrium without a leadership - the Marshall equilibrium,<br />
which involves many firms active in the market <strong>and</strong> earning zero pr<strong>of</strong>its.<br />
Let us now consider a related situation with a different cost pattern for<br />
the firms (see Section 1.2.1). When marginal costs are substantially increasing<br />
in the production level or, more generally, when the average costs have a U-<br />
shape, a market leader facing endogenous entry <strong>of</strong> competitors may not have<br />
incentives to deter entry, but would still behave in an aggressive way. In such<br />
a case, given the strategy <strong>of</strong> the leader, all the entrants maximize their own<br />
pr<strong>of</strong>its <strong>and</strong> therefore they price above the marginal cost. However, endogenous<br />
entry reduces the equilibrium price to a level that is just high enough to<br />
cover the fixed costs <strong>of</strong> production. Notice that this equilibrium generates<br />
a production below the efficient scale (which should equate marginal <strong>and</strong><br />
average costs). Also in this case, the leader takes into account these elements<br />
<strong>and</strong>, in particular, takes as given the equilibrium price emerging from the<br />
endogenous entry <strong>of</strong> the competitors. Accordingly, the leader finds it optimal<br />
to produce as much to equate its marginal cost to the price, which requires<br />
a production above the efficient scale. Since marginal costs are increasing for<br />
such a high production level, the leader is pricing above its average cost, <strong>and</strong><br />
hence obtains positive pr<strong>of</strong>its. In this case the strategy <strong>of</strong> the leader does not<br />
even affect the market price, which is fully determined by endogenous entry<br />
<strong>of</strong> firms. Nevertheless, the leader obtains a larger market share than its rivals<br />
<strong>and</strong> positive pr<strong>of</strong>its. Moreover, we have shown that the aggressive behavior<br />
<strong>of</strong> the leader, that adopts a price equal to the marginal cost, improves the<br />
allocation <strong>of</strong> resources compared to the same market with free entry <strong>and</strong> no<br />
leadership. 12<br />
A similar situation emerges when goods are not homogeneous but differ in<br />
quality (see Section 1.2.2). This happens when consumer needs or tastes are<br />
quite differentiated, as is the case in many sectors where the design <strong>and</strong> the<br />
inner quality <strong>of</strong> products play an important role. Under these circumstances,<br />
firms <strong>of</strong>ten compete in prices by choosing different mark-ups for different<br />
products. When quality differs, it is important to have a number <strong>of</strong> firms<br />
producing different varieties <strong>of</strong> goods. A competitive market typically satisfies<br />
12 As we saw in the general model <strong>of</strong> Chapter 3, under Stackelberg competition<br />
with endogenous entry, the followers equate the price to their average total cost<br />
following the st<strong>and</strong>ard mark-up rule, <strong>and</strong> the leader equates the price to its own<br />
marginal cost:<br />
p =<br />
c0 (q)<br />
1 − 1/ = c(q)+F<br />
q<br />
= c 0 (q L)<br />
where is the elasticity <strong>of</strong> dem<strong>and</strong>. It follows that the equilibrium output <strong>of</strong> the<br />
leader, q L , is always higher than the one <strong>of</strong> the followers, q.
182 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
this requirement, but it tends to induce excessive proliferation <strong>of</strong> products.<br />
The presence <strong>of</strong> market leaders is again beneficial: they will not conquer the<br />
entire market, but they will exp<strong>and</strong> production <strong>and</strong> consequently reduce their<br />
prices below the prices <strong>of</strong> their competitors, some <strong>of</strong> which will be driven<br />
out <strong>of</strong> the market. Consumers will then face a lower variety <strong>of</strong> alternative<br />
products, but pay less for some <strong>of</strong> them. 13<br />
The crucial lesson from this analysis is that we should be careful in drawing<br />
any conclusion from indexes <strong>of</strong> concentration or from the market shares.<br />
We have seen examples in which the equilibrium outcome <strong>of</strong> a market with<br />
free entry was characterized by a single active firm enjoying positive pr<strong>of</strong>its,<br />
<strong>and</strong> other examples where the outcome was less drastic, but not too different.<br />
Notice that in all these cases, the market leader was adopting aggressive<br />
strategies which were reducing entry but increasing welfare nevertheless. It<br />
is important to emphasize that strategies that are aimed at reducing entry<br />
are not necessarily negative for consumers, especially when entry is not fully<br />
deterred, but simply limited due to a low level <strong>of</strong> the prices, so that some<br />
competitors are still active in the market <strong>and</strong> able to exert a competitive<br />
pressure on the leader. As a matter <strong>of</strong> fact, this is a good example <strong>of</strong> how<br />
real competition works.<br />
Of course, a predatory behavior can still be associated with aggressive<br />
strategies aimed at foreclosure <strong>and</strong> with negative consequences on consumers.<br />
This can be the case under two circumstances: 1) when these strategies are<br />
implemented by leaders with genuine market power which is not constrained<br />
by effective entry, <strong>and</strong> 2) when the same leader has built barriers to artificially<br />
constrain entry without efficiency reasons (see the Appendix in Chapter 1). 14<br />
Finally, notice that a complete analysis <strong>of</strong> the consequences <strong>of</strong> entry deterrence<br />
would require a dynamic model taking into account the behavior <strong>of</strong> the<br />
13 Consider a generalization <strong>of</strong> the examples <strong>of</strong> Chapter 1 with both product differentiation<br />
<strong>and</strong> U-shaped costs. Under the pr<strong>of</strong>it function:<br />
π i = q i (a − q i − b j6=i q j) − cq i − dq 2 i /2 − F<br />
a Stackelberg equilibrium with endogenous entry is characterized by followers<br />
producing:<br />
<br />
2F<br />
q =<br />
2+d<br />
<strong>and</strong> a leader producing:<br />
q L = 2 − b + d<br />
<br />
2F<br />
2 − 2b + d 2+d<br />
14 Artifical barriers associated with bureaucracy <strong>and</strong> lobbying activity on government<br />
processes are emphasized by the Chicago school as well (see Bork, 1993,<br />
Ch. 18).
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous Entry 183<br />
leader before <strong>and</strong> after deterrence, which is beyond the scope <strong>of</strong> this book.<br />
However, simple models can reveal a lot. Our point here is simply to warn<br />
against the risk <strong>of</strong> directly associating aggressive price strategies that reduce<br />
entry with strategies that harm consumers.<br />
<strong>Competition</strong>inPrices.Another important implication <strong>of</strong> the theory <strong>of</strong><br />
market leaders emerges quite clearly under competition in prices. In this typical<br />
situation, the traditional analysis <strong>of</strong> Stackelberg oligopolies shows that<br />
dominant firms are either accommodating (setting high prices) or trying to<br />
exclude rivals by setting low enough prices: the first case happens when the<br />
fixed costs <strong>of</strong> entry are small (<strong>and</strong> predation would be too costly), the second<br />
when they are high enough. 15 Such an outcome implies the risk <strong>of</strong> erroneously<br />
associating an aggressive price strategy with an entry deterring strategy in<br />
a systematic fashion. As we have seen (in Section 1.3), when we endogenize<br />
entry in the market, leaders never adopt accommodating pricing strategies<br />
while they are always aggressive. Again, in equilibrium with endogenous entry,<br />
leaders increase their market shares <strong>and</strong> obtain positive pr<strong>of</strong>its. Of course<br />
an aggressive pricing strategy will still reduce entry, even if it will not exclude<br />
all rivals. Nevertheless, we must be more careful in associating aggressive<br />
pricing with predatory purposes. The reason why predatory strategies are<br />
anti-competitive is that they exclude competition in the future allowing the<br />
dominant firm to behave in a monopolistic fashion once competitors are out<br />
<strong>of</strong> the market. Clearly, if an aggressive pricing strategy is aimed at excluding<br />
some but not all competitors, this anti-competitive element is more limited.<br />
Notice that competition in prices is quite typical <strong>of</strong> markets where product<br />
differentiation is relevant <strong>and</strong> firms have more autonomy in choosing their<br />
prices directly. The results are also relevant in oligopolistic markets in which<br />
prices determine the volume <strong>of</strong> business, as in the banking sector, where the<br />
interest rates on loans determine how much firms borrow from a bank, <strong>and</strong> the<br />
interest rates on deposits determine how much households lend to a bank. 16<br />
15 Accommodating high prices are chosen by the leader when fixed costs <strong>of</strong> entry<br />
are small. The problem is that this is exactly when there are incentives for other<br />
firms to enter, hence the duopolistic equilibrium is quite weak, <strong>and</strong> the study <strong>of</strong><br />
endogenous entry becomes crucial.<br />
16 Price competition is typical <strong>of</strong> the banking sector, where banks choose both the<br />
interest rates on loans, ι, <strong>and</strong> on deposits, r (see Freixas <strong>and</strong> Rochet, 1997). When<br />
entry in the sector is exogenous we would expect leaders to <strong>of</strong>fer worse terms<br />
to their customers, when it is endogenous we would expect the opposite. For<br />
instance, imagine that banks compete only on the deposit side, <strong>and</strong> the supply<br />
<strong>of</strong> deposit for firm i is S(r i,<br />
j6=i<br />
h(rj)) with S1 > 0 <strong>and</strong> S2 < 0. Adopting the<br />
usual notation, pr<strong>of</strong>its for bank i are:<br />
π i =(ι − r i − c)S (r i,β i ) − F<br />
where c is the marginal cost <strong>of</strong> intermediation. A Stackelberg equilibrium with<br />
endogenous entry is characterized by the optimality <strong>and</strong> free entry conditions for
184 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
Strategic Commitments. In general, as we have seen in Chapter 2, the<br />
spirit <strong>of</strong> our result on the aggressive behavior <strong>of</strong> leaders goes through when<br />
leaders cannot commit to output or price strategies, but can undertake preliminary<br />
investments that change their incentives in the market. For instance,<br />
a market leader facing an exogenous number <strong>of</strong> competitors may want to<br />
underinvest or overinvest strategically in cost reducing R&D according to<br />
the kind <strong>of</strong> competition (in prices or in quantities), because it may want to<br />
commit through these investments to adopt an accommodating or an aggressive<br />
strategy in the market: in particular, underinvesting is optimal before<br />
price competition, while overinvesting is optimal before quantity competition.<br />
However, this ambiguity collapses if the leader is facing endogenous entry <strong>of</strong><br />
competitors. In such a case, it is always optimal to adopt the strategy that<br />
allows one to be aggressive in the market: strategic overinvestment in cost<br />
reducing R&D is optimal independently from the form <strong>of</strong> competition, because<br />
it allows one to be aggressive against competitors (see Section 2.6). 17<br />
A similar role is attached to investment in production capacity, to debt as a<br />
financing tool issued to commit management to produce higher output, <strong>and</strong><br />
to many other strategic investments.<br />
An interesting situation for antitrust purposes emerges when dem<strong>and</strong> is<br />
characterized by network effects. In such a case, market leaders tend to underprice<br />
their products initially to attract customers in the future. As known,<br />
these strategies may induce pricing below marginal cost without entry deterrence<br />
purposes. Moreover, in Section 2.9 we have seen that leaders facing<br />
endogenous entry may have further strategic incentives to reduce initial prices<br />
(or exp<strong>and</strong> initial production): by doing so, they enhance network externalities<br />
<strong>and</strong> are able to reduce their prices also in the future. Therefore, antitrust<br />
authorities should be careful in evaluating aggressive pricing in the presence<br />
the followers:<br />
ι − r − c =<br />
S(r, β)<br />
S 1 (r, β) = F<br />
S(r, β)<br />
<strong>and</strong> by the optimality condition for the leader:<br />
ι − r L − c =<br />
S(r L ,β L )<br />
S 1 (r L ,β L ) − S 2 (r L ,β L )h 0 (r L )<br />
which implies r L >r. Only under the pressure <strong>of</strong> free entry, the leader affords<br />
to compensate deposits with a high rate than its followers.<br />
17 Both effective <strong>and</strong> potential competition are crucial here. On this point, we are<br />
close to early informal insights <strong>of</strong> the Chicago school. For instance, Posner (2001,<br />
p. 145) notices that “notions <strong>of</strong> potential competition cannot <strong>and</strong> should not be<br />
banished entirely from antitrust law... a monopolist who creates excess capacity<br />
inordertoreducehismarginalcost,sothatentrants(whohavetobeableto<br />
cover their average total cost if they are to make a go <strong>of</strong> entry) are deterred, is<br />
reacting to potential competition.”
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous Entry 185<br />
<strong>of</strong> network effects. Finally, this point applies in particular to multi-sided markets,<br />
where network effects take place between different kinds <strong>of</strong> customers,<br />
<strong>and</strong> firms can charge their different customers differently. In such an environment<br />
market leaders tend to price quite aggressively one <strong>of</strong> the sides, but<br />
again without exclusionary purposes. We will return to this point with more<br />
details in the next chapter.<br />
The same care in judging aggressive strategies is needed in cases <strong>of</strong> complementary<br />
strategies that virtually induce aggressive behaviors. One <strong>of</strong> these<br />
is bundling. In an influential paper, Whinston (1990) has studied bundling<br />
in a market with two goods. The primary good is monopolized by one firm,<br />
which competes with a single rival in the market for the secondary good. Under<br />
price competition in the secondary market, the monopolist becomes more<br />
aggressive in its price choice in the case <strong>of</strong> bundling <strong>of</strong> its two goods. Since a<br />
more aggressive strategy leads to lower prices for both firmsaslongasboth<br />
are producing, the only reason why the monopolist may want to bundle its<br />
two goods is to deter entry <strong>of</strong> the rival in the secondary market. This conclusion<br />
can be highly misleading because it neglects the possibility <strong>of</strong> further<br />
entry in the market. As we have seen in Section 2.10, if the secondary market<br />
is characterized by endogenous entry, the monopolist would always like to be<br />
aggressive in this market <strong>and</strong> bundling may be the right way to commit to an<br />
aggressive strategy. Bundling would not necessarily deter entry in this case,<br />
especially if there is a high degree <strong>of</strong> product differentiation in the secondary<br />
market, but may increase competition in this market <strong>and</strong> reduce prices with<br />
positive effects on the consumers. 18<br />
Another application <strong>of</strong> the theory <strong>of</strong> market leaders concerns vertical restraints<br />
affecting inter-br<strong>and</strong> competition (Bonanno <strong>and</strong> Vickers, 1988; Rey<br />
<strong>and</strong> Stiglitz, 1988). Also in this case, the behavior <strong>of</strong> the market leader can be<br />
anticompetitive depending on the entry conditions. In particular, under price<br />
competition, a contract delegating distribution to a downstream firm tends<br />
to s<strong>of</strong>ten price competition when entry in the market is exogenous (because<br />
the upstream firm imposes high prices through direct or indirect contractual<br />
restraints), but it strengthens price competition when entry is endogenous<br />
(in which case the upstream firm can only gain by inducing an aggressive<br />
behavior <strong>of</strong> the downstream firm): the consequences on consumers tend to be<br />
negative in the former case <strong>and</strong> positive in the latter case (Section 2.11).<br />
We encounter a more complex situation when we consider price discrimination<br />
versus uniform pricing, since they can both s<strong>of</strong>ten or strengthen price<br />
competition in different markets. However, we have shown an example where,<br />
when price discrimination emerges between two groups <strong>of</strong> customers, it is also<br />
likely to s<strong>of</strong>ten price competition compared to uniform pricing (Section 2.12).<br />
If this is the case, price discrimination is adopted by a firm competing with<br />
18 Notice that the same limit <strong>of</strong> the analysis <strong>of</strong> Whinston, namely the exogenous<br />
assumption that there are just two firms <strong>and</strong> further endogenous entry is not<br />
taken into account, applies to many other duopolistic models <strong>of</strong> bundling.
186 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
an exogenous number <strong>of</strong> competitors, but not when entry is endogenous. Accordingly,<br />
when it takes place price discrimination is likely to harm at least<br />
some consumers.<br />
Conclusion. Our final remark is about the presence <strong>of</strong> high <strong>and</strong> sustained<br />
pr<strong>of</strong>its (sometimes called supernormal pr<strong>of</strong>its) <strong>of</strong> the market leaders. These<br />
high pr<strong>of</strong>its are frequently, but sometimes erroneously, associated by the traditional<br />
approaches with a situation <strong>of</strong> market dominance <strong>and</strong> barriers to<br />
entry that prevents competition from driving down the rate <strong>of</strong> return to its<br />
competitive level. As we have seen repeatedly, even in the presence <strong>of</strong> free<br />
entry market leaders with a first mover advantage or with a more reasonable<br />
chance to undertake preliminary investments are able to obtain positive<br />
pr<strong>of</strong>its, or, in other words, they are able to preserve a rate <strong>of</strong> return above<br />
the opportunity cost <strong>of</strong> capital. Evidently, these sustained pr<strong>of</strong>its in a market<br />
where entry is free are not a symptom <strong>of</strong> dominance per se. Notice that there<br />
are other important reasons why sustained pr<strong>of</strong>its may persist in innovative<br />
markets, but we will look at them in the next section on the competition for<br />
the market <strong>and</strong> its implications.<br />
The bottom line <strong>of</strong> this discussion on competition in the market is that in<br />
evaluating market structures <strong>and</strong> the behavior <strong>of</strong> market leaders we should<br />
give special attention to the entry conditions. St<strong>and</strong>ard results on aggressive<br />
price <strong>and</strong> non-price strategies with exclusionary purposes emerging for markets<br />
with an incumbent <strong>and</strong> an entrant can change in radical ways when we<br />
take in consideration the possibility <strong>of</strong> endogenous entry by other firms. After<br />
all, antitrust policy in an uncertain world should derive from a comparison <strong>of</strong><br />
the expected losses from incorrectly challenging a practice that benefits consumers<br />
(a Type I error ) versus the expected losses from incorrectly failing to<br />
challenge a practice that harms consumers (a Type II error ). We believe that<br />
while the Chicago School has been extremely biased to reduce the first kind<br />
<strong>of</strong> losses (exactly because it largely ignored strategic interactions), the post-<br />
Chicago approach has been excessively biased in the opposite sense (exactly<br />
because it <strong>of</strong>ten neglected endogenous entry). 19<br />
5.2.2 <strong>Competition</strong> for the <strong>Market</strong> <strong>and</strong> Policy Implications<br />
<strong>Competition</strong> in high-tech markets is dynamic in the Schumpeterian sense that<br />
it takes place as competition for the market in a so-called winner-takes-allrace,<br />
<strong>and</strong> such an element requires a deeper evaluation <strong>of</strong> competition policy<br />
than that suggested in the analysis <strong>of</strong> the previous section, which was mostly<br />
focused on a static concept <strong>of</strong> competition in the market. 20<br />
19 On the design <strong>of</strong> optimal procedures for competition policy see Katsoulacos <strong>and</strong><br />
Ulph (2007).<br />
20 In the terminology <strong>of</strong> the famous growth-share matrix popularized by the Boston<br />
Consulting Group, new products in the growing high-tech markets are Stars (for
5.2 The <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong> Endogenous Entry 187<br />
Economic research has emphasized the positive relationship linking patents<br />
to investments in innovation <strong>and</strong> these investments to technological progress<br />
<strong>and</strong> growth. In high-tech sectors (hardware, s<strong>of</strong>tware, pharmaceuticals, biotechnology)<br />
firms compete mainly by innovating. This is possible as long as there<br />
are well defined intellectual property rights (IPRs), <strong>and</strong> especially patents,<br />
protecting their innovations <strong>and</strong> investments, which is ultimately what leads<br />
to technological progress in our economies. In Section 1.4 we have even suggested<br />
that this form <strong>of</strong> dynamic competition can be a valid substitute for<br />
the competition in the market: when entry in the competition for the market<br />
is free, an increase in the degree <strong>of</strong> competition in the market cannot provide<br />
further incentives to invest in innovation (because any escape competition<br />
effect disappears).<br />
Moreover, even if most economists are used to thinking about market<br />
leaders as firms with weaker incentives to invest in R&D, recent theoretical<br />
<strong>and</strong> empirical research has also found support for an old idea associated with<br />
the institutional work <strong>of</strong> Schumpeter (1943), Galbraith (1952) <strong>and</strong> Ch<strong>and</strong>ler<br />
(1990), according to which market leaders play a crucial role in the innovative<br />
activity. The theory <strong>of</strong> market leaders has clarified the mechanics <strong>of</strong><br />
these results. IPRs drive competition through innovation in these markets<br />
<strong>and</strong> induce technological progress led by incumbent monopolists under two<br />
conditions: their leadership in the contest to innovate <strong>and</strong> free entry <strong>of</strong> outsiders<br />
in this same contest. In particular, in Chapter 4 we contrasted two<br />
scenarios. According to traditional theories, in the absence <strong>of</strong> strategic advantages,<br />
a technological leader that is also an incumbent monopolist in its<br />
market, would have less incentives to invest in R&D compared to other firms,<br />
since its relative gain from improving its own technology is smaller than the<br />
gain <strong>of</strong> the outsiders from replacing the incumbent monopolist. This result,<br />
sometimes called the Arrow’s paradox, has <strong>of</strong>ten been used to suggest that<br />
incumbent patent-holders invest less than other firms <strong>and</strong> stifle innovation.<br />
However, we have also seen that when an incumbent monopolist is the leader<br />
in the contest for innovating, the pressure<strong>of</strong>acompetitivefringe<strong>of</strong>potential<br />
innovators leads this monopolist to invest more than any other firm. The<br />
competitive environment spurs investment by leaders <strong>and</strong> consequently induces<br />
a chance that their leadership persists. Finally, we have also suggested<br />
that when the leadership persists because <strong>of</strong> the endogenous investment in<br />
R&D by the leaders, the same value <strong>of</strong> becoming a leader is increased, which<br />
strengthens even further the incentives for any other firm to invest, <strong>and</strong> so<br />
on. Paradoxically, the persistence <strong>of</strong> a leadership in high-tech sectors can be a<br />
sign <strong>of</strong> effective dynamic competition for the market, which leads to a faster<br />
rate <strong>of</strong> technological progress in the interest <strong>of</strong> consumers.<br />
market leaders) or Question Marks (for the followers), as opposed to the mature<br />
markets typical <strong>of</strong> the traditional sectors that can be characterized by products<br />
with high market shares (Cash Cows) or low market shares (Dogs).
188 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
Notice that our results on the relation between entry in the competition<br />
for the market <strong>and</strong> investment by the incumbent monopolists can be seen as<br />
strengthening our initial claim that st<strong>and</strong>ard indices <strong>of</strong> market concentration<br />
or market shares should not be related to the degree <strong>of</strong> competition in a<br />
market. In high-tech markets where competition is mostly for the market, it<br />
is natural that better products conquer large shares <strong>and</strong>, exactly when entry<br />
is free, incumbent patent-holders have more incentives to invest <strong>and</strong> their<br />
leadership is more likely to persist. There is no basis to relate in a significant<br />
way market shares <strong>and</strong> market power in dynamic sectors.<br />
This mechanism is even more radical in markets with network effects,<br />
where the natural outcome is a sequence <strong>of</strong> dominant paradigms associated<br />
with market leaders whose behavior is still constrained by innovative competitive<br />
pressure. Scotchmer (2004, p. 296), in her discussion <strong>of</strong> network effects,<br />
emphasizes this point neatly:<br />
“All this calls into question whether an incumbent’s share <strong>of</strong> a<br />
network market is a good test <strong>of</strong> market power for antitrust purposes.<br />
With tippy markets, any snapshot <strong>of</strong> the market will find some firm<br />
with a dominant market share. But sequential monopoly is only a<br />
problem for competition policy if the price charged by each sequential<br />
monopolist is high [...] the price is constrained at first by the<br />
proprietor’s need to attract users <strong>of</strong> the previous product, <strong>and</strong> later<br />
by a fear <strong>of</strong> scaring users into embracing a successor. The same fears<br />
will cause the incumbent to keep innovating.”<br />
Of course, we do not want to give the message that persistent monopolies<br />
are necessarily the fruit <strong>of</strong> effective competition for the market, but rather<br />
that they can be the fruit <strong>of</strong> effective competition. What we would like to<br />
emphasize is the importance <strong>of</strong> entry conditions in the market for innovations.<br />
This is in line with an old Chicago-style position associated with Demsetz<br />
(1973), who pointed out that: 21<br />
“Under the pressure <strong>of</strong> competitive rivalry, <strong>and</strong> in the apparent<br />
absence <strong>of</strong> effective barriers to entry, it would seem that the concentration<br />
<strong>of</strong> an industry’s output in a few firms could only derive from<br />
their superiority in producing <strong>and</strong> marketing products ... an industry<br />
will become more concentrated under competitive conditions only if a<br />
differential advantage in exp<strong>and</strong>ing output develops in some firms ...<br />
The cost advantage that gives rise to increased concentration may be<br />
reflected in scale economies or in downward shifts in positively sloped<br />
marginal cost curves, or it may be reflected in better products which<br />
satisfy dem<strong>and</strong> at a lower cost” (Demsetz, 1973).<br />
Consequently, industrial policy, including antitrust policy, should primarily<br />
promote, <strong>and</strong> possibly subsidize, investment in R&D, <strong>and</strong> it should be less<br />
21 See Hughes (2007) for a recent <strong>and</strong> successful test <strong>of</strong> the Demsetz hypothesis.
5.3 A Digression on IPRs Protection 189<br />
relevant whether the incumbent monopolist or new comers invest in R&D <strong>and</strong><br />
innovate once entry is free. On the other side, the protection <strong>of</strong> IPRs should<br />
be established at a legislative level (possibly even coordinated at an international<br />
level) because its stability is essential to foster investments, while<br />
the discretionary activity <strong>of</strong> antitrust authorities should not affect the basic<br />
principles <strong>of</strong> IPRs protection.<br />
The credibility <strong>of</strong> innovation policy is crucial to give incentives to firms to<br />
innovate, because investment in R&D depends mainly on the expectations on<br />
the protection <strong>of</strong> IPRs. This point is quite similar to st<strong>and</strong>ard results in monetary<br />
<strong>and</strong> fiscal policy. A commitment to low inflation is essential because price<br />
setting decisions are based on the expectations <strong>of</strong> inflation: surprise inflation<br />
may push the economy in the short run, but will just increase inflationinthe<br />
long run. A commitment to a capital income tax is essential because savings<br />
decisions are based on the expectations <strong>of</strong> capital income taxes: unexpected<br />
higher capital taxation can raise more tax revenue in the short run, but it<br />
will mainly reduce savings <strong>and</strong> tax revenue in the long run. Analogously, a<br />
commitment to a level <strong>of</strong> protection <strong>of</strong> IPRs is essential because investment<br />
decisions are based on the expectation <strong>of</strong> this protection: forcing disclosure<br />
<strong>of</strong> IPRs can have some positive effects for outsider firms in the short run, but<br />
will have devastating effects on innovation <strong>and</strong> growth in the long run.<br />
This leads us to another important aspect <strong>of</strong> industrial policy, the protection<br />
<strong>of</strong> IPRs, which has an old <strong>and</strong> well recognized tradition in developed<br />
market economies. 22<br />
5.3 A Digression on IPRs Protection<br />
To underst<strong>and</strong> the crucial role <strong>of</strong> IPRs <strong>and</strong> patents in promoting technological<br />
progress <strong>and</strong> growth we rely on an old important theory developed<br />
by Nordhaus (1969). In general, patents assign a temporary monopolistic<br />
power for the innovators which creates price distortions <strong>and</strong> hence carries a<br />
social cost, but also constitutes an incentive for many firms to invest <strong>and</strong><br />
try to gain market leadership through innovations. This effect leads to social<br />
benefits because innovations have a social value that can be higher than the<br />
22 Patent protection was recognized in Renaissance Italy. In 1474 the Republic <strong>of</strong><br />
Venice issued a decree by which new <strong>and</strong> inventive devices, once they had been<br />
put into practice, had to be communicated to the Republic in order to obtain<br />
legal protection against potential infringers. Galileo applied for a patent on an<br />
hydraulic system in 1593 noticing that “it does not suit me that the invention,<br />
which is my property <strong>and</strong> was created by me with great effort <strong>and</strong> cost, should<br />
become the common property <strong>of</strong> just anyone.” The Venetian Senate assigned<br />
him a patent valid for twenty years. For more on the history <strong>of</strong> innovations <strong>and</strong><br />
IPRs see Scotchmer (2004).
190 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
private value (<strong>and</strong> that cannot be fully appropriated by the innovators) 23 <strong>and</strong><br />
because they drive technological progress <strong>and</strong> growth. Clearly social benefits<br />
<strong>and</strong> costs can be different for different inventions <strong>and</strong> generally for different<br />
fields <strong>of</strong> technology, so, in theory, the optimal length <strong>and</strong> breadth <strong>of</strong> patents<br />
should depend on the field it applies to. For simplicity, <strong>and</strong> to avoid discriminations<br />
between fields <strong>of</strong> technology, patents typically have a uniform<br />
length <strong>and</strong> common principles on infringement regulation. Nevertheless, from<br />
a strictly economic point <strong>of</strong> view, one may question this uniformity <strong>and</strong> consider<br />
the advantages <strong>of</strong> providing different terms <strong>of</strong> protection in different<br />
sectors (at least this could avoid the inefficient choice <strong>of</strong> radically excluding<br />
certain innovations from patentability rather than allowing a more limited<br />
protection). More importantly, an evaluation <strong>of</strong> the social benefits <strong>and</strong> costs<br />
<strong>of</strong> patents for different fields is essential in judging the net benefit <strong>of</strong>apatent<br />
system. 24<br />
5.3.1 Patents in Dynamic Sectors <strong>and</strong> <strong>Innovation</strong>s<br />
Consider the pharmaceutical sector, where the role <strong>of</strong> patents on new drugs<br />
is, to say the least, at the basis <strong>of</strong> competition in the market <strong>and</strong> <strong>of</strong> scientific<br />
progress in the world. These kinds <strong>of</strong> patents have been <strong>of</strong>ten criticized<br />
for jeopardizing health defense around the world <strong>and</strong> especially in developing<br />
countries, where western drugs are very important but very expensive.<br />
Nevertheless, one should not forget that those same patents induced many<br />
firms to invest <strong>and</strong> some <strong>of</strong> them to invent new drugs which are now available,<br />
something which would have hardly happened or would have happened<br />
later without patent protection: in other words the social benefit <strong>of</strong>patents<br />
on drugs is very high. Fortunately, there are ways to reduce the problems<br />
related with the pricing <strong>of</strong> drugs <strong>and</strong> their adoption depends mostly on the<br />
public sector. For instance, governments could buy drugs <strong>and</strong> distribute them<br />
at lower prices through the medical system, or just pay part <strong>of</strong> the prices.<br />
They may even directly buy the same patents from the innovators <strong>and</strong> produce<br />
the drugs (or outsource their production) <strong>and</strong> sell them at lower prices.<br />
Finally, western governments could redirect their international aid toward<br />
similar initiatives in favor <strong>of</strong> developing countries. These solutions, widely<br />
discussed between economists, may preserve the proper incentives to invest<br />
<strong>and</strong> discover new drugs while spreading their effects globally. Ultimately, this<br />
suggests that patents in the pharmaceutical sectors are a crucial determinant<br />
<strong>of</strong> innovation <strong>and</strong> progress <strong>and</strong> should be protected while finding alternative<br />
solutions to guarantee health defense for poor classes <strong>and</strong> poor countries.<br />
23 For a “public choice” perspective on this <strong>and</strong> the relation with the theory <strong>of</strong><br />
market leaders see Reksulak et al. (2006).<br />
24 When the social value <strong>of</strong> patents is very high, public sponsorship <strong>of</strong> R&D activities<br />
<strong>and</strong> public-private partnership can be useful (on the latter experience see<br />
the empirical works by Baarsma et al., 2004, <strong>and</strong> Ambrosanio et al., 2004).
5.3 A Digression on IPRs Protection 191<br />
Another field in which patents are particularly valuable <strong>and</strong> induce high<br />
investments in R&D is the New Economy. In the last few years the European<br />
Union tried (without success) to complete a process <strong>of</strong> harmonization<br />
<strong>of</strong> the patent system for computer-implemented inventions (CIIs). 25 We believe<br />
that the rationale for these patents is strong: while their main social<br />
gain is to promote innovation in the most dynamic sectors, the social cost<br />
is smaller than for other patents since in these sectors competition mainly<br />
works through frequent price-reducing <strong>and</strong> quality-improving innovations,<br />
therefore price distortions are less relevant <strong>and</strong> do not last long anyway.<br />
Neglecting these traditional economic insights, opponents <strong>of</strong> the patent system<br />
have <strong>of</strong>ten claimed that patents stifle innovation. There is not, however,<br />
consistent theory or empirical evidence behind these claims. In US, the extension<br />
<strong>of</strong> patent protection to CIIs started in 1980 (the firstpatent<strong>of</strong>this<br />
kind was granted by the US Patent <strong>and</strong> Trademark Office in 1981), <strong>and</strong> it<br />
was associated with a clear increase in R&D investment during the eighties.<br />
The R&D/sales ratio for US firms innovating on computer, telecommunications<br />
<strong>and</strong> electronic components (the relevant field here) increased from 5.5%<br />
to above 8% in 1989 (see Etro, 2007c). In a careful empirical study Mann<br />
(2005) has shown that patents bestow significant benefits, especially for start<br />
up companies, in terms <strong>of</strong> traditional appropriability, information signalling<br />
<strong>and</strong> cross-licensing revenue, while Merges (2006), looking at patent data in<br />
the US s<strong>of</strong>tware market, finds out that “new firms entry remains robust, despite<br />
the presence <strong>of</strong> patents (<strong>and</strong>, in some cases, perhaps because <strong>of</strong> them).<br />
Successful incumbent firms have adjusted to the advent <strong>of</strong> patents by learning<br />
to put a reasonable amount <strong>of</strong> effort into the acquisition <strong>of</strong> patents <strong>and</strong> the<br />
building <strong>of</strong> patent portfolios. Patent data on incumbent firms shows that several<br />
well-accepted measures <strong>of</strong> ‘patent effort’ correlate closely with indicators<br />
<strong>of</strong> market success such as revenue <strong>and</strong> employee growth.” 26<br />
5.3.2 Open-Source <strong>Innovation</strong>s<br />
While IPRs are fundamental drivers <strong>of</strong> innovation in all sectors, s<strong>of</strong>tware development<br />
has recently been characterized by a large amount <strong>of</strong> innovation<br />
25 After a long procedure, the Common Position adopted by the European Council<br />
in March 2005 proposed the patentability <strong>of</strong> CIIs when they provide a technical<br />
contribution to a field <strong>of</strong> technology. While this positive proposal simply reaffirmed<br />
the requirements already adopted in Europefortheprevioustwodecades<br />
<strong>and</strong> it excluded from patentability any pure s<strong>of</strong>tware, business methods <strong>and</strong> consulting<br />
practices (which are patentable in US), the European Parliament ended<br />
up rejecting the Directive in July 2005. See Etro (2005a,b).<br />
26 See Bessen <strong>and</strong> Maskin (2002) for a theoretical <strong>and</strong> empirical position against<br />
s<strong>of</strong>tware patents, <strong>and</strong> Etro (2007c) for a critical view <strong>of</strong> their theoretical <strong>and</strong><br />
empirical results.
192 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
obtained in a decentralized, voluntary <strong>and</strong> uncompensated way by programmers<br />
within the so-called open source movement. 27 While many private corporations<br />
support it because they supply products that are complementary<br />
to open source s<strong>of</strong>tware (IBM first <strong>of</strong> all, but also Hewlett-Packard, Intel,<br />
Sun, Oracle,...), it remains surprising that such a large innovative process<br />
can take place, at least in part, through directly unrewarded efforts. 28 Lerner<br />
<strong>and</strong> Tirole (2002, 2006) have provided a few explanations for the incentives<br />
<strong>of</strong> these individual programmers: career concern, ego gratification <strong>and</strong> signalling<br />
activity are quite powerful <strong>and</strong> effective in this field. Unfortunately,<br />
the same nature <strong>of</strong> these incentives shows the possible limitations <strong>of</strong> the innovative<br />
activity in the open source community: it is limited by the usual<br />
free riding problems emerging in the private provision <strong>of</strong> public goods, it<br />
requires a complementary activity in the for-pr<strong>of</strong>it sector (to motivate the<br />
career concern <strong>and</strong> the signalling activity), it may be biased by research efforts<br />
that are different from general consumer needs <strong>and</strong> by adverse selection<br />
<strong>of</strong> the contributors, <strong>and</strong> it may be effective to solve a number <strong>of</strong> small <strong>and</strong><br />
short term problems, but less effective to solve multi-sided challenges <strong>and</strong><br />
approach long term projects. 29 While the development <strong>of</strong> this new form <strong>of</strong><br />
user-driven innovation is a symptom <strong>of</strong> high competitive pressure in the sec-<br />
27 Open source s<strong>of</strong>tware is made available for direct use <strong>and</strong> modification (through<br />
direct access to the source code) under limited protection. For instance, the GPL<br />
(General Public License, first used in 1981 by Richard Stallman, the leader <strong>of</strong> the<br />
Free S<strong>of</strong>tware Movement) grants unlimited right to use, modify <strong>and</strong> distribute<br />
s<strong>of</strong>tware as long as its redistribution makes available the modified source code<br />
<strong>and</strong> does not impose further restrictions on the rights granted by the GPL.<br />
These enforcement mechanisms make cooperative innovation quite effective <strong>and</strong><br />
immune from free riding, but can create problems when an innovation includes<br />
both open source s<strong>of</strong>tware <strong>and</strong> licensed proprietary s<strong>of</strong>tware.<br />
28 Major results <strong>of</strong> the open source movement are Linux, an operating system based<br />
on Unix (an old OS first created at Bell Labs) <strong>and</strong> developed in 1991 by Linus<br />
Torvalds, Apache, a world wide web (HTTP) server, <strong>and</strong> Mozilla Firefox, a web<br />
browser. Besides s<strong>of</strong>tware that is freely distributed, there is an increasing number<br />
<strong>of</strong> companies, like Red Hat <strong>and</strong> Novell, that pr<strong>of</strong>it from collateral services supplied<br />
jointly with free s<strong>of</strong>tware. In theory, any rival could resell Red Hat s<strong>of</strong>tware<br />
at a lower price because it is under GPL (<strong>and</strong> some firmsactuallydoit),butRed<br />
Hat managed to sidestep this problem protecting its products with trademark<br />
law. In this sense, the difference between proprietary s<strong>of</strong>tware <strong>and</strong> open source<br />
s<strong>of</strong>tware appears much less relevant: the former earns from licenses to end-users,<br />
the latter mainly licenses s<strong>of</strong>tware free <strong>of</strong> charge <strong>and</strong> earns from selling support<br />
description needed by end-users to install <strong>and</strong> run the s<strong>of</strong>tware. For empirical<br />
evidence on the open source development see Koski (2007).<br />
29 It is <strong>of</strong>ten claimed that open source s<strong>of</strong>tware is more effective than proprietary<br />
s<strong>of</strong>tware in debugging activity (since many programmers find <strong>and</strong> solve many<br />
defects within a s<strong>of</strong>tware <strong>and</strong> make the solutions freely available), but may have
5.3 A Digression on IPRs Protection 193<br />
tor, it does not provide any evidence against the fundamental role <strong>of</strong> IPRs in<br />
driving core innovations. 30 Actually, we believe that the current coexistence<br />
<strong>of</strong> open source s<strong>of</strong>tware <strong>and</strong> proprietary s<strong>of</strong>tware exerts a positive impact on<br />
innovation on both sides. 31<br />
In a fascinating work, Boldrin <strong>and</strong> Levine (2005) adopted open source s<strong>of</strong>tware<br />
as a main example <strong>of</strong> innovation created without commercialization <strong>of</strong><br />
IPRs, <strong>and</strong> collected some anecdotal evidence suggesting that innovations can<br />
perfectly take place in the absence <strong>of</strong> what they call “intellectual monopoly”.<br />
Their idea is that the first mover advantage <strong>of</strong> the innovator in the competition<br />
in the market preserves a certain amount <strong>of</strong> pr<strong>of</strong>its even when entry<br />
<strong>of</strong> imitators is free, <strong>and</strong> this Stackelberg advantage can be sufficient to probig<br />
problems confronting issues as synchronization <strong>of</strong> upgrades <strong>and</strong> efficient levels<br />
<strong>of</strong> backward compatibility.<br />
30 In a debate on the Financial Times, the author <strong>of</strong> this book expressed a related<br />
point: “It is true that the competitive pressure from open source s<strong>of</strong>tware has<br />
led technological leaders to continue investing in research <strong>and</strong> development, but<br />
major advances such as the iPhone or Micros<strong>of</strong>t Surface keep arriving from the<br />
commercial s<strong>of</strong>tware world. Moreover, restrictive open source s<strong>of</strong>tware creates a<br />
fundamental asymmetry. On one side, open source s<strong>of</strong>tware companies (allied<br />
with big business, such as IBM) can use proprietary s<strong>of</strong>tware within their products<br />
<strong>and</strong> freely distribute them while covering licence expenditures through customer<br />
support services. On the other side, commercial companies cannot pursue<br />
their business model when including open source within their s<strong>of</strong>tware, because<br />
they would infringe the "copyleft" if they apply a price to the licence. This asymmetry<br />
can create substantial problems for the conventional business model, <strong>and</strong><br />
may inhibit or bias consumer-driven innovation. Finally, the notion that European<br />
competitiveness vis a vis China will be enhanced by the promotion <strong>of</strong> the<br />
open source s<strong>of</strong>tware model is preposterous. The general public licence is built<br />
on the proposition that anything you do <strong>and</strong> distribute can be freely appropriated<br />
by anyone within or outside Europe, in fact h<strong>and</strong>ing over the result <strong>of</strong> your<br />
investment to China on a silver platter. Rather than democratising innovation,<br />
we should protect it.” (Etro, 2007,e).<br />
31 To see why, think <strong>of</strong> a different sort <strong>of</strong> open source activity: Wikipedia is a famous<br />
<strong>and</strong> successful online encyclopedia where anybody can post a new voice or edit<br />
an old one (www.wikipedia.org). While it contains a lot <strong>of</strong> useful <strong>and</strong> constantly<br />
updated information (especially in certain fields, like those related to the online<br />
community), it <strong>of</strong>ten includes unmotivated <strong>and</strong> misleading references or mistakes<br />
that are the normal consequences <strong>of</strong> overlapping additions by heterogeneous<br />
contributors whose preparation is not properly controlled <strong>and</strong> whose effort is not<br />
rewarded. Traditional encyclopedias based on rewarded contributions by selected<br />
experts are not constantly updated like Wikipedia, but they provide a st<strong>and</strong>ard<br />
<strong>of</strong> quality <strong>and</strong> a balanced unifying structure that Wikipedia lacks. The trade<strong>of</strong>f<br />
for the end users is clear, <strong>and</strong> coexistence appears natural. See Aghion <strong>and</strong><br />
Modica (2006).
194 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
mote innovation. Since a main leitmotif <strong>of</strong> our book has been showing that<br />
market leaders with a first mover advantage can obtain positive pr<strong>of</strong>its even<br />
in markets where entry is free by adopting aggressive strategies (Chapters<br />
2-3), <strong>and</strong> since our formalization <strong>of</strong> competition for the market was perfectly<br />
compatible with incentives deriving from the pr<strong>of</strong>its <strong>of</strong> a market leader facing<br />
free entry, rather than deriving from a monopolistic position (Chapter 4),<br />
we certainly do not dislike the idea <strong>of</strong> Boldrin <strong>and</strong> Levine. But the point is:<br />
how much innovation can be promoted by the simple first mover advantage?<br />
Without an analysis <strong>of</strong> the industrial organization <strong>of</strong> the competition for the<br />
market it is quite hard to answer this question, <strong>and</strong> the general equilibrium<br />
analysis <strong>of</strong> Boldrin <strong>and</strong> Levine (1998) concluding that perfectly competitive<br />
innovations can achieve the first best amount <strong>of</strong> innovation neglects, to a<br />
large extent, the industrial organization <strong>of</strong> the market for innovations. Unfortunately,<br />
whether the incentives to invest are efficient or not is more an<br />
empirical question than a theoretical one, <strong>and</strong> we still do not see a consistent<br />
piece <strong>of</strong> evidence showing that current patent systems provide excessive<br />
incentives in a systematic way, or that we should totally eliminate IPRs legalizing<br />
“intellectual expropriation”- as Boldrin <strong>and</strong> Levine (2005) actually<br />
suggest. As a matter <strong>of</strong> fact, the opposite may be true. Recently, Denicolò<br />
(2007) analyzed a general model <strong>of</strong> the organization <strong>of</strong> innovations <strong>and</strong> obtainedasimplerulefortheoptimallevel<br />
<strong>of</strong> patent protection: his empirical<br />
estimates suggest that current patent systems do not over-compensate innovators,<br />
while they may actually induce too limited incentives to invest in<br />
R&D. 32<br />
5.3.3 Conclusions on IPRs Protection<br />
Before returning to the core discussion on antitrust issues, we list a number<br />
<strong>of</strong> implications <strong>of</strong> the economic debate on IPR protection that in our opinion<br />
are quite relevant:<br />
1) the optimal patent system should trade-<strong>of</strong>f the social benefits <strong>of</strong> the<br />
incentives to innovate <strong>and</strong> the social costs due to temporary price distortions,<br />
<strong>and</strong> the protection <strong>of</strong> IPRs is crucial in those fields, such as in the New<br />
Economy, where the net benefits <strong>of</strong> patents are higher or in those fields, like<br />
the pharmaceutical sector, where social benefits are higher <strong>and</strong> there are<br />
proper policies which can reduce the social costs;<br />
2) restrictions to the patentability <strong>of</strong> innovations in high-tech sectors for<br />
one country or a group <strong>of</strong> countries could severely jeopardize investment in<br />
innovation <strong>and</strong> technological progress in the leading high-tech sectors with<br />
negative consequences on growth <strong>and</strong> competition in the global economy<br />
<strong>and</strong> would shift investments toward other countries where IPRs are better<br />
protected without regard for comparative advantage;<br />
32 On the same point see Erkal (2005) <strong>and</strong> Cozzi <strong>and</strong> Galli (2007).
5.4 Reforming <strong>Antitrust</strong> 195<br />
3) improvements <strong>of</strong> the effectiveness <strong>of</strong> the current patent systems should<br />
rather promote access to patents, especially for small <strong>and</strong> medium size enterprises<br />
which are traditionally less able to exploit this opportunity, <strong>and</strong><br />
enhance the spillovers created by the patent system on the diffusion <strong>of</strong> knowledge<br />
through further requirements on a disclosure <strong>of</strong> the patented inventions<br />
whichshouldbesufficiently clear <strong>and</strong> complete to be carried out by a person<br />
skilled in the art;<br />
4) a proper industrial policy promoting competition for the market should<br />
adequately protect <strong>and</strong> subsidize R&D investments, <strong>and</strong> at the same time<br />
guarantee open access to the markets for innovations.<br />
5.4 Reforming <strong>Antitrust</strong><br />
In the last few years there has been a lot <strong>of</strong> academic <strong>and</strong> political debate<br />
on how to reform the EU approach to antitrust, <strong>and</strong> in particular on issues<br />
concerning abuse <strong>of</strong> dominance, moving toward an economic based approach<br />
more similar to the US approach. European Commission (2005) proposed a<br />
new approach to exclusionary abuses under Article 82 which is the subject <strong>of</strong><br />
an open debate <strong>and</strong> gives an important indication as to how the Commission<br />
may approach antitrust cases <strong>of</strong> abuse <strong>of</strong> dominance in the future. We will<br />
comment on this debate focusing on the general principles <strong>of</strong> EU antitrust<br />
policy, but our discussion tries to provide principles for antitrust policy that<br />
could be applied to any national antitrust authority.<br />
The EU approach appears to move toward a purpose <strong>of</strong> competition policy<br />
associated with the protection <strong>of</strong> competition in the market as a means <strong>of</strong> enhancing<br />
consumer welfare <strong>and</strong> <strong>of</strong> ensuring an efficient allocation <strong>of</strong> resources.<br />
This implies that antitrust should protect competition <strong>and</strong> not competitors,<br />
<strong>and</strong> be based on an economic analysis aimed at the maximization <strong>of</strong> consumer<br />
welfare <strong>and</strong> allocative efficiency rather than based on a legalistic analysis, a<br />
new direction which appears much more in line with the consolidated US<br />
approach.<br />
While the aim is to enhance consumer welfare <strong>and</strong> to protect competition<br />
<strong>and</strong> not competitors, we have some concern that these principles are not fully<br />
carried through into certain aspects <strong>of</strong> the current EU competition policy <strong>and</strong><br />
<strong>of</strong> the proposal <strong>of</strong> European Commission (2005). 33 As a matter <strong>of</strong> fact, until<br />
now the approach <strong>of</strong> the European Commission has <strong>of</strong>ten been in line with<br />
outdated views, for instance when stressing an excessive reliance on market<br />
33 This section is partly derived by Etro (2006c). See International Chamber <strong>of</strong><br />
Commerce (2006,2007) for a more extensive <strong>and</strong> related treatment. Here we<br />
will focus on issues related to our previous theoretical analysis, namely market<br />
dominance, predatory pricing, bundling <strong>and</strong> protection <strong>of</strong> IPRs. I am grateful to<br />
Martti Virtanen <strong>of</strong> the Finnish <strong>Competition</strong> authority for precious comments on<br />
this section.
196 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
shares in determining dominance. The analysis <strong>of</strong> whether an undertaking has<br />
engaged in abusive conduct under Article 82 should ultimately turn on the<br />
conduct’s actual effects on efficiency <strong>and</strong> consumer welfare. Thus, we believe<br />
that, if the pro-consumer benefits <strong>of</strong> a dominant undertaking’s conduct are<br />
significant, it should be immune from liability even if it disadvantages certain<br />
competitors. Inventing better products or more efficient methods <strong>of</strong> distribution,<br />
reducing prices or <strong>of</strong>fering better terms <strong>of</strong> trade, <strong>and</strong> more quickly<br />
adapting to changes in the market can disadvantage rivals <strong>and</strong> maybe even<br />
cause them to exit the market. Yet, these forms <strong>of</strong> conduct <strong>of</strong>ten also enhance<br />
efficiency <strong>and</strong> consumer welfare.<br />
Thefocusontheeffects for consumers is particularly important with respect<br />
to fast-moving markets such as those commonly found in high-tech<br />
<strong>and</strong> New Economy industries which are <strong>of</strong>ten characterized by massive R&D<br />
investments, strong reliance on IPRs <strong>and</strong> other intangible assets, network effects,<br />
high sunk costs <strong>and</strong> low marginal costs. As we already noticed, under<br />
competition for the market, leading firms might enjoy high market shares<br />
yet be subject to massive competitive pressure to constantly create better<br />
products at lower prices due to threats from innovative competitors <strong>and</strong> potential<br />
entrants. Undertakings that hold a significant share <strong>of</strong> the market at<br />
any given point in time may see this share decrease rapidly <strong>and</strong> significantly<br />
following the development <strong>and</strong> supply <strong>of</strong> a new <strong>and</strong> more attractive product<br />
by an actual or potential competitor. Nevertheless, the current EU approach<br />
is still characterized by a close association between market shares <strong>and</strong> market<br />
dominance without any reference to the kind <strong>of</strong> market that is under<br />
consideration.<br />
5.4.1 Efficiency Defense<br />
Since the main purpose <strong>of</strong> antitrust policy should be the protection <strong>of</strong> consumers<br />
<strong>and</strong> <strong>of</strong> the efficient allocation <strong>of</strong> resources within sectors, it is important<br />
that strategies that create efficiency gains remain outside the realm <strong>of</strong><br />
abusive strategies.<br />
The proposal on the efficiency defenses for dominant firms contained in<br />
European Commission (2005) appears to be going in the right direction since<br />
it allows otherwise abusive strategies if they create a net efficiency gain (which<br />
benefits consumers). 34 In our view, conduct that generates efficiencies should<br />
not be deemed abusive unless it is demonstrated that the impact <strong>of</strong> this conduct<br />
on competition will result in consumer harm outweighing these efficien-<br />
34 This can happen in two ways: through an objective necessity defense “where the<br />
dominant company is able to show that the otherwise abusive conduct is actually<br />
necessary conduct on the basis <strong>of</strong> objective factors external to the parties involved<br />
<strong>and</strong> in particular external to the dominant company”, or a meeting competition<br />
defense “where the dominant company is able to show that the otherwise abusive<br />
conduct is actually a loss-minimising reaction to competition from others”.
5.4 Reforming <strong>Antitrust</strong> 197<br />
cies. Nevertheless, the proposal <strong>of</strong> the European Commission (2005) provides<br />
relatively limited scope for taking efficiencies into account. First, according<br />
to the proposed approach, it will fall on dominant undertakings to prove the<br />
extent to which their conduct was justified on grounds <strong>of</strong> efficiency. However,<br />
such a system would send the wrong signal to the business community: investigations<br />
would <strong>of</strong>ten move quite far along before efficiency considerations<br />
fully come into play. Placing the burden <strong>of</strong> pro<strong>of</strong> on competition authorities,<br />
by contrast, would make more sense as they are likely to be in a better<br />
position to obtain relevant evidence from the dominant undertaking as well<br />
as other market participants (such as consumer organizations) on whether<br />
challenged conduct promotes efficiency.<br />
Second, to assert a successful efficiency defense under the proposed framework,<br />
dominant undertakings will be required to show that there are no other<br />
economically practicable <strong>and</strong> less anticompetitive alternatives to achieve the<br />
claimed efficiencies. This condition means that liability could be imposed even<br />
on conduct whose efficiency <strong>and</strong> consumer benefits far outweigh its adverse<br />
effect on competitors simply because there exists an alternative that would<br />
have disadvantaged rivals less. We doubt that such a rule would have any<br />
economic justification <strong>and</strong> any basis in commercial reality.<br />
Finally, the effectiveness <strong>of</strong> these rules in safeguarding consumer welfare<br />
would be weakened under the proposal <strong>of</strong> European Commission (2005) for<br />
which some firms are virtually excluded from the possibility <strong>of</strong> an efficiency<br />
defense: according to this proposal, the protection <strong>of</strong> competitors would be<br />
given priority over efficiency when the dominant undertaking holds a market<br />
share above seventy-five per cent. In our view, efficiencies should be assessed<br />
in the same manner in all cases, regardless <strong>of</strong> the defendant’s market<br />
share: undertakings that generate pro-competitive efficiencies that benefit<br />
consumers should not be penalized regardless <strong>of</strong> the level <strong>of</strong> market share or<br />
potential impact on less efficient competitors.<br />
5.4.2 Predatory Pricing<br />
Predatory pricing is defined by European Commission (2005) as “the practice<br />
where a dominant company lowers its prices <strong>and</strong> thereby deliberately incurs<br />
losses or foregoes pr<strong>of</strong>its in the short run so as to eliminate or discipline one<br />
or more rivals or to prevent entry by one or more potential rivals thereby<br />
hindering the maintenance or the degree <strong>of</strong> competition still existing in the<br />
market or the growth <strong>of</strong> that competition”. The st<strong>and</strong>ard antitrust approach<br />
uses a number <strong>of</strong> cost benchmarks in order to assess whether “predatory pricing”<br />
by a dominant undertaking has actually taken place, <strong>and</strong> in particular<br />
it sets a cut-<strong>of</strong>f such that pricing below this cut-<strong>of</strong>f gives rise to a rebuttable<br />
presumption that the pricing is predatory. This strategy is supported by the<br />
traditional idea that pricing below marginal cost should have an exclusionary<br />
purpose in st<strong>and</strong>ard markets, while pricing above marginal cost should not.
198 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
The theory <strong>of</strong> market leaders emphasizes some limits <strong>of</strong> this way <strong>of</strong> thinking:<br />
pricing at or below marginal cost by the market leader does not need to<br />
exclude (equally efficient) competitors <strong>and</strong> it does not even need to induce<br />
short run losses for the same leader. To see why, let us remember that, as<br />
we noticed in Chapter 3 <strong>and</strong> in Section 5.2.1, a leader in a st<strong>and</strong>ard market<br />
with quantity competition <strong>and</strong> endogenous entry can generally choose<br />
between two alternative strategies. The first one is to price below the rivals<br />
<strong>and</strong> allow their entry with a price equal to their average cost but above the<br />
marginal cost. The second strategy is to choose a limit price such that entry<br />
is not pr<strong>of</strong>itable for any firm. The former strategy is optimal when marginal<br />
costs are increasing enough in the production level <strong>and</strong>/or products are differentiated,<br />
while the latter strategy is optimal in the case <strong>of</strong> decreasing or<br />
constant marginal costs <strong>and</strong>/or homogenous goods.<br />
Let us focus on the first situation. When goods are homogenous, the<br />
equilibrium strategy <strong>of</strong> the leader is simply to price at marginal cost, <strong>and</strong> its<br />
pr<strong>of</strong>its are positive because production is in the region where average total<br />
costs are increasing. When goods are differentiated, the equilibrium price <strong>of</strong><br />
the leader is above its marginal cost, <strong>and</strong> pr<strong>of</strong>its are again positive. As we have<br />
seen, in this equilibrium entry occurs, <strong>and</strong> is not deterred. Moreover, if the<br />
leader can obtain positive pr<strong>of</strong>its in equilibrium by pricing at marginal cost,<br />
positive pr<strong>of</strong>its could be preserved even by pricing slightly below marginal<br />
cost as long as the scale <strong>of</strong> production is large enough.<br />
Let us focus on the second situation now. The leader can deter entry when<br />
marginal costs are constant or decreasing <strong>and</strong>/or goods are homogenous, <strong>and</strong><br />
this happens with a price <strong>of</strong> the leader above the marginal cost. Nevertheless,<br />
when entry is endogenous this is a normal competitive strategy <strong>of</strong> a firm able<br />
to exploit scale economies <strong>and</strong> reduce average costs <strong>of</strong> production (<strong>and</strong>, as<br />
we have seen in Section 1.1, this strategy does not even reduce welfare).<br />
Finally, in Section 2.9 we saw that in dynamic (<strong>and</strong> multi-sided) markets<br />
where dem<strong>and</strong> is characterized by network externalities or supply is characterized<br />
by learning by doing, leaders may want to price below the marginal<br />
cost without entry deterrence purposes. The purpose <strong>of</strong> pricing below marginal<br />
cost would be to develop network effects or decrease costs for the future<br />
<strong>and</strong> to be more aggressive in the future competition.<br />
In conclusion, it is highly questionable that the marginal cost should be<br />
the right theoretical cut-<strong>of</strong>f below which predation can be presumed, <strong>and</strong> we<br />
do believe that a rule <strong>of</strong> reason should be applied also in this case, because<br />
different sectors <strong>and</strong> different cost <strong>and</strong> dem<strong>and</strong> structures require different<br />
approaches to the definition <strong>of</strong> predatory pricing.<br />
For the sake <strong>of</strong> argument, suppose we could agree that marginal cost<br />
pricing represents a crucial cut-<strong>of</strong>f under some circumstances. The problem<br />
is that it is quite difficult to measure such a figure. Therefore, many antitrust<br />
scholars, notably Areeda <strong>and</strong> Turner (1974), have proposed to substitute it<br />
with the average variable cost:
5.4 Reforming <strong>Antitrust</strong> 199<br />
“the incremental cost <strong>of</strong> making <strong>and</strong> selling the last unit cannot<br />
readily be inferred from conventional business accounts, which<br />
typically go no further than showing observed average variable cost.<br />
Consequently it may well be necessary to use the latter as an indicator<br />
<strong>of</strong> marginal cost”<br />
This rule has influenced antitrust policy worldwide, but one should always<br />
keep in mind that there are (dem<strong>and</strong> <strong>and</strong> technological) conditions<br />
under which its premise, the marginal cost as a cut-<strong>of</strong>f below which pricing<br />
is predatory, is not valid.<br />
On the basis <strong>of</strong> our theoretical discussion, we can now try to draw our<br />
conclusions on the proper approach to predatory pricing. As we have noticed<br />
repeatedly in this book, one can not judge the pricing behavior <strong>of</strong> a market<br />
leader in a correct way without taking the entry conditions into account.<br />
When entry is endogenous, in the practical sense that entry is driven by<br />
pr<strong>of</strong>itable opportunities <strong>and</strong> it is rapid, no firm can manipulate the market<br />
at its will. As McGee (1958) noticed in his pioneering work on predatory<br />
pricing, a necessary condition for the success, <strong>and</strong> therefore the viability, <strong>of</strong><br />
a predatory strategy is that entry must be exogenously blocked:<br />
“Obstacles to entry are necessary conditions for success. Entry is<br />
the nemesis <strong>of</strong> monopoly. It is foolish to monopolize an area or market<br />
into which entry is quick <strong>and</strong> easy. Moreover, monopolization that<br />
produces a firm <strong>of</strong> greater than optimum size is in for trouble if entry<br />
can occur even over a longer period. In general, monopolization will<br />
not pay if there is no special qualification for entry, or no relatively<br />
long gestation period for the facilities that must be committed for<br />
successful entry.”<br />
Only when entry is not feasible (even when it could be pr<strong>of</strong>itable), a leader<br />
can hope to exclude the current rivals <strong>and</strong> monopolize the market.<br />
On the basis <strong>of</strong> these considerations, we propose the following rule based<br />
on two steps:<br />
1) the <strong>Antitrust</strong> Authority should evaluate whether the undertaking is<br />
effectively constrained by endogenous entry <strong>of</strong> competitors in his strategic<br />
choices: if entry is endogenous dismiss the case, otherwise proceed.<br />
2) the <strong>Antitrust</strong> Authority should evaluate the relation between price, average<br />
total cost (ATC) <strong>and</strong> average variable cost (AVC):<br />
a) a price above ATC should be lawful without exceptions;<br />
b) a price below ATC but above AVC should be presumed lawful with the<br />
burden <strong>of</strong> proving the contrary on the <strong>Antitrust</strong> Authority, <strong>and</strong> on the basis<br />
<strong>of</strong> the consequences on consumers <strong>and</strong> allocative efficiency;<br />
c) a price below AVC should be presumed unlawful with the burden <strong>of</strong><br />
proving the contrary on the undertaking, through an efficiency defense or<br />
proving that dem<strong>and</strong> or technological conditions reduce the relevant cut-<strong>of</strong>f<br />
below the AVC.
200 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
Notice that the first step we propose is different from the traditional one,<br />
which simply evaluates whether there is a dominant position in the relevant<br />
market. 35 The traditional step is based on the idea that after excluding the<br />
rivals, a dominant firm can monopolize the market <strong>and</strong> recoup its initial losses<br />
with higher prices. But, this is impossible when entry in the competition in<br />
the market is endogenous (there is no way to recoup losses by increasing<br />
future prices if a price increase attracts entry), <strong>and</strong> it is extremely unlikely<br />
when entry in the competition for the market is endogenous (there is a low<br />
probability to recoup losses by increasing future prices <strong>of</strong> goods that may<br />
be soon replaced by innovations <strong>of</strong> other firms). The traditional definition<br />
<strong>of</strong> dominance (associated with the market share <strong>and</strong> the related indexes <strong>of</strong><br />
concentration) should not be the relevant element to establish the likelihood<br />
<strong>of</strong> recoupment, particularly in high-tech markets. We believe that the focus<br />
should not be on the market leader in the first step <strong>of</strong> an antitrust investigation<br />
for abuse <strong>of</strong> dominance, but on the followers <strong>and</strong> on the chances that<br />
these followers have to exploit pr<strong>of</strong>itable opportunities in the market.<br />
Concerning the second step in the evaluation <strong>of</strong> predatory strategies, EU<br />
antitrust has also adopted a similar approach (but an efficiency defense still<br />
needs to be formally introduced). However, the recent proposal by European<br />
Commission (2005) has suggested to substitute the AVC with an average<br />
avoidable cost (AAC), the average <strong>of</strong> the costs that could have been avoided<br />
if the undertaking had not produced a discrete amount <strong>of</strong> extra output (this<br />
extra output is usually the amount allegedly subject to abusive conduct), a<br />
sort <strong>of</strong> average marginal (or incremental) cost <strong>of</strong> the extra output to serve<br />
the predatory sales. Unfortunately, the AAC can be quite higher than the<br />
right theoretical concept whenever it accounts for fixed costs. Moreover, the<br />
AAC can be much more difficult to measure than the AVC, since it is almost<br />
always impossible to precisely define which costs are sustained for a given<br />
output <strong>and</strong> isolate the extra output (supposedly the predatory output) from<br />
the total one. Finally, there are well known conditions, as in the presence<br />
<strong>of</strong> network externalities <strong>and</strong> multi-sided markets, under which extremely ag-<br />
35 The definition <strong>of</strong> the relevant market generally depends on an empirical analysis<br />
<strong>of</strong> the way dem<strong>and</strong> <strong>of</strong> substitute products changes with changes in the price <strong>of</strong><br />
the hypothetical dominant firm. Such an analysis can be problematic because the<br />
market price could be above its competitive level. For instance, a widely used<br />
method is the SSNIP-test, which defines the relevant market as the smallest<br />
market where a Small but Significant Non-transitory Increase in Prices (say <strong>of</strong><br />
5-10%) increases the pr<strong>of</strong>its <strong>of</strong> a hypothetical monopolist. This test is ideal when<br />
prices are close to the competitive level, but otherwise it is biased <strong>and</strong> leads to a<br />
too-wide market definition (which in turn may lead to a finding<strong>of</strong>nodominance<br />
inawidemarket).Thisproblemisknownasthe‘cellophane fallacy’, from the<br />
subject <strong>of</strong> the du Pont case (1956). It should be noticed that such bias should<br />
not emerge (<strong>and</strong> the SSNIP-test at the prevailing prices should be valid) when<br />
the market leader is constrained by endogenous entry (see Etro, 2007,b).
5.4 Reforming <strong>Antitrust</strong> 201<br />
gressive pricing is a normal competitive strategy for a market leader. For<br />
instance, it is a st<strong>and</strong>ard practice for multi-sided markets to charge less one<br />
side <strong>of</strong> the market (as readers for a newspaper or end-users for video game<br />
consoles) <strong>and</strong> more the other side (advertisers <strong>and</strong> game developers in these<br />
examples), 36 without an exclusionary purpose but only to create network effects<br />
<strong>and</strong> increase the value <strong>of</strong> the interactions between the two sides. Often,<br />
the price on one side is not only below cost, but even below zero (the sale is<br />
subsidized with free add ons), <strong>and</strong> nevertheless even such a strategy is not<br />
necessarily predatory. For these reasons, we believe that the traditional AVC<br />
remains a better reference than the AAC. 37<br />
5.4.3 Bundling<br />
Looking at the approach <strong>of</strong> the European Commission (2005) on bundling,<br />
again it appears that its positive principles are not fully carried through.<br />
Indeed, economists today generally acknowledge that tying can produce positive<br />
efficiencies <strong>and</strong> consumer benefits, <strong>and</strong> that a rule <strong>of</strong> reason should be<br />
adopted in evaluating its anti-competitive effects. The pro-competitive effects<br />
are particularly pronounced in the case <strong>of</strong> technical tying (when companies<br />
innovate by linking formerly separate technologies or products, efficiencies <strong>of</strong>ten<br />
emerge through improved performance <strong>and</strong> quality). Moreover, they can<br />
also emerge because tying strengthens price competition, <strong>and</strong> so it can be<br />
used as an aggressive strategy by leaders facing endogenous entry in the secondary<br />
market: as we have seen in Section 2.10, under product differentiation<br />
in this market, such an aggressive strategy by the leader would induce low<br />
prices without eliminating product diversification in the secondary market.<br />
The current EU approach, however, perpetuates a biased position against<br />
bundling per se.<br />
36 According to the Rochet-Tirole (2003) rule, the higher price should be for the<br />
market side with higher elasticity <strong>of</strong> dem<strong>and</strong> <strong>and</strong> vice versa. But this goes right<br />
against the fundamental principle <strong>of</strong> monopoly pricing for which a higher price<br />
should be for the market side with lower elasticity <strong>of</strong> dem<strong>and</strong> <strong>and</strong> vice versa.<br />
37 In certain sectors, the same proposal uses a long-run average incremental cost<br />
benchmark (LAIC), instead <strong>of</strong> AAC. This is usually the case in industries where<br />
fixed costs are high <strong>and</strong> variable costs very low. In these cases, the LAIC benchmark<br />
is used as the benchmark below which predation is presumed. The same<br />
considerations as before hold also here: there are not economic justifications for a<br />
change <strong>of</strong> st<strong>and</strong>ard from AVC to LAIC. Moreover, we believe that the LAIC st<strong>and</strong>ard<br />
is inconsistent with business reality because it requires companies to price<br />
to cover their own average sunk fixed costs that are unrecoverable: this approach<br />
ignores the economic reality that, when market leaders decide how to price a<br />
product, they do not consider their own costs that are sunk or unrecoverable,<br />
even if not a single product is sold.
202 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
We have many doubts on the same definition <strong>of</strong> tying adopted in the EU<br />
approach, which places too much emphasis on consumer dem<strong>and</strong> for the tied<br />
product in the context <strong>of</strong> its “distinct products test” as a proxy for determining<br />
whether the tying arrangement produces efficiencies. To exemplify our<br />
doubts, notice that, while there is clearly consumer dem<strong>and</strong> for shoelaces, this<br />
should not mean that shoes <strong>and</strong> shoelaces are distinct products for the purposes<br />
<strong>of</strong> tying analysis. This issue can only be addressed by asking whether<br />
there is consumer dem<strong>and</strong> for shoes without shoelaces. In sum, whether or<br />
not consumer dem<strong>and</strong> exists for the tied product is the wrong question; the<br />
correct question is whether there is any significant consumer dem<strong>and</strong> for the<br />
tying product without the tied product. Unless the analysis focuses on this<br />
question, there is a danger that the mere existence <strong>of</strong> consumer dem<strong>and</strong> for<br />
the tied product may prevent the emergence <strong>of</strong> efficient tying arrangements<br />
<strong>and</strong> end up protecting suppliers <strong>of</strong> tied products at the expense <strong>of</strong> consumers<br />
<strong>and</strong> innovation.<br />
Moreover, in the case <strong>of</strong> technical integration <strong>of</strong> two products that were<br />
previously distinct, the distinct products test itself may not be helpful for<br />
underst<strong>and</strong>ing market dynamics because, by definition, this test is backwardlooking.<br />
A better approach in these cases would be simply to ask whether the<br />
company integrating the previously distinct products can make a plausible<br />
showing <strong>of</strong> efficiency gains: since technical tying is normally efficient, market<br />
leaders would be able to continue producing innovative products benefiting<br />
consumers without running afoul <strong>of</strong> the prohibitions on tying. Finally, since<br />
tying usually enhances price competition, it should never be abusive when<br />
it is st<strong>and</strong>ard commercial practice (which is also indirect evidence that such<br />
tying generates efficiencies, or that there is no dem<strong>and</strong> for the unbundled<br />
product).<br />
We are also concerned that the current approach fails to acknowledge<br />
that bundling can be used to create value for consumers in markets that experience<br />
network effects <strong>and</strong> in multi-sided markets (further analyzed in the<br />
next chapter). For instance, bundling is a valuable strategy to gain broader<br />
distribution <strong>of</strong> the products or services that are subject to network effects.<br />
And the broader the distribution, the greater the value produced for all consumers.<br />
This is particularly true when the product or service in question has<br />
low (or zero) marginal costs, because the supplier can include the product or<br />
service in bundles with other products at no cost. In a multi-sided market<br />
multiple types <strong>of</strong> customers gain from reciprocal interaction, as in the case<br />
<strong>of</strong> advertisers <strong>and</strong> readers for a journal: complex business models resulting<br />
from multi-sided markets <strong>of</strong>ten require bundling practices because the consumption<br />
on one side <strong>of</strong> the market is being “sold” on the other side <strong>of</strong> the<br />
market, <strong>and</strong> piece-meal consumption on one side <strong>of</strong> the market would break<br />
down the interdependent ecosystem. 38<br />
38 On the antitrust implications <strong>of</strong> multi-sided markets see Evans (2003b).
5.4 Reforming <strong>Antitrust</strong> 203<br />
Finally, in the EU approach the st<strong>and</strong>ard <strong>of</strong> pro<strong>of</strong> the antitrust authority<br />
is required to meet to establish harmful foreclosure effects is too low, particularly<br />
in light <strong>of</strong> the fact that the analysis <strong>of</strong> foreclosure effects can be<br />
speculative in nature. According to the current EU approach to bundling,<br />
actual market foreclosure effects are not required: it is enough that such effects<br />
are “likely” to occur. In other words, the mere risk <strong>of</strong> foreclosure can<br />
result in a finding against a dominant company. A st<strong>and</strong>ard <strong>of</strong> pro<strong>of</strong> that<br />
requires convincing evidence would rather help ensure that companies will<br />
not be deterred from bringing new products to market as a result <strong>of</strong> concerns<br />
about remote <strong>and</strong> potential foreclosure effects.<br />
5.4.4 Intellectual Property Rights<br />
Finally, we want to look at the relationship between antitrust <strong>and</strong> the protection<br />
<strong>of</strong> IPRs. While we noticed that the latter should be the focus <strong>of</strong><br />
legislation <strong>and</strong> not <strong>of</strong> the discretionary behavior <strong>of</strong> antitrust authorities, the<br />
current EU approach deals with IPRs in the discipline on refusals to supply,<br />
that is, situations where a dominant company denies a buyer access to an<br />
input in order to exclude that buyer from participating in an economic activity.<br />
In general, four conditions have to be fulfilled in order to find a refusal<br />
to supply be abusive: i) the behavior must be properly characterized as a<br />
termination <strong>of</strong> a previous supply arrangement; ii) the refusing undertaking<br />
must be dominant; iii) the refusal must be likely to have a negative effect<br />
on competition; <strong>and</strong> iv) the refusal must not be justified objectively or by<br />
efficiencies. Only when the dominant supplier has not previously supplied the<br />
input to a potential buyer, as for IPRs, an additional criterion is added: v)<br />
the input must be “indispensable” to carry on normal economic activity in<br />
the downstream market (a so-called “essential facility”).<br />
Nevertheless, the European Commission (2005) correctly points out that<br />
“to maintain incentives to invest <strong>and</strong> innovate, the dominant firm must not<br />
be unduly restricted in the exploitation <strong>of</strong> valuable results <strong>of</strong> the investment.<br />
For these reasons the dominant firm should normally be free to seek compensation<br />
for successful projects that is sufficient to maintain investment<br />
incentives, taking the risk <strong>of</strong> failed projects into account. To achieve such<br />
compensation, it may be necessary for the dominant firm to exclude others<br />
from access to the input for a certain period <strong>of</strong> time.” The proposal clearly<br />
states the priority <strong>of</strong> IPR protection, saying that “imposing on the holder<br />
<strong>of</strong> the rights the obligation to grant to third parties a licence for the supply<br />
<strong>of</strong> products incorporating the IPR, even in return for a reasonable royalty,<br />
would lead to the holder being deprived <strong>of</strong> the substance <strong>of</strong> the exclusive<br />
right”. Therefore, another more restrictive criterion is added in the case <strong>of</strong><br />
a refusal to license IPRs: the undertaking which requests the licence should<br />
intend to produce new goods or services not <strong>of</strong>fered by the owner <strong>of</strong> the IPRs<br />
<strong>and</strong> for which there is a potential consumer dem<strong>and</strong>. This additional criterion<br />
is in line with established case law, but an exception to this criterion is
204 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
introduced by the European Commission (2005). This states that a refusal to<br />
license IPR-protected technology which is indispensable for follow-on innovation<br />
may be abusive even if the license is not sought to directly incorporate<br />
the technology in clearly identifiable new goods <strong>and</strong> services, since the refusal<br />
to license an IPR-protected technology “should not impair consumers’ ability<br />
to benefit from innovation brought about by the dominant undertaking’s<br />
competitors”. This exception is inconsistent with economic analysis. As we<br />
have seen in Chapter 4, there are no clear economic arguments supporting<br />
the view that weakening IPRs could ever strengthen innovation in the long<br />
run, even when innovation is sequential. As a matter <strong>of</strong> fact, the opposite is<br />
true: the protection <strong>of</strong> IPRs for sequential innovations is more important to<br />
promote innovation <strong>and</strong> growth because it creates a multiplicative effect on<br />
the incentives to innovate <strong>and</strong> it fosters technological progress <strong>and</strong> growth.<br />
Finally, concerning the refusal to supply information needed for interoperability,<br />
the proposal in European Commission (2005) states that leveraging<br />
market power from one market to another may be an abuse <strong>of</strong> a dominant<br />
position <strong>and</strong> it may not be appropriate to apply the same high st<strong>and</strong>ards for<br />
intervention even if such information may be considered a trade secret. The<br />
framework for assessing how such leveraging may occur or when trade secrets<br />
do not deserve the same high st<strong>and</strong>ards for protection has not been developed<br />
yet. Again, such a broad policy intervention could have chilling effects<br />
on the incentives to invest <strong>and</strong> innovate <strong>and</strong> could ultimately end up protecting<br />
inefficient competitors that may free ride on the risks <strong>and</strong> investments <strong>of</strong><br />
the dominant undertaking, therefore in contradiction with the objective <strong>of</strong><br />
protecting competition on the merits.<br />
5.5 Conclusions<br />
In this book we have proposed an alternative approach to antitrust policy.<br />
Taking into account the endogeneity <strong>of</strong> entry, we have seen that st<strong>and</strong>ard results<br />
<strong>of</strong> the post-Chicago literature can be radically modified. The main implications,<br />
analyzed in this chapter, concern the behavior <strong>of</strong> market leaders <strong>and</strong>,<br />
consequently, the antitrust approach to abuse <strong>of</strong> dominance. In other parts<br />
<strong>of</strong> this book, we have also derived implications for the antitrust approach<br />
to mergers, collusion <strong>and</strong> state aids. The overall flavor <strong>of</strong> our approach to<br />
antitrust is reminiscent <strong>of</strong> the Chicago school. Nevertheless, our analysis is<br />
based on solid game theoretic foundations that the original Chicago view did<br />
not have.<br />
The theory <strong>of</strong> market leaders has shown that whether entry in a market<br />
is exogenous or endogenous makes a lot <strong>of</strong> difference for the way leaders<br />
behave. In markets where entry is independent from the pr<strong>of</strong>itability conditions,<br />
market leaders can adopt accommodating strategies to increase prices<br />
or aggressive ones to exclude rivals, <strong>and</strong> their strategies can harm consumers.
5.5 Conclusions 205<br />
When entry is endogenously dependent on the pr<strong>of</strong>itability conditions in the<br />
market, the leaders always adopt aggressive strategies which typically do not<br />
harm consumers. For instance, a firm competing with a single rival could<br />
engage in accommodating pricing to increase mark ups, or could engage<br />
in predatory pricing to induce the exit <strong>of</strong> the rival, but a firm facing endogenous<br />
entry <strong>of</strong> competitors will ordinarily engage in aggressive pricing<br />
strategies without exclusionary purposes. A monopolist in a primary market<br />
competing with a single rival on a secondary market may bundle its goods to<br />
monopolize the secondary market as well, but when the secondary market is<br />
characterized by endogenous entry the only purpose <strong>of</strong> bundling can be the<br />
strengthening <strong>of</strong> price competition. A firm facing a single rival could adopt<br />
vertical restraints on its retailers or price discrimination strategies to s<strong>of</strong>ten<br />
price competition, but when the same firm faces endogenous entry <strong>of</strong> rivals<br />
these anti-competitive practices will not be in its interest. Of course, notice<br />
that efficiency reasons can still motivate the adoption <strong>of</strong> bundling, vertical<br />
restraints, price discrimination or other strategies.<br />
The theory <strong>of</strong> endogenous entry delivers a related <strong>and</strong> strong result on<br />
horizontal mergers (see Section 2.13). As well known, even in the absence <strong>of</strong><br />
cost efficiencies, these mergers are <strong>of</strong>ten pr<strong>of</strong>itable when entry is exogenous<br />
because they allow the merged entity to increase prices or restrict production<br />
so as to enhance pr<strong>of</strong>itability. These effects are counterproductive when entry<br />
is endogenous because any accommodating strategy attracts entry. Therefore,<br />
the only rationale for mergers in markets with endogenous entry must be a<br />
cost efficiency large enough to (more than) compensate the strategic disadvantages<br />
associated with the merger. In these cases, mergers are welfare<br />
improving.<br />
The theory <strong>of</strong> endogenous entry also has some implications for the case in<br />
which collusive cartels are organized between a restricted number <strong>of</strong> firms (see<br />
Section 3.5). These cartels, as with any price fixing agreements, always lead<br />
to higher prices <strong>and</strong> lower welfare when the number <strong>of</strong> firms in the market<br />
is exogenous. However, when entry in the market is endogenous collusive<br />
cartels are ineffective, unless they act as leaders. In this last case, the cartels<br />
coordinate aggressive strategies aimed at increasing the market shares <strong>of</strong> their<br />
members through low prices, <strong>and</strong> their implementation is always sustainable<br />
<strong>and</strong> does not harm consumers.<br />
The theory <strong>of</strong> market leaders <strong>and</strong> endogenous entry can be applied to<br />
st<strong>and</strong>ard problems <strong>of</strong> strategic policy to evaluate the role <strong>of</strong> state aids aimed<br />
at promoting exports. These are always optimal when the domestic firms export<br />
in markets where entry is endogenous, <strong>and</strong> they do not harm domestic<br />
or foreign consumers. Therefore limitations to state aids <strong>and</strong> export subsidies<br />
should be exempted when they concern firms competing in international<br />
markets where entry is free.<br />
Finally, the theory <strong>of</strong> market leaders <strong>and</strong> endogenous entry has implications<br />
also in the case <strong>of</strong> competition for the market. <strong>Market</strong> leaders invest
206 5. <strong>Antitrust</strong> <strong>and</strong> Abuse <strong>of</strong> Dominance<br />
more in R&D when threatened by a competitive pressure, while they tend<br />
to stifle innovation in the absence <strong>of</strong> such a pressure: hence, persistence <strong>of</strong> a<br />
leadership in high-tech sectors can be consistent with effective dynamic competition<br />
for the market, which leads to a faster rate <strong>of</strong> technological progress<br />
in the interest <strong>of</strong> consumers.<br />
It is clear that the relevance <strong>of</strong> our results depends on the relevance <strong>of</strong><br />
the hypothesis that entry is endogenous. As we repeatedly pointed out, it<br />
does not matter what constrains entry, but simply that some fixed costs <strong>of</strong><br />
production or some opportunity costs <strong>of</strong> participating to the competition<br />
endogenously limit entry <strong>of</strong> firms in the market. One may argue that entry<br />
can be regarded as endogenous in the medium <strong>and</strong> long run, but not in the<br />
short run. If this is the case, <strong>and</strong> if antitrust policy is aimed at correcting<br />
distortions in the medium <strong>and</strong> long run (as opposed to short run distortions),<br />
then our results are potentially relevant.<br />
In the next chapter we move on to examine in more detail the markets<br />
<strong>of</strong> the New Economy. A recent important article by Segerstrom (2007) on<br />
technological progress in the New Economy has a simple <strong>and</strong> suggestive title,<br />
“Intel Economics”. This title emphasizes the importance <strong>of</strong> market leaders<br />
<strong>of</strong> the high-tech sectors in driving innovation <strong>and</strong> global growth. In the next<br />
chapter, we will borrow the style <strong>of</strong> that title to refer to what is probably<br />
the major market leader <strong>of</strong> the New Economy <strong>and</strong> the subject <strong>of</strong> some <strong>of</strong> the<br />
most representative antitrust cases in recent history.
6. Micros<strong>of</strong>t Economics<br />
After examining theoretical <strong>and</strong> institutional aspects <strong>of</strong> the behavior <strong>of</strong> market<br />
leaders <strong>and</strong> <strong>of</strong> the role <strong>of</strong> antitrust policy, this chapter approaches an<br />
important example <strong>of</strong> market leadership <strong>and</strong> technological leadership which<br />
is also associated with well known antitrust issues. The choice <strong>of</strong> Micros<strong>of</strong>t<br />
as a symbol <strong>of</strong> market leadership is somewhat natural: Micros<strong>of</strong>t is one <strong>of</strong><br />
the most visible <strong>and</strong> relevant companies in the New Economy, one <strong>of</strong> the<br />
most innovative firms in one <strong>of</strong> the most dynamic industries. The antitrust<br />
cases in which this company has been involved in both the US <strong>and</strong> the EU<br />
attracted primary attention <strong>of</strong> media, policymakers <strong>and</strong> observers. Many important<br />
economists were involved in these antitrust cases in both the US <strong>and</strong><br />
the EU, <strong>and</strong> many others were inspired by them while developing theoretical<br />
<strong>and</strong> empirical analysis on the structure <strong>of</strong> the s<strong>of</strong>tware market, on the role<br />
<strong>of</strong> Micros<strong>of</strong>t in this market <strong>and</strong> on the role <strong>of</strong> antitrust policy for the New<br />
Economy.<br />
In a recent important book, Evans et al. (2006) have emphasized the<br />
crucial role that s<strong>of</strong>tware platforms are playing in shaping our economies,<br />
in enhancing the development <strong>of</strong> many traditional sectors, <strong>and</strong> ultimately<br />
in affecting our way <strong>of</strong> living. These “invisible engines”, as they call the<br />
s<strong>of</strong>tware platforms, power not only the PC industry but also other industries<br />
like those associated with mobile phones <strong>and</strong> other h<strong>and</strong>held devices,<br />
video games, digital music, <strong>and</strong> (with strong externalities for the rest <strong>of</strong> the<br />
economy) on-line auctions, online searches <strong>and</strong> web-based advertising. Their<br />
claim is that s<strong>of</strong>tware platforms <strong>and</strong> microprocessors are at the basis <strong>of</strong> a new<br />
industrial revolution, exactly as the steam engine had been at the basis <strong>of</strong><br />
the first industrial revolution (1760-1830) <strong>and</strong> electric power at the basis <strong>of</strong><br />
the second industrial revolution (1850-1930). The third industrial revolution<br />
began with the introduction <strong>of</strong> commercial PCs in the early 80s <strong>and</strong> had a<br />
second phase starting in the mid 90s with the diffusion <strong>of</strong> the Internet. 1 Ob-<br />
1 The Internet is a global network <strong>of</strong> interconnected computer networks (linked<br />
by copper wires, fiber-optic cables <strong>and</strong> wireless connections) that transmit data<br />
by packet switching using the st<strong>and</strong>ard Internet Protocol (IP) <strong>and</strong> the Transfer<br />
Control Protocol (TCP). The World Wide Web (WWW) is a collection <strong>of</strong> interconnected<br />
documents <strong>and</strong> other resources (linked by hyperlinks <strong>and</strong> Uniform<br />
Resource Locators, or URLs) that is accessible via the Internet, as are many
208 6. Micros<strong>of</strong>t Economics<br />
servers have talked about “Intel economics”, “Micros<strong>of</strong>t economics” or the<br />
“Internet economics” to refer to this period <strong>of</strong> innovations in general purpose<br />
technologies, <strong>and</strong> to describe the ultimate engine <strong>of</strong> growth in the New<br />
Economy. 2<br />
What follows in this chapter surveys the wide academic debate on these<br />
issues. Our aim is not to provide a comprehensive analysis <strong>of</strong> the s<strong>of</strong>tware<br />
market or <strong>of</strong> the role <strong>of</strong> Micros<strong>of</strong>t, but to point out relations between our<br />
theoretical results on the behavior <strong>of</strong> market leaders <strong>and</strong> the structure <strong>of</strong><br />
this market, <strong>and</strong> use this theoretical background to evaluate antitrust issues<br />
involving Micros<strong>of</strong>t.<br />
The chapter is organized as follows. Section 6.1 describes the development<br />
<strong>of</strong> the s<strong>of</strong>tware market within the New Economy, <strong>and</strong> the role <strong>of</strong> Micros<strong>of</strong>t<br />
in this environment. Section 6.2 describes the genesis <strong>of</strong> the antitrust cases<br />
which involved Micros<strong>of</strong>t <strong>and</strong> the remaining sections adopt our theoretical<br />
instruments in evaluating the basic issues emerging in these cases: whether<br />
Micros<strong>of</strong>t is a monopoly in Section 6.3, whether its bundling strategies are<br />
predatory in Section 6.4, <strong>and</strong> whether its innovations should be disclosed to<br />
promote interoperability in Section 6.5. We conclude in Section 6.6.<br />
6.1 The S<strong>of</strong>tware <strong>Market</strong><br />
The s<strong>of</strong>tware market was developed in the last few decades. 3 In the 1960s, the<br />
computer industry was dominated by IBM, which manufactured mainframe<br />
computers used by large enterprise customers. These computers were expensive<br />
to purchase <strong>and</strong> expensive to maintain. As a result, very few consumers<br />
had access to computers. Apart from IBM, mainframes were <strong>of</strong>fered by firms<br />
such as Sperry, Control Data, Philco, Burroughs, General Electric, NCR, ect.<br />
other services like the emails. Other protocols or applications run on top <strong>of</strong> this<br />
structure. In 1958 United States created the Advanced Research Projects Agency,<br />
which supported first the research <strong>of</strong> the MIT Lincoln Laboratory in networking<br />
country-wide radar systems, <strong>and</strong> then the development <strong>of</strong> ARPANET, the<br />
main predecessor <strong>of</strong> the Internet, activated in 1969 at UCLA. The first TCP/IP<br />
wide area network was operational by 1983, when the American National Science<br />
Foundation constructed a university network backbone that would later become<br />
the NSFNet. In 1991 the European CERN launched the new WWW project<br />
after having created the Hypertext markup language (html), the predominant<br />
language for the creation <strong>of</strong> web pages, the Hypertext transfer protocol (http),<br />
the application that links <strong>and</strong> provides access to the files, documents <strong>and</strong> other<br />
resources <strong>of</strong> the WWW, <strong>and</strong> the first web pages. Popular web browsers emerged<br />
soon after that.<br />
2 On the role <strong>of</strong> the Information <strong>and</strong> Communication Technology in the recent<br />
growth experience see Dosi et al. (2007)<br />
3 Part <strong>of</strong> this section is based on Etro (2007d).
6.1 The S<strong>of</strong>tware <strong>Market</strong> 209<br />
In the mid 70s the US market was still dominated by IBM followed by Honeywell,<br />
Burroughs, Sperry, Control Data, NCR, Digital, G.E. <strong>and</strong> Hewlett-<br />
Packard (see Sutton, 1998, Ch. 15). There was little or no interoperability<br />
among mainframes from different vendors. For the most part, an enterprise<br />
customer was required to choose an all IBM solution or an all Sperry solution.<br />
In the 1970s, Digital Equipment achieved considerable success with<br />
a line <strong>of</strong> less expensive minicomputers that were well-suited to engineering<br />
<strong>and</strong> scientific tasks. Again, however, there was little or no interoperability<br />
between these minicomputers <strong>and</strong> mainframes <strong>of</strong>fered by IBM <strong>and</strong> others.<br />
The structure <strong>of</strong> the industry at that time was still largely vertical. By 1980,<br />
a number <strong>of</strong> companies had started <strong>of</strong>fering less expensive microcomputers<br />
which, again, were not interoperable with one another. Early PCs by Altair,<br />
T<strong>and</strong>y, Apple, Texas Instruments, Commodore <strong>and</strong> Atari ran their own operating<br />
systems, meaning that applications written for one br<strong>and</strong> <strong>of</strong> PC would<br />
not run on any other br<strong>and</strong>: the industry was fragmented. Apple, founded by<br />
Steve Jobs <strong>and</strong> Steve Wozniak in 1976, developed a very successful s<strong>of</strong>tware<br />
platform, especially because <strong>of</strong> VisiCalc, an electronic spreadsheet which was<br />
introduced in 1979 <strong>and</strong> soon became a killer application for Apple II.<br />
In the early 80s, IBM announced plans to introduce an IBM personal<br />
computer. The first one was <strong>of</strong>fered with operating systems (OSs) produced<br />
by others: CP/M-86 from Digital Research (a rewrite <strong>of</strong> the leading OS at<br />
the time), UCSD-p System by S<strong>of</strong>tech Microsystems, <strong>and</strong> PC-DOS developed<br />
by Micros<strong>of</strong>t, a company founded by Bill Gates, a young s<strong>of</strong>tware architect<br />
who dropped out <strong>of</strong> Harvard University to create what was going to become<br />
a symbol <strong>of</strong> market leadership in the New Economy. Micros<strong>of</strong>t’s OS won<br />
the race mainly because it was cheaper than CP/M-86 ($ 40 against $ 240)<br />
<strong>and</strong> faster than p-System. Moreover, Micros<strong>of</strong>t managed to keep the right to<br />
license its OS to other PC makers, under the name MS-DOS: this drove its<br />
success in the s<strong>of</strong>tware market. As Evans et al. (2006) noticed,<br />
“having multiple operating systems run on a hardware platform is<br />
a poor strategy. The idea, <strong>of</strong> course, was to ensure that the hardware,<br />
not the operating system, became the st<strong>and</strong>ard that defined the platform<strong>and</strong>determineditsevolution.Indeed,IBMfollowedanimportant<br />
economic principle for traditional industries: all firms would like<br />
everyone else in the supply chain to be competitive. IBM didn’t seem<br />
to recognize that this was far from a traditional industry... Applications<br />
are generally written for s<strong>of</strong>tware platforms, not the underlying<br />
hardware. The more fragmented the installed base <strong>of</strong> operating systems,<br />
the less attractive it is to write an application for any one <strong>of</strong><br />
them.”<br />
Not surprisingly, IBM’s multiple-OS strategy did not work, the hardware<br />
sector became always more fragmented, with many PC manufacturers producing<br />
clones <strong>of</strong> the IBM PC <strong>and</strong> most <strong>of</strong> them running MS-DOS, the exact<br />
replica <strong>of</strong> the operating system running on IBM PCs. In the second half <strong>of</strong>
210 6. Micros<strong>of</strong>t Economics<br />
the 80s IBM reacted by developing a new operating system, OS/2, while<br />
Micros<strong>of</strong>t independently developed Windows, whose lead at that point became<br />
unreachable. According to some observers, IBM based its strategy on<br />
its br<strong>and</strong> name <strong>and</strong> its research capacity, while Micros<strong>of</strong>t invested more in<br />
supporting the developers <strong>of</strong> s<strong>of</strong>tware applications <strong>and</strong> in what is <strong>of</strong>ten called<br />
“evangelization”: convincing s<strong>of</strong>tware producers to develop applications for<br />
Windows. This was the winning strategy: the share <strong>of</strong> IBM in the market for<br />
the so-called IBM-compatible PCs decreased over time (in 2004 IBM arrived<br />
to the point <strong>of</strong> selling its PC division to Lenovo), while the market share <strong>of</strong><br />
Micros<strong>of</strong>t in the s<strong>of</strong>tware market increased.<br />
Over time, the computer industry had moved from the old vertical structure<br />
toward a horizontal structure. This was characterized by a market for<br />
chips (Intel as a leader, Motorola, ARM, TI, AMD,..), one for hardware <strong>and</strong><br />
peripheral equipment (IBM, Dell, Hewlett-Packard, Packard Bell, Compaq,<br />
Gateway, Acer, Fujitsu,...), one for operating systems (Windows as a leader,<br />
OS/2, Unix, Linux, Solaris,..), one for application s<strong>of</strong>tware (Office, Scientific<br />
Workplace, Adobe Acrobat, Macromedia Dreamweaver,..) <strong>and</strong> one for sales<br />
<strong>and</strong> distribution, with competition within horizontal levels <strong>and</strong> higher interoperability<br />
across levels. A similar horizontal structure has emerged in the<br />
industries for mobile phones <strong>and</strong> personal organizers. This is not by chance:<br />
such decentralized structures can work well when technical interactions between<br />
complementary products are stable <strong>and</strong> well defined, while vertical<br />
structures would become too rigid to control them. Apple, the only large<br />
player remaining a fully integrated structure producing both hardware <strong>and</strong><br />
s<strong>of</strong>tware for its PCs <strong>and</strong> for other devices, had to become quite active in<br />
attracting applications from other s<strong>of</strong>tware developers, in order to build network<br />
externalities. 4<br />
6.1.1 Network Effects<br />
A s<strong>of</strong>tware platform is a s<strong>of</strong>tware program that makes services available to<br />
other s<strong>of</strong>tware programs through external “hooks” called Application Programming<br />
Interfaces (APIs). Examples are the operating systems running on<br />
PCs as Windows, Mac OS or Linux, those employed by videogame consoles<br />
as the Sony one for PlayStation or Windows 2000 for the Xbox, the Symbian<br />
4 Famous is the 1985 letter by Bill Gates to Apple, which advised its future main<br />
competitor to license the Mac OS to PC manufacturers to create network effects<br />
<strong>and</strong> establish a st<strong>and</strong>ard (at the time Micros<strong>of</strong>t was still earning its revenues<br />
mainly from s<strong>of</strong>tware applications, most <strong>of</strong> which, like Word <strong>and</strong> Excel, were important<br />
applications for the Macintosh). As well known, Apple chose the harder<br />
way, but it is still a strong <strong>and</strong> extremely innovative company in the PC industry<br />
today.
6.1 The S<strong>of</strong>tware <strong>Market</strong> 211<br />
operating system for cellular phones, 5 Palm OS for personal digital assistants<br />
(PDAs), RIM for the BlackBerry, Mac OS for the Apple iPod <strong>and</strong> iPhone,<br />
<strong>and</strong> so forth.<br />
To underst<strong>and</strong> the peculiarities <strong>of</strong> the s<strong>of</strong>tware market in general it is convenient<br />
to focus briefly on the main functions <strong>of</strong> PC operating systems. The<br />
main one is to serve as a platform on which applications can be created by<br />
s<strong>of</strong>tware developers. OSs supply different types <strong>of</strong> functionality, referred to as<br />
system services, that s<strong>of</strong>tware developers can call upon in creating their applications.<br />
These system services are made available through APIs. When an<br />
application calls a particular API, the operating system supplies the system<br />
service associated with that API by causing the microprocessor to execute a<br />
specified set <strong>of</strong> instructions. S<strong>of</strong>tware developers need well-defined platforms<br />
that remain stable over time. They need to know whether the system services<br />
on which their applications rely will be present on any given PC. Otherwise<br />
theywouldhavetowritethes<strong>of</strong>twarecodetoprovideequivalentfunctionality<br />
in their own applications, generating redundancy, inefficiency <strong>and</strong> a lack <strong>of</strong><br />
interoperability. Moreover, modern OSs provide a user interface, the means<br />
by which users interact with their computers. User interfaces for computers<br />
have evolved dramatically over the last decades, from punch card readers,<br />
to teletype terminals, to character-based user interfaces, to graphical user<br />
interfaces, first introduced (at a low price) by Apple with Macintosh in 1984.<br />
Finally, OSs enable users to find <strong>and</strong> use information contained in various<br />
storage devices: local ones, such as a floppy diskette, a CD-ROM drive, a<br />
jump drive or the hard drive built into a PC, or remote ones, such as local<br />
area networks that connect computers in a particular <strong>of</strong>fice, wide area networks<br />
that connect computers in geographically separated <strong>of</strong>fices, <strong>and</strong> the<br />
Internet.<br />
Over time, the OS <strong>of</strong> Micros<strong>of</strong>t became the most popular because Micros<strong>of</strong>t<br />
continually added new functionality <strong>and</strong> licensed it to a wide range<br />
<strong>of</strong> computer manufacturers with extremely aggressive price strategies. Micros<strong>of</strong>t<br />
recognized early on that an OS that served as a common platform<br />
for developing applications <strong>and</strong> could run on a wide range <strong>of</strong> PCs would<br />
provide substantial benefits to consumers. Among other advantages, development<br />
costs would fall <strong>and</strong> a broader array <strong>of</strong> products would become available<br />
because products could be developed for the common platform rather than<br />
for a large number <strong>of</strong> different platforms. By providing a single OS that ran<br />
on multiple br<strong>and</strong>s <strong>of</strong> PCs, Micros<strong>of</strong>t enabled s<strong>of</strong>tware developers to create<br />
applications, confident that users could run those applications on PCs from<br />
many different computer manufacturers. In addition, applications developed<br />
for a single platform are more easily interoperable because they rely on the<br />
same functionality supplied by the underlying OS.<br />
5 Symbian is a joint venture founded by Nokia, Ericsson <strong>and</strong> Motorola (which left<br />
it in 2003). It is currently owned (in order <strong>of</strong> shares <strong>of</strong> stocks) by Nokia, Ericsson,<br />
Sony Ericsson, Panasonic, Siemens AG <strong>and</strong> Samsung.
212 6. Micros<strong>of</strong>t Economics<br />
The original winning strategy <strong>of</strong> Micros<strong>of</strong>t was the creation <strong>of</strong> these network<br />
effects between hardware producers, s<strong>of</strong>tware developers <strong>and</strong> consumers:<br />
computer manufacturers benefit because their PCs can run the many applications<br />
written for Windows <strong>and</strong> because users are familiar with the Windows<br />
user interface; s<strong>of</strong>tware developers benefit because their applications can rely<br />
on system services exposed by Windows via published APIs <strong>and</strong> because they<br />
can write applications with assurance that they will run on a broad range<br />
<strong>of</strong> PCs; consumers benefit because they can choose from among thous<strong>and</strong>s<br />
<strong>of</strong> PC models <strong>and</strong> applications that will all work well with one another <strong>and</strong><br />
because such broad compatibility fosters intense competition among computer<br />
manufacturers <strong>and</strong> s<strong>of</strong>tware developers to deliver improved products<br />
at attractive prices. But this argument should not be overemphasized: for<br />
many years, PC-DOS <strong>and</strong> OS/2 had as many applications as Windows, but<br />
IBM’s decline did not stop. There is indeed another <strong>and</strong> more traditional<br />
element that is fundamental also in the s<strong>of</strong>tware market: the other crucial aspect<br />
<strong>of</strong> the strategy <strong>of</strong> Micros<strong>of</strong>t was its aggressive pricing strategy. This was<br />
strengthened through the development <strong>of</strong> the same network effects: conquering<br />
market shares, Micros<strong>of</strong>t could spread its huge fixed costs <strong>of</strong> production<br />
over a larger market <strong>and</strong> reduce the price, which in turn could enhance the<br />
network effects.<br />
6.1.2 Multi-sided Platforms<br />
S<strong>of</strong>tware platforms, as we have seen, deal with multiple sides. Micros<strong>of</strong>t deals<br />
with at least three: consumers, s<strong>of</strong>tware developers <strong>and</strong> PC manufacturers.<br />
Apple produces hardware internally, hence it deals with the remaining two<br />
sides: consumers <strong>and</strong> s<strong>of</strong>tware developers. Sometimes relationships are even<br />
more complex, as in the platform for smart mobile phones where, beyond<br />
OSs, s<strong>of</strong>tware developers <strong>and</strong> h<strong>and</strong>set makers, there are network operators (as<br />
Vodafone, NTT, T-Mobile, Orange, China Mobile, Telecom Italia Mobile,..)<br />
playing a coordinating role. 6<br />
In the presence <strong>of</strong> multiple sides with network effects between them, the<br />
choice <strong>of</strong> which ones should be charged more to use the platform is not simple.<br />
Rochet <strong>and</strong> Tirole (2003), Caillaud <strong>and</strong> Jullien (2003) <strong>and</strong> Evans (2003a) have<br />
noticed that s<strong>of</strong>tware platforms, as other similar multi-sided platforms, give<br />
rise to market structures that are quite different from the traditional ones.<br />
For simplicity, here we will refer to two-sided platforms, which connect two<br />
sides in such a way that for each side the valuation <strong>of</strong> the interactions with the<br />
other side depends on the number <strong>of</strong> agents on the others side. These network<br />
externalities,<strong>and</strong>inparticularthenonneutral impact <strong>of</strong> the pricing structure<br />
on both sides (<strong>and</strong> therefore on these externalities) distinguishes a two-sided<br />
6 Moreover in this ecosystem not only competition within layers is strong, but also<br />
competition between layers is relevant.
6.1 The S<strong>of</strong>tware <strong>Market</strong> 213<br />
market from a traditional one-sided market with different consumers (<strong>and</strong><br />
possibly price-discrimination between them). 7<br />
An analogous situation to s<strong>of</strong>tware platforms emerges in many completely<br />
different contexts. A classic example, useful to underst<strong>and</strong> the implications<br />
<strong>of</strong> any kind <strong>of</strong> platforms, is given by newspapers. They are sold to readers,<br />
but they also sell advertising space to advertisers: the reader is not only a<br />
“customer” <strong>of</strong> the newspaper, the reader is also a supplier <strong>of</strong> “eyeballs” that<br />
the newspaper sells to advertisers. In this case network effects emerge because<br />
advertisers value their advertising more in a newspaper when the number <strong>of</strong><br />
its readers is higher (the effect in the other direction may exist but is typically<br />
less important). This has crucial consequences on the pricing structure<br />
since a low price for the readers increases the number <strong>of</strong> sold copies <strong>and</strong> in<br />
turn enhances the value <strong>of</strong> advertising. Such a phenomenon is stronger when<br />
a newspaper is competing with other newspapers, <strong>and</strong> a low price reduces<br />
the readers <strong>of</strong> competing newspapers <strong>and</strong> the value <strong>of</strong> advertising on these<br />
competing newspapers.<br />
Other two-sided platforms include other media networks as television<br />
channels, real estate agencies, traditional auction houses, shopping malls,<br />
night clubs, payment card systems, telephone networks <strong>and</strong> many industries<br />
<strong>of</strong> the New Economy as those related with video game consoles, smart phones,<br />
digital music, PDAs, i-Mode, search engines (Google), on line communication<br />
(Yahoo! <strong>and</strong> Skype), on line social networks (MySpace, asmallworld, or Second<br />
Life), on line academic articles (JSTORE or SSRN) <strong>and</strong> on line shopping<br />
(Amazon <strong>and</strong> eBay). In many <strong>of</strong> these markets, multi-homing on at least one<br />
<strong>of</strong> the two sides is common: people <strong>of</strong>ten buy more than one journal or watch<br />
more TV channels (as companies advertise on multiple medias), hold multiple<br />
credit cards (as merchants accept multiple cards) <strong>and</strong> s<strong>of</strong>tware developers<br />
prepare applications for multiple OSs (while individuals typically use only<br />
one).<br />
In each one <strong>of</strong> these examples, network externalities are crucial to the<br />
success <strong>of</strong> a s<strong>of</strong>tware platform, <strong>and</strong> the pricing structure toward buyers <strong>and</strong><br />
sellers is crucial to the creation <strong>of</strong> these network effects. In particular, a platform<br />
typically ends up charging one <strong>of</strong> the two sides less than the other, taking<br />
into account dem<strong>and</strong> elasticities <strong>and</strong> which side values the other side more:<br />
the aim is to get on board as many agents as possible from one side, so as to<br />
increase the value <strong>of</strong> the platform for the other side <strong>and</strong> earn more revenue<br />
from it. For instance, when the price is the strategic variable, it is optimal to<br />
charge the side whose dem<strong>and</strong> is more elastic, because this allows one to maximize<br />
the total volume <strong>of</strong> interactions. 8 Prices will be constrained downward<br />
7 See Section 2.9 for a theoretical analysis <strong>of</strong> this aspect. An early contribution on<br />
two-sided markets is due to Baxter (1983).<br />
8 Consider the simplest case <strong>of</strong> a monopolistic platform charging a group, say the<br />
buyers, a price p B per interaction with the other group, say the sellers, <strong>and</strong><br />
charging the sellers a price p S per interaction with the buyers. If total dem<strong>and</strong>
214 6. Micros<strong>of</strong>t Economics<br />
when there are competing platforms (especially in the case <strong>of</strong> multi-homing),<br />
<strong>and</strong> further bias may emerge for strategic reasons, 9 but the general principles<br />
on the balanced price structure between the two sides remain unchanged. In<br />
extreme cases, one side may even receive its goods or its services for free or<br />
even be subsidized so as to maximize earnings from the other side.<br />
The above theoretical results are fully confirmed by what happens in<br />
the above mentioned two-sided markets, whose companies typically settle on<br />
pricing structures that are heavily skewed toward one side <strong>of</strong> the market or, in<br />
other words, adopt what is sometimes called a “divide <strong>and</strong> conquer” strategy.<br />
Newspapers, television networks <strong>and</strong> even websites typically earn more from<br />
advertisers than from consumers, real estate agencies earn more from sellers<br />
(or from l<strong>and</strong>lords) than from buyers (or renters), auction houses from sellers<br />
rather than from the buyers, shopping malls from stores rather than from the<br />
shoppers, night clubs from men rather than from women <strong>and</strong> payment card<br />
companies from merchants rather than from cardholders. Similarly, phone<br />
operators earn more from originating calls rather than from receiving ones,<br />
video game platforms from royalties on game developers rather than from<br />
<strong>of</strong> interactions is D(p B ) for the buyers <strong>and</strong> D(p S ) for the sellers, the number <strong>of</strong><br />
interactions is D(p B)D(p S). Given a marginal cost per interaction c the pr<strong>of</strong>its<br />
<strong>of</strong> the platform are:<br />
π =(p B + p S − c)D(p B )D(p S )<br />
whose maximization provides the following Rochet-Tirole (2003) rule p B + p S −<br />
c = p B / B = p S / S ,where i is the elasticity <strong>of</strong> dem<strong>and</strong> for i = B,S. Similar<br />
outcomes emerge in case dem<strong>and</strong> on each side depends on dem<strong>and</strong> on the other<br />
side, with more complex pricing structures <strong>and</strong> with competition between platforms<br />
(see Armstrong, 2006, Rochet <strong>and</strong> Tirole, 2006, <strong>and</strong> Goldfain <strong>and</strong> Kováč,<br />
2007).<br />
9 Strategic reasons may bias the pricing structure <strong>of</strong> platform leaders. In the example<br />
<strong>of</strong> the previous footnote, suppose product differentiation on one side occurs<br />
<strong>and</strong>, with the usual notation, pr<strong>of</strong>its <strong>of</strong> a representative firm are:<br />
π =(p B + p S − c)D(p B )D(p S ,β S )<br />
Suppose that the leader can commit to a price for the buyers, while, for simplicity,<br />
the others are given. Firms compete on the prices for the sellers. To verify the<br />
incentives <strong>of</strong> the leader, notice that, with the usual notation <strong>of</strong> Chapter 2 (with<br />
k =1/p B as preliminary commitment), we have:<br />
Π L 13(1/p S,β S , 1/p B)=(p Sp B) 2 D(p B)D 1(p S,β S ) < 0<br />
Assuming that SC holds, this implies that when entry is not free the leader will<br />
tend to underprice buyers to be accomodating in the competition for the sellers.<br />
However, when entry in the platform competition is endogenous, the leader will<br />
tend to overprice buyers to be aggressive in the competition for the sellers.
6.1 The S<strong>of</strong>tware <strong>Market</strong> 215<br />
buyers <strong>of</strong> consoles (that are <strong>of</strong>ten sold below cost), while most <strong>of</strong> the other<br />
s<strong>of</strong>tware platforms, including PC OSs, earn more from end users rather than<br />
from s<strong>of</strong>tware developers. 10<br />
Notice that, in spite <strong>of</strong> the network effects, most <strong>of</strong> these two-sided markets<br />
are also characterized by a certain degree <strong>of</strong> fragmentation between platform<br />
providers, <strong>of</strong>ten associated with a certain degree <strong>of</strong> differentiation. Only<br />
when technological innovation is particularly important <strong>and</strong> fixed costs <strong>of</strong><br />
investment in R&D are high, while marginal costs <strong>of</strong> production are particularly<br />
low, the number <strong>of</strong> competing platforms is endogenously reduced, as in<br />
the above mentioned markets <strong>of</strong> the New Economy. Nevertheless, tipping on<br />
a single leader rarely happens, especially when product differentiation <strong>and</strong><br />
multi-homing have a role, as for video games. And even in these cases competition<br />
for the market can be quite effective <strong>and</strong> induce periods <strong>of</strong> persistent<br />
leadership with occasional replacement <strong>of</strong> the leader: pathbreaking innovations<br />
(or “killer applications”) are what competitive firms really look for. For<br />
instance, in the console video game industry, sequential innovations brought<br />
to leadership a number <strong>of</strong> companies as Atari (that reached 80% share <strong>of</strong> the<br />
market in 1980), Nintendo (90% <strong>of</strong> the market in 1987), Sega (leader in the<br />
early 90s), Nintendo again (in the mid 90s) <strong>and</strong> Sony with the PlayStation in<br />
different improved versions (during the last decade): recently Micros<strong>of</strong>t Xbox<br />
started gaining market shares, <strong>and</strong> Nintendo is still active, but the leadership<br />
<strong>of</strong> Sony (65% market share in 2004) does not appear under threat yet,<br />
especially after the recent successful launch <strong>of</strong> PlayStation 3. Similarly, after<br />
a number <strong>of</strong> unsuccessful attempts by many companies, Palm’s PDA gained<br />
success <strong>and</strong> leadership in the market for OSs for organizers thanks to a simple<br />
h<strong>and</strong>writing recognition system (65% market share in 2000) until Micros<strong>of</strong>t<br />
competing platform <strong>and</strong> other h<strong>and</strong>held devices, including Blackberry <strong>and</strong><br />
(in perspective) Apple’s iPhone, started gaining success.<br />
Having described the role <strong>of</strong> network effects <strong>and</strong> multi-sided relations, it<br />
is now time to return to the s<strong>of</strong>tware market, where these elements play a<br />
crucial role.<br />
6.1.3 Micros<strong>of</strong>t<br />
Micros<strong>of</strong>t was founded in 1975 by Bill Gates <strong>and</strong> Paul Allen to develop BA-<br />
SIC interpreters for the first PC, Altair 8800, <strong>and</strong> then other programming<br />
languages. Only later, did it start producing major s<strong>of</strong>tware programs. In<br />
1981, Micros<strong>of</strong>t released its first operating system, MS-DOS, which had a<br />
10 This happens in different ways however: Micros<strong>of</strong>t licenses Windows, Palm <strong>and</strong><br />
Symbian license their OSs to manufacturers <strong>of</strong> PCs, PDAs <strong>and</strong> cellular phones,<br />
while RealNetworks licenses access to digital content <strong>and</strong> Apple sells PCs <strong>and</strong><br />
iPods, but none <strong>of</strong> these companies charges content owners (Apple <strong>and</strong> RealNetworks<br />
actually pay them) or s<strong>of</strong>tware developers (which are typically subsidized).
216 6. Micros<strong>of</strong>t Economics<br />
character-based user interface that required users to type specific instructions<br />
to perform tasks. In 1985, Micros<strong>of</strong>t introduced a new product called<br />
Windows that included a graphical user interface, enabling users to perform<br />
tasks by clicking on icons on the screen using a pointing device called a<br />
mouse (basically the only piece <strong>of</strong> hardware produced by Micros<strong>of</strong>t for PCs).<br />
Windows 3.0, shipped in 1990, was the first commercially successful version<br />
<strong>of</strong> Windows. In 1995, Micros<strong>of</strong>t released Windows 95, which integrated the<br />
functionality <strong>of</strong> Windows 3.1 <strong>and</strong> MS-DOSinasingleoperatingsystem.In<br />
2000, Micros<strong>of</strong>t shipped Windows 2000 Pr<strong>of</strong>essional, a new generation <strong>of</strong> PC<br />
operating system built on a more stable <strong>and</strong> reliable s<strong>of</strong>tware code base than<br />
earlier versions <strong>of</strong> Windows. Windows XP represented a further evolution<br />
with a range <strong>of</strong> added functionality for both business <strong>and</strong> home users. In<br />
2007 Windows Vista has been released worldwide: it was the fruit <strong>of</strong> five<br />
years <strong>of</strong> work by eight thous<strong>and</strong> designers, programmers <strong>and</strong> testers <strong>and</strong> <strong>of</strong><br />
an estimated investment <strong>of</strong> $ 10 billion to rewrite from scratch a new code.<br />
This impressive effort was probably related to the competitive pressure coming<br />
from the open source community, which is strongly supported by many<br />
large corporations willing to strengthen valid alternatives to Windows.<br />
Even if complete <strong>and</strong> homogenous data are unavailable, consistent evidence<br />
suggests that the market share <strong>of</strong> Windows on sales <strong>of</strong> OSs for PCs<br />
rapidly increased towards 80% in the first half <strong>of</strong> the 90s to gradually arrive<br />
at 92% in 1996, 94% in 1997, 95% in 1999, 96% 2001, <strong>and</strong> remained above<br />
90% since then (while the average consumer price <strong>of</strong> Windows, calculated as<br />
average revenue per licence to PC manufacturers based on Micros<strong>of</strong>t sales,<br />
remained around $ 44-45). Nevertheless, one should keep in mind that Linux,<br />
after having made inroads into corporations’ server computers, is now exp<strong>and</strong>ing<br />
into a much broader market, that <strong>of</strong> employees’ PCs, that a minor<br />
group <strong>of</strong> PC users (but strongly increasing in number, especially between expert<br />
users) downloads open source OSs from the Internet, 11 <strong>and</strong> that on the<br />
11 Estimates for the percentage <strong>of</strong> server computers running Linux worldwide are<br />
in the range <strong>of</strong> 20-25%, while desktop computers running Linux are around 3%.<br />
According to the Wall Street Journal (March13, 2007, Linux Starts to Find Home<br />
on Desktops), “market researcher IDC said licenses <strong>of</strong> both free <strong>and</strong> purchased<br />
versions <strong>of</strong> Linux s<strong>of</strong>tware going into PCs world-wide rose 20.8% in 2006 over the<br />
previous year <strong>and</strong> forecast that licenses will increase 30% this year over last. That<br />
compares with 10.5% growth in 2004, according to IDC. Whether Linux gains a<br />
stronger footing in PCs depends partly on whether PC makers start supporting<br />
it more strongly. To date, neither Dell Inc.norHewlett-PackardCo.have<strong>of</strong>fered<br />
PCs preloaded with Linux. But Dell has been soliciting input from its customers<br />
to help guide its plans for Linux — which some industry observers say could lead<br />
the company to start making Linux PCs [...] H-P says it has recently signed deals<br />
— on an ad hoc, custom basis — to provide Linux PCs to large customers.”
6.1 The S<strong>of</strong>tware <strong>Market</strong> 217<br />
top <strong>of</strong> this market there are Apple computers running Mac OS. 12 It is clear<br />
that Micros<strong>of</strong>t has reached a robust leadership in the PC operating systems<br />
market for Intel-compatible computers. In line with our previous discussion,<br />
Evans et al. (2006) state four key strategies that have driven Micros<strong>of</strong>t to<br />
become the leader <strong>of</strong> the PC industry:<br />
“(1) <strong>of</strong>fering lower prices to users than its competitors; (2) intensely<br />
promoting API-based s<strong>of</strong>tware services to developers; (3) promoting<br />
the development the development <strong>of</strong> peripherals, sometimes<br />
through direct subsidies, in order to increase the value <strong>of</strong> the Windows<br />
platform to developers <strong>and</strong> users; <strong>and</strong> (4) continually developing<br />
s<strong>of</strong>tware services that provide value to developers directly <strong>and</strong> to end<br />
users indirectly.”<br />
Beyond OSs, Micros<strong>of</strong>t is the leader in other markets for s<strong>of</strong>tware applications.<br />
Some essential applications have been freely bundled with the operating<br />
system: for instance a basic word processing s<strong>of</strong>tware (WordPad), a<br />
browser to access Internet (Internet Explorer) <strong>and</strong> media player functionality<br />
(Windows Media Player) have been gradually added for free to subsequent<br />
versions <strong>of</strong> Windows when they became st<strong>and</strong>ard components <strong>of</strong> a modern<br />
OS. Other more sophisticated applications are supplied separately. Most notablythisisthecase<strong>of</strong>theOffice<br />
Suite consisting <strong>of</strong> the word processor<br />
Word (first edition released in 1983), the spreadsheet Excel (1985), the presentation<br />
s<strong>of</strong>tware PowerPoint (1987) <strong>and</strong> more. The main two applications,<br />
Word <strong>and</strong> Excel, have been successfully competing against alternative products<br />
like WordPerfect, WordStar, AmiPro <strong>and</strong> others on one side <strong>and</strong> Lotus<br />
1-2-3, Quattro <strong>and</strong> others on the other side. Liebowitz <strong>and</strong> Margolis (1999)<br />
have shown convincing evidence for which a better quality-price ratio together<br />
with network effects were at the basis <strong>of</strong> this evolution (it is important to note<br />
that Micros<strong>of</strong>t achieved leadership in the Macintosh market, hence without<br />
exploiting the presence <strong>of</strong> its own OS, considerably earlier than in the PC<br />
market).<br />
In the market for word processing applications, Micros<strong>of</strong>t’s market share<br />
was hardly above 10% at the end <strong>of</strong> the 80s, but gradually increased to 28%<br />
in 1990, 40% in 1991, 45% in 1992, 50% in 1993, 65% in 1994, 79% in 1995,<br />
90% in 1996, 94% in 1997 <strong>and</strong> arrived to 95% in 1998, remaining around this<br />
level afterward. Meanwhile the average consumer price <strong>of</strong> Word (calculated as<br />
average revenue per license) decreased from $ 235 in 1988 to $ 39 in 2001. In<br />
the market for spreadsheet applications, Micros<strong>of</strong>t followed a similar progress,<br />
with a market share <strong>of</strong> 18% in 1990, 34% in 1991, 43% in 1992, 46% in 1993,<br />
68% in 1994, 77% in 1995, 84% in 1996, 92% in 1997 <strong>and</strong> 94% in 1998, with<br />
12 As pointed out by Foncel <strong>and</strong> Ivaldi (2005), there is a certain variability between<br />
countries. The DOS/Windows platform is on 88 % <strong>of</strong> PC sales in Japan <strong>and</strong><br />
98% in Germany. Data for the second half <strong>of</strong> the 90s are from International Data<br />
Corporation.
218 6. Micros<strong>of</strong>t Economics<br />
minor progress in the following years, while the average consumer price <strong>of</strong><br />
Excel was decreasing from $ 249 in 1988 to $ 42 in 2001.<br />
Finally, Micros<strong>of</strong>t is also active in other strategic markets as a follower,<br />
in particular with the personal finance s<strong>of</strong>tware Money (the leader being<br />
Intuit Quicken), the operating systems for smart phones Windows Mobile<br />
(the leader being Symbian, with a 60% market share in 2004), the video<br />
game console Xbox (the leader being Sony PlayStation, with a 65% market<br />
share in 2004), the search engine based portal Windows Live (the leader being<br />
Google, with more than 80% <strong>of</strong> searches on the Internet) <strong>and</strong> more. In 2006,<br />
Micros<strong>of</strong>t, led by the CEO Steve Ballmer, had revenues <strong>of</strong> $ 44.2 billion, 60%<br />
<strong>of</strong> which derives from Windows <strong>and</strong> Office, <strong>and</strong> net income <strong>of</strong> $ 12.4 billion,<br />
80-90% <strong>of</strong> which derives from Windows <strong>and</strong> Office.<br />
6.2 The <strong>Antitrust</strong> Cases<br />
Micros<strong>of</strong>t’s leading position induced large opposition in the industry <strong>and</strong> the<br />
emergence <strong>of</strong> multiple antitrust cases with importance at the global level. 13<br />
Micros<strong>of</strong>t has been under investigations in the US by the Federal Trade Commission<br />
<strong>and</strong> the Department <strong>of</strong> Justice since 1990, primarily for its contracts<br />
with computer manufacturers <strong>and</strong> for bundling secondary products with its<br />
OSs. 14 However, the most important US case began only in the late ’90s under<br />
the Democratic Clinton Administration, followed after a few years by the<br />
EU case.<br />
6.2.1 The US Case<br />
In the main Micros<strong>of</strong>t vs. US case, started in 1998, the s<strong>of</strong>tware company was<br />
accused <strong>of</strong> protecting its monopoly in the OS market from the joint threat<br />
<strong>of</strong> the Internet browser Netscape Navigator <strong>and</strong> the Java programming language,<br />
15 which could have developed a potential substitute for OSs allowing<br />
13 This section is partly derived from Etro (2006d).<br />
14 Alreadyinthemid90swecouldseeimportanteconomistsinactionintheseearly<br />
cases. In 1995 the Nobel prize Kenneth Arrow intervened saying that “Micros<strong>of</strong>t<br />
appearstohaveachieveditsdominantpositioninitsmarketasaconsequence<br />
<strong>of</strong> good fortune <strong>and</strong> possibly superior products <strong>and</strong> business acumen” <strong>and</strong> that<br />
Micros<strong>of</strong>t’s licensing practices toward original equipment manufacturers “made<br />
only a minor contribution to the growth <strong>of</strong> Micros<strong>of</strong>t’s installed base. Even this<br />
minor contribution overstates the impact <strong>of</strong> Micros<strong>of</strong>t’s licensing practices on its<br />
installed base barrier to the entry <strong>and</strong> growth <strong>of</strong> competing operating systems”<br />
(Declaration <strong>of</strong> Kenneth Arrow, U.S..v.Micros<strong>of</strong>tCorp., Civil Action No. 94-<br />
1564 (SS), January 17, p. 11-12).<br />
15 The dramatic expansion <strong>of</strong> the World Wide Web started in 1993 after the development,<br />
by a team from the University <strong>of</strong> Illinois, <strong>of</strong> the first graphical web
6.2 The <strong>Antitrust</strong> Cases 219<br />
s<strong>of</strong>tware applications to run on hardware independently from the desktop<br />
OS. Basically, the hypothetical threat for Micros<strong>of</strong>t was the development <strong>of</strong><br />
an alternative to the s<strong>of</strong>tware platform based on the OS, a sort <strong>of</strong> middleware<br />
platform or a web-based platform leading to the “commoditization” <strong>of</strong><br />
the OS (as ten years before the s<strong>of</strong>tware platform led to the commoditization<br />
<strong>of</strong> hardware), <strong>and</strong> hence to the loss <strong>of</strong> leadership <strong>of</strong> Micros<strong>of</strong>t. Micros<strong>of</strong>t<br />
reacted by improving its Internet Explorer (IE) browser, engaging in contractual<br />
agreements with computer manufacturers <strong>and</strong> Internet service providers<br />
to promote preferential treatment for IE (notably AOL, whose “You’ve got<br />
mail” sound track was attracting more than 20 millions Americans at the<br />
time), <strong>and</strong> finally tying Windows with IE. For perspectives by economists<br />
who were active in the case see Fisher <strong>and</strong> Rubinfeld (2001) <strong>and</strong> Bresnahan<br />
(2001) on the side <strong>of</strong> US Government, <strong>and</strong> the essays in Evans (2002),<br />
especially Elzinga et al. (2002), on the opposite side. 16<br />
As Klein (2001) pointed out in an academic survey on the Journal <strong>of</strong><br />
Economic Perspectives (Symposium on the Micros<strong>of</strong>t case), “Micros<strong>of</strong>t spent<br />
hundreds <strong>of</strong> millions <strong>of</strong> dollars developing an improved version <strong>of</strong> its browser<br />
s<strong>of</strong>tware <strong>and</strong> then marketed it aggressively, most importantly by integrating<br />
it into Windows, pricing it at zero <strong>and</strong> paying online service providers <strong>and</strong><br />
personal computer manufacturers for distribution. All <strong>of</strong> this was aimed at<br />
increasing use <strong>of</strong> Micros<strong>of</strong>t’s Internet Explorer browser technology, both by<br />
end users <strong>and</strong> s<strong>of</strong>tware developers, to blunt Netscape’s threat to the dominance<br />
by Windows <strong>of</strong> the market for personal computer operating systems.”<br />
Micros<strong>of</strong>t’s investments in browser technology, which largely improved IE until<br />
it became a superior product compared to Netscape Navigator (see the<br />
empirical analysis in Liebowitz <strong>and</strong> Margolis, 1999), <strong>and</strong> Micros<strong>of</strong>t’s pricing<br />
<strong>of</strong> IE at zero (as always since then) appear to us as examples <strong>of</strong> aggressive<br />
strategic investment <strong>and</strong> aggressive pricing by a market leader facing competition,<br />
<strong>and</strong> not as anti-competitive strategies. 17 According to Klein (2001),<br />
browser, Mosaic. Netscape, founded in 1994, hired most <strong>of</strong> its developers to create<br />
Navigator. Java was developed at the same time by Sun as a middleware<br />
product to allow programmers to write applications that would run on line on<br />
any computer regardless <strong>of</strong> the underlying OS.<br />
16 For economic surveys on the case see Gilbert <strong>and</strong> Katz (2001), Klein (2001) <strong>and</strong><br />
Economides (2001). For retrospective views see Motta (2004, Ch. 7) <strong>and</strong> Evans<br />
et al. (2005).<br />
17 In our view, the US case was characterized by a too limited focus on rigorous<br />
economic arguments in support <strong>of</strong> the different thesis. It is ironic that Micros<strong>of</strong>t’s<br />
internal documents <strong>and</strong> emails including aggressive expressions toward competitors<br />
were used to support the idea that Micros<strong>of</strong>t undertook its browser development<br />
for entry deterrence purposes. It is hard to see how the aggressive language<br />
<strong>of</strong> business people can prove more than competitive intent (on the use <strong>of</strong> internal<br />
documents to prove antitrust violations, see Manne <strong>and</strong> Williamson, 2005).
220 6. Micros<strong>of</strong>t Economics<br />
“a crucial condition for anticompetitive behavior in such cases is<br />
that the competitive process is not open. In particular, we should be<br />
concerned only if a dominant firm abuses its market power in a way<br />
that places rivals at a significant competitive disadvantage without<br />
any reasonable business justification. Only under these circumstances<br />
can more efficient rivals be driven out <strong>of</strong> the market <strong>and</strong> consumers<br />
not receive the full benefits <strong>of</strong> competition for dominance. The only<br />
Micros<strong>of</strong>t conduct ... that may fit this criteria for anticompetitive<br />
behavior are the actions Micros<strong>of</strong>t took in obtaining browser distribution<br />
through personal computer manufacturers”<br />
This is correct: a number <strong>of</strong> contractual restraints imposed by Micros<strong>of</strong>t<br />
on its distributors were potentially harmful <strong>and</strong> have been correctly forbidden.<br />
After a failed attempt by Judge Richad Posner to mediate in settlement<br />
negotiations, Judge Thomas Penfield Jackson decided to impose heavy behavioral<br />
<strong>and</strong> structural remedies on Micros<strong>of</strong>t, including the break up in an<br />
operating system <strong>and</strong> an application company (the so-called “Baby Bills”, as<br />
Baby Bells were the companies derived from the 1984 break up <strong>of</strong> AT&T).<br />
At the time, this draconian remedy was criticized by many economists with<br />
different perspectives on the case, for excessively penalizing the company<br />
without a clear relation between the punishment <strong>and</strong> the alleged crime, <strong>and</strong><br />
for inducing perverse consequences for consumers. For instance, on the pages<br />
<strong>of</strong> The New York Times, Paul Krugman pointed out the risk <strong>of</strong> creating two<br />
monopolists engaging in double marginalization:<br />
“The now ‘naked’ operating-system company would ab<strong>and</strong>on its<br />
traditional pricing restraints <strong>and</strong> use its still formidable monopoly<br />
power to charge much more. And at the same time applications s<strong>of</strong>tware<br />
that now comes free would also start to carry heftly price tags”<br />
(Krugman, 2000a). 18<br />
18 Judge Jackson was later disqualifiedforviolatinganumber<strong>of</strong>ethicalprecepts<br />
<strong>and</strong> being manifestly biased against Micros<strong>of</strong>t. The government proposal <strong>of</strong> splitting<br />
Micros<strong>of</strong>t into two companies, which was adopted by the Judge without<br />
substantive changes, had been supported by declarations <strong>of</strong> important economists,<br />
including Paul Romer <strong>and</strong> Carl Shapiro. For instance, Shapiro declared<br />
that, while “network monopolies can be very strong, they are most vulnerable to<br />
attack by firms with a strong position in the provision <strong>of</strong> a widely-used complementary<br />
product”, hence “the proposed reorganization <strong>of</strong> Micros<strong>of</strong>t into separate<br />
applications <strong>and</strong> operating systems businesses will lower entry barriers, encourage<br />
competition <strong>and</strong> promote innovation” (Declaration <strong>of</strong> Carl Shapiro, U.S. v.<br />
Micros<strong>of</strong>t Corp., Civil Action N0. 98-1232 (TPJ), p. 7 <strong>and</strong> p 29). Nevertheless,<br />
other economists were critical <strong>of</strong> the consequences <strong>of</strong> the break up on innovation.<br />
Krugman (2000b) again criticized a systematic reference to the promotion<br />
<strong>of</strong> innovation as a vague justification for the remedy: “we don’t know very much
6.2 The <strong>Antitrust</strong> Cases 221<br />
After the appeal phase <strong>and</strong> the return <strong>of</strong> the Republican Administration<br />
with George W. Bush, the DOJ changed attitude looking for a settlement.<br />
The November 2002 ruling <strong>of</strong> the District <strong>of</strong> Court decided on behavioral<br />
remedies aimed at preventing Micros<strong>of</strong>t from adopting exclusionary strategies<br />
against firms challenging its market power in the market for OSs. Moreover,<br />
the Court adopted forward looking remedies that required limited disclosure<br />
<strong>of</strong> APIs, communication protocols, <strong>and</strong> related technical information in order<br />
to facilitate interoperability, <strong>and</strong> created a system <strong>of</strong> monitoring <strong>of</strong> Micros<strong>of</strong>t’s<br />
compliance which has been working quite well in the last years. Since other<br />
derivative private actions have also been dismissed or settled, it seems that<br />
this long-st<strong>and</strong>ing conflicthasarrivedtoitsendintheUS.<br />
6.2.2 The EU Case<br />
The Micros<strong>of</strong>t vs. EU case was subsequently developed on somewhat similar<br />
issues. In particular, Micros<strong>of</strong>t has been accused <strong>of</strong> abuse <strong>of</strong> dominance in<br />
the market for OSs through technological leveraging <strong>and</strong> in particular in two<br />
ways: first, by bundling Windows with Media Player, a s<strong>of</strong>tware for downloading<br />
audio/video content, <strong>and</strong>, second, by refusing to supply competitors with<br />
the interface information needed to achieve interoperability between work<br />
group server OSs 19 <strong>and</strong> Windows. Contrary to the US case, the bundling<br />
part <strong>of</strong> the EU case is a traditional case <strong>of</strong> bundling, since the competitors in<br />
the secondary market, notably RealNetworks, do not represent a threat for<br />
Windows, the primary product <strong>of</strong> Micros<strong>of</strong>t.<br />
In the famous antitrust decision <strong>of</strong> March 24, 2004, <strong>Competition</strong> Commissioner<br />
Mario Monti imposed on Micros<strong>of</strong>t the largest fine in the history<br />
<strong>of</strong> antitrust (€ 497 million), required Micros<strong>of</strong>t to issue a version <strong>of</strong> its Windows<br />
operating system without Media Player, <strong>and</strong> m<strong>and</strong>ated the licensing <strong>of</strong><br />
intellectual property to enable interoperability between Windows PCs <strong>and</strong><br />
about what promotes innovation, <strong>and</strong> even some <strong>of</strong> what we think we know may<br />
not be true. For example, advocates <strong>of</strong> the breakup <strong>of</strong> Micros<strong>of</strong>t like to point<br />
to the breakup <strong>of</strong> AT&T, which everyone thinks was purely positive in its effects<br />
on innovation. It’s a bad parallel in many ways, but still it is interesting<br />
to notice that next-generation telecommunications is not yet a hot issue in the<br />
United States, because thanks to the fragmentation <strong>of</strong> our cellular system we<br />
are lagging well behind Europe <strong>and</strong> Japan in mobile phone technology. And that<br />
fragmentation is in part a legacy <strong>of</strong> the AT&T breakup. My point is not that it<br />
is wrong to consider the impact <strong>of</strong> policy on innovation; it is that because the<br />
determinants <strong>of</strong> innovation are not well understood, clever advocates can invoke<br />
technological progress as an all-purpose justification for whatever policy they<br />
favor”.<br />
19 A work group server OS is a s<strong>of</strong>tware providing services to share files <strong>and</strong> printers<br />
<strong>and</strong> other administration services to a group <strong>of</strong> users connected in a network,<br />
typically in <strong>of</strong>fice environments.
222 6. Micros<strong>of</strong>t Economics<br />
work group servers on one side, <strong>and</strong> competitor products on the other side.<br />
After this decision, Micros<strong>of</strong>t paid the fine, developed <strong>and</strong> released a version<br />
<strong>of</strong> Windows without Media Player, <strong>and</strong> entered into extensive discussions<br />
with the Commission about the implementation <strong>of</strong> the remedies concerning<br />
interoperability. In the original decision this required to prepare a complete<br />
<strong>and</strong> accurate interface documentation describing portions <strong>of</strong> Micros<strong>of</strong>t server<br />
operating system s<strong>of</strong>tware <strong>and</strong> to license innovations created by Micros<strong>of</strong>t<br />
under “reasonable <strong>and</strong> non discriminatory” (so-called RAND) terms to competitors.<br />
These imply that the royalties should be set at levels that enable<br />
use by other developers in a commercially practicable way with reference to<br />
st<strong>and</strong>ard valuation techniques, to an assessment <strong>of</strong> whether the protocols are<br />
innovative, <strong>and</strong> with reference to market rates for comparable technologies.<br />
Over time, the new <strong>Competition</strong> Commissioner Neelie Kroes has continued<br />
to extend the scope <strong>of</strong> the information required, from information that<br />
would enable interoperability with Windows PCs <strong>and</strong> servers for the purpose<br />
<strong>of</strong> creating new products for which there is unmet consumer dem<strong>and</strong>, to<br />
information that would allow a competitor to produce clones or “drop-in replacements”<br />
<strong>of</strong> the Windows server OS. Even more controversially, the Commission’s<br />
<strong>Competition</strong> Directorate-General has sought to loosen the terms<br />
under which Micros<strong>of</strong>t would be able to licence its information, so as to allow<br />
products implementing its technical specifications to be released under<br />
so-called Open Source licences (DG <strong>Competition</strong> was prepared to make an<br />
exception for technologies that involved an inventive step <strong>and</strong> were considered<br />
novel by comparison with the prior art, thus meeting the criteria for<br />
patentability). Such release, by revealing to the world Micros<strong>of</strong>t’s own implementations<br />
<strong>of</strong> its technical specifications, would irreparably undermine<br />
the trade secret protection to which these technologies, some <strong>of</strong> which are<br />
not patented, are subject. In a further shift, the Commission made clear in<br />
Spring <strong>of</strong> 2007 that it expected Micros<strong>of</strong>t to forego royalty payments on any<br />
technologies that were not covered by patents. With the compliance process<br />
made more difficult on both sides by the technical complexity <strong>of</strong> the material<br />
<strong>and</strong> key policy differences (e.g. over the intellectual property issues),<br />
DG <strong>Competition</strong> challenged Micros<strong>of</strong>t to comply with the interoperability<br />
remedy by 15 December 2005, on pain <strong>of</strong> massive penalty payments for noncompliance.<br />
In early 2006, Micros<strong>of</strong>t provided further information needed for<br />
interoperability purposes, <strong>and</strong> even made available to its competitors selective<br />
access to the source code <strong>of</strong> Windows. Nevertheless, in July 2006 the<br />
Commission levied fines <strong>of</strong> € 1.5 million a day from the December hearing<br />
onwards (for a total <strong>of</strong> other € 280.5 million), <strong>and</strong> threatened to double the<br />
fine if the company did not comply. At the time <strong>of</strong> writing, the case is still<br />
unresolved: Micros<strong>of</strong>t’s Appeal <strong>of</strong> the Commission’s 2004 l<strong>and</strong>mark decision
6.3 Is Micros<strong>of</strong>t a Monopolist? 223<br />
was heard by the European Court <strong>of</strong> First Instance in April 2006, 20 <strong>and</strong> a<br />
decision is expected by September 2007, after this book will be completed.<br />
In either case, both Micros<strong>of</strong>t <strong>and</strong> the Commission may also appeal to the<br />
European Court <strong>of</strong> Justice, which is the EU’s highest court. 21<br />
A common element in both the US <strong>and</strong> EU cases has been the substantial<br />
involvement <strong>of</strong> competitors <strong>of</strong> Micros<strong>of</strong>t on the side <strong>of</strong> the antitrust authorities.<br />
In a neat article about the US vs. Micros<strong>of</strong>t case on Business Week,<br />
Robert Barro noticed that:<br />
“a sad sidelight in the Micros<strong>of</strong>t case is the cooperation <strong>of</strong> its<br />
competitors, Netscape, Sun <strong>and</strong> Oracle Corp., with the government.<br />
One might have expected these robust innovators to rise above the<br />
category <strong>of</strong> whiner corporations [...] The real problem is that whining<br />
can sometimes be pr<strong>of</strong>itable, because the political process makes it<br />
so. The remedy requires a shift in public policies to provide less<br />
reward for whining. The bottom line is that the best policy for the<br />
government in the computer industry is to stay out <strong>of</strong> it” (Barro,<br />
1998)<br />
Nevertheless,IBM,Sun,Oracle,Novell<strong>and</strong>thewideopensourcemovement<br />
have also been active against Micros<strong>of</strong>t in the EU case.<br />
6.3 Is Micros<strong>of</strong>t a Monopolist?<br />
While a comprehensive analysis on the PC operating system market <strong>and</strong><br />
<strong>of</strong> the role <strong>of</strong> Micros<strong>of</strong>t is beyond our scope, we can try to provide a basic<br />
interpretation <strong>of</strong> a few features <strong>of</strong> this market through the simple ideas developed<br />
in the theoretical part <strong>of</strong> this book. The technological conditions in<br />
the s<strong>of</strong>tware market are relatively simple. Production <strong>of</strong> an operating system,<br />
as any other s<strong>of</strong>tware, takes a very high up-front investment <strong>and</strong> a roughly<br />
constant <strong>and</strong> low (close to zero) marginal cost. 22 Dem<strong>and</strong> conditions are<br />
20 Also in this case important economists played a crucial role: for instance with<br />
Joseph Stiglitz on the European Commission side <strong>and</strong> David Evans on the Micros<strong>of</strong>t<br />
side.<br />
21 Another antitrust case focused on bundling issues has taken place in South Korea:<br />
in 2005 Micros<strong>of</strong>t had to pay a fine <strong>of</strong> $ 32 million <strong>and</strong> produce more than<br />
one version <strong>of</strong> Windows for the country (one with Windows Media Player <strong>and</strong><br />
Windows Messenger <strong>and</strong> one without them).<br />
22 However, notice that the initial fixed costs <strong>of</strong> production are not independent<br />
from the future scale <strong>of</strong> production: higher production requires a better product,<br />
which requires higher R&D investments. Moreover, diminishing returns to scale<br />
are typical in the production <strong>of</strong> new s<strong>of</strong>tware, which may explain why the price<br />
<strong>of</strong> s<strong>of</strong>tware has not been declining as much as the price <strong>of</strong> hardware in the last<br />
decades.
224 6. Micros<strong>of</strong>t Economics<br />
more complex. What drives dem<strong>and</strong> is not the traditional concept <strong>of</strong> product<br />
differentiation, which is <strong>of</strong> course present, but the development <strong>of</strong> network<br />
externalities: network effects are crucial in the development <strong>of</strong> a market for<br />
an OS <strong>and</strong> the pricing structure is fundamental to get on board both end<br />
users <strong>and</strong> application developers. Beyond this, a firm producing OSs faces<br />
competitors: the entry conditions in the market for OSs are quite debated,<br />
but there are good reasons to believe that even though entry into the s<strong>of</strong>tware<br />
market may entail large costs, it is substantially endogenous. First <strong>of</strong><br />
all, there are already many companies distributing OSs (for instance Solaris<br />
by Sun Microsystems, many versions <strong>of</strong> Unix <strong>and</strong> Linux, those by Red Hat<br />
<strong>and</strong> Novell), there are many firms producing OSs for related industries (smart<br />
phones, PDAs <strong>and</strong> videogames) which could be scaled-up to run on desktop<br />
computers (especially on low cost PCs), <strong>and</strong> there are even more potential<br />
entrants (think <strong>of</strong> the giants in adjacent sectors <strong>of</strong> the New Economy, hardware<br />
<strong>and</strong> telecommunications in particular). Second, it is hard to think <strong>of</strong><br />
a market which is more “global” than the s<strong>of</strong>tware market: dem<strong>and</strong> comes<br />
from all over the world, transport costs are virtually zero, <strong>and</strong> the knowledge<br />
required to build s<strong>of</strong>tware is accessible worldwide.<br />
Nevertheless, it has been claimed that in the market for OSs, the high<br />
number <strong>of</strong> applications developed by many different firms for Windows represents<br />
a substantial barrier to entry. It is probably true that high quality<br />
products make life harder to the competing products, but this should not lead<br />
to the conclusion that quality is a barrier to entry, especially in sectors where<br />
innovation should drive competition. Moreover, it is true that competitors<br />
need to <strong>of</strong>fer a number <strong>of</strong> st<strong>and</strong>ard <strong>and</strong> technologically mature applications<br />
upon entry to match the high quality <strong>of</strong> the Windows package <strong>and</strong> create<br />
network effects (<strong>and</strong> some do <strong>of</strong>fer many already), but the cost <strong>of</strong> <strong>of</strong>fering<br />
these applications is unlikely to be prohibitive compared to the global size <strong>of</strong><br />
this market. 23 There are at least two reasons for this. First, notice that the<br />
alleged “applications barrier to entry” is <strong>of</strong>ten erroneously associated with<br />
thous<strong>and</strong>s <strong>of</strong> applications written for Windows, while it is actually limited<br />
to a h<strong>and</strong>ful <strong>of</strong> applications such as word processing, spreadsheet, graphics,<br />
internet access <strong>and</strong> media player s<strong>of</strong>tware, which really satisfy the needs <strong>of</strong><br />
most active computer users (McKenzie, 2001). Second, the competitors <strong>of</strong><br />
Micros<strong>of</strong>t should not (<strong>and</strong> the existing ones do not) even finance the development<br />
<strong>of</strong> all the needed applications: they should just fund <strong>and</strong> encourage<br />
23 There are many examples <strong>of</strong> markets with network effects where subsequent<br />
entrants managed to create network effects <strong>and</strong> challenge incumbents. Within<br />
traditional sectors, examples abound in the fashion industry <strong>and</strong> many industries<br />
where creating new successful br<strong>and</strong>s requires building network effects. In the<br />
New Economy a clear example emerges in the case <strong>of</strong> payment cards (where<br />
network effects are crucial, to say the least), which absorbed sequential entry by<br />
Diners Club (1950), American Express (1958), Visa (1966), MasterCard (1966)<br />
<strong>and</strong> Discover (1985), to name the most famous actors <strong>of</strong> this market.
6.3 Is Micros<strong>of</strong>t a Monopolist? 225<br />
other firms to write applications for their OSs, or have old applications originally<br />
written for other OSs “ported to” theirs, which is what already happens<br />
since multi-homing is common practice on the side <strong>of</strong> s<strong>of</strong>tware developers. 24<br />
In essence, the s<strong>of</strong>tware market is characterized by network effects, high<br />
fixed costs <strong>of</strong> R&D, constant marginal costs <strong>of</strong> production close to zero <strong>and</strong><br />
substantially open access by competitors able to create new s<strong>of</strong>tware. According<br />
to the theory <strong>of</strong> market leaders, these are the ideal conditions under<br />
which we should expect a leader to produce for the whole market with very<br />
aggressive (low) prices. Hence, it should not be surprising that, at least in<br />
the market for operating systems, a single firm, Micros<strong>of</strong>t, has such a large<br />
market share. We can see the same fact from a different perspective: since<br />
entry into the s<strong>of</strong>tware market is endogenous, the leader has to keep prices<br />
low enough to exp<strong>and</strong> its market share to almost the whole market.<br />
6.3.1 Why Is the Price <strong>of</strong> Windows so Low?<br />
Many economists agree on the fact that Micros<strong>of</strong>t sells Windows at an extremely<br />
low price. For instance, Fudenberg <strong>and</strong> Tirole (2000) notice that<br />
both sides in the US Micros<strong>of</strong>t case admit that “Micros<strong>of</strong>t’s pricing <strong>of</strong> Windows<br />
does not correspond to short run pr<strong>of</strong>it maximization by a monopolist.<br />
Schmalensee’s direct testimony argues that Micros<strong>of</strong>t’s low prices are due at<br />
least in part to its concern that higher prices would encourage other firms<br />
to develop competing operating systems” even if, they add, “neither side has<br />
proposed a formal model where such ‘limit pricing’ would make sense.”<br />
To verify in the simplest way that the price <strong>of</strong> Windows does not correspond<br />
to the monopolistic price for the OS market, assume for simplicity<br />
that the marginal cost <strong>of</strong> producing Windows is zero, <strong>and</strong> that the price <strong>of</strong><br />
hardware is constant <strong>and</strong> independent from the price <strong>of</strong> Windows. Dem<strong>and</strong><br />
for Windows is clearly a derived dem<strong>and</strong>, in the sense that it depends on<br />
the dem<strong>and</strong> for PCs <strong>and</strong> on the total price <strong>of</strong> PCs in particular. St<strong>and</strong>ard<br />
economic theory implies that the monopolistic price for an operating system<br />
should be the price <strong>of</strong> the hardware divided by − 1, where is the elasticity<br />
<strong>of</strong> dem<strong>and</strong> for PCs (including both hardware <strong>and</strong> s<strong>of</strong>tware): it means that a<br />
1% increase in the price <strong>of</strong> PCs reduces dem<strong>and</strong> by %. 25 Now, this rela-<br />
24 In 2000, it has been estimated that 68 % <strong>of</strong> s<strong>of</strong>tware companies developed applications<br />
for Windows, 19 % for Apple (which requires adapting to both unique<br />
s<strong>of</strong>tware <strong>and</strong> hardware), 48 % for various versions <strong>of</strong> Unix <strong>and</strong> Linux <strong>and</strong> 36 %<br />
for other proprietary OSs (see Lerner, 2001). Notice that the respective percentages<br />
in 1992 were 71%, 12%, 33%<strong>and</strong> 31%, therefore competing OSs experienced<br />
an increase in s<strong>of</strong>tware developers compared to Micros<strong>of</strong>t during the 90s.<br />
25 Formally, let us assume that the price <strong>of</strong> the hardware is fixed <strong>and</strong> independent<br />
from that <strong>of</strong> the s<strong>of</strong>tware. Given a dem<strong>and</strong> D(h + w) decreasing in the price <strong>of</strong><br />
the hardware h plus the price <strong>of</strong> Windows w, the gross pr<strong>of</strong>it <strong>of</strong>amonopolist<br />
in the OS market would be wD(h + w) <strong>and</strong> would be maximized by a price <strong>of</strong>
226 6. Micros<strong>of</strong>t Economics<br />
tionship tells us that, if the basic price <strong>of</strong> the hardware is € 1000, which is<br />
about the current average price for a PC, the monopolistic price for Windows<br />
would be other € 1000 if =2,itwouldbe€ 500 if =3,itwouldbe€<br />
333 if =4<strong>and</strong> so on. Foncel <strong>and</strong> Ivaldi (2005) estimate this elasticity on<br />
the basis <strong>of</strong> a panel data <strong>of</strong> all PC br<strong>and</strong>s sold in the G7 countries over the<br />
period 1995-1999 <strong>and</strong> derive a value between =1.5 <strong>and</strong> =3with a best<br />
guess slightly above two. The moral is that it would take really unreasonable<br />
values <strong>of</strong> dem<strong>and</strong> elasticity to even get close to the real price <strong>of</strong> Windows,<br />
which is around € 50. 26 Moreover, the above estimate <strong>of</strong> the monopolistic<br />
price is very conservative. In the real world, we can imagine that the price<br />
<strong>of</strong> hardware is not completely independent from the price <strong>of</strong> Windows: if<br />
the latter would double tomorrow, hardware producers would be forced to<br />
somewhat reduce their prices (eventually switching to lower cost techniques<br />
<strong>and</strong>/or lower quality products). 27 Even if this effect may be limited by the<br />
high level <strong>of</strong> competition in the hardware sector, it works in the direction <strong>of</strong><br />
further increasing the hypothetical monopolistic price, that is, even beyond<br />
the actual price <strong>of</strong> Windows. Finally, let us remember what we pointed out<br />
in our previous discussion on the s<strong>of</strong>tware platforms: a two-sided platform<br />
like Windows earns its revenue entirely from end-users, <strong>and</strong> not from s<strong>of</strong>tware<br />
developers, which are typically subsidized by Micros<strong>of</strong>t to develop new<br />
<strong>and</strong> better applications to strengthen network externalities. Hence, the low<br />
Windows w ∗ such that D(h + w ∗ )+w ∗ D 0 (h + w ∗ )=0or:<br />
w ∗ =<br />
h<br />
− 1<br />
with elasticity <strong>of</strong> dem<strong>and</strong>.<br />
26 This point was first made by Richard Schmalensee, a testimony in the Micros<strong>of</strong>t<br />
vs. US case on behalf <strong>of</strong> the Micros<strong>of</strong>t Corporation (see Schmalensee, 2000).<br />
It was criticized by the US Government’s economic witnesses at the trial (see<br />
Fisher <strong>and</strong> Rubinfeld, 2001), who argued that Micros<strong>of</strong>t did not maximize short<br />
run pr<strong>of</strong>its, but did actually maximize long run pr<strong>of</strong>its taking into account the<br />
positive impact <strong>of</strong> a lower price on network effects. This last point may have<br />
been correct for the initial pricing strategies <strong>of</strong> Micros<strong>of</strong>t, but it does not explain<br />
why the price <strong>of</strong> Windows has remained so low after two decades.<br />
27 Formally, think that the price <strong>of</strong> hardware h(w) is decreasing in that <strong>of</strong> Windows<br />
(we could endogenize the actual effect but this is beyond the scope <strong>of</strong> this<br />
discussion). Then, we can rework the monopolistic price <strong>of</strong> Windows as:<br />
w ∗ =<br />
h(w ∗ )<br />
[1 + h 0 (w ∗ )] − 1<br />
which is higher that in absence <strong>of</strong> this translation effect (remember that h 0 (w ∗ ) <<br />
0): a monopolist would price Windows even more because part <strong>of</strong> the potential<br />
reduction in dem<strong>and</strong> due to a higher price would be avoided thanks to an induced<br />
reduction in the price <strong>of</strong> the hardware.
6.3 Is Micros<strong>of</strong>t a Monopolist? 227<br />
price <strong>of</strong> Windows appears further away from what should be the hypothetical<br />
monopolistic price.<br />
Hall <strong>and</strong> Hall (2000) developed similar calculations to the one above assuming<br />
Nash competition in quantities in the hardware market <strong>and</strong> suggested<br />
that Micros<strong>of</strong>t has to adopt a low price for Windows as a rational strategy in<br />
front <strong>of</strong> endogenous entry in the PC market. Their conclusion is consistent<br />
with the results <strong>of</strong> the theory <strong>of</strong> market leaders <strong>and</strong> endogenous entry: “not<br />
only is the price <strong>of</strong> Windows brought down to a small fraction <strong>of</strong> its monopoly<br />
price, but the social waste <strong>of</strong> duplicative investment in operating systems is<br />
avoided as well.”<br />
It has been claimed that low Windows pricing may be explained with<br />
higher pricing <strong>of</strong> the complementary applications, as the Micros<strong>of</strong>t Office<br />
suite. However, the combined price <strong>of</strong> Windows <strong>and</strong> the average application<br />
package sold with it is still below the monopolistic price. Moreover, these applications<br />
are not sold at lower prices for other OSs. As Nicholas Economides<br />
pointed out:<br />
“Windows has the ability to collect surplus from the whole assortment<br />
<strong>of</strong> applications that run on top <strong>of</strong> it. Keeping Windows’<br />
price artificially low would subsidize not only MS-Office, but also<br />
the whole array <strong>of</strong> tens <strong>of</strong> thous<strong>and</strong>s <strong>of</strong> Windows applications that<br />
are not produced by Micros<strong>of</strong>t. Therefore, even if Micros<strong>of</strong>t had a<br />
monopoly power in the Office market, keeping the price <strong>of</strong> Windows<br />
low is definitely not the optimal way to collect surplus” (Economides,<br />
2001).<br />
What does all this tell us? Simply that Micros<strong>of</strong>t is not an unconstrained<br />
price-setter, while its prices are limited well below the monopolistic price<br />
to compete aggressively with the other firms active in the operating system<br />
market <strong>and</strong> with the potential entrants in it. Economides (2001) concludes in<br />
a similar fashion: “Micros<strong>of</strong>t priced low because <strong>of</strong> the threat <strong>of</strong> competition.<br />
This means that Micros<strong>of</strong>t believed that it could not price higher if it were to<br />
maintain its market position.” The empirical work <strong>of</strong> Foncel <strong>and</strong> Ivaldi (2005)<br />
supports the same conjecture: “Micros<strong>of</strong>t seems to behave as if it fears that<br />
charging monopoly prices today would cause it to lose substantial pr<strong>of</strong>its to<br />
competitors in the future.”<br />
Indeed, we can say more than just that Micros<strong>of</strong>t is not a monopoly.<br />
What the post-Chicago approach suggested about leaders in markets with<br />
price competition was that they should be accommodating <strong>and</strong> exploit their<br />
market power, setting higher prices than competitors, or otherwise engage<br />
in predatory pricing <strong>and</strong>, after having conquered the whole market, increase<br />
prices. But in the last 10-15 years <strong>of</strong> global leadership, Micros<strong>of</strong>t has done<br />
neither <strong>of</strong> these things. Micros<strong>of</strong>t has been constantly aggressive, which, according<br />
to the theory <strong>of</strong> market leaders developed in this book, is exactly<br />
what a leader under the threat <strong>of</strong> competitive pressure would do.
228 6. Micros<strong>of</strong>t Economics<br />
The theory <strong>of</strong> market leaders has shown that a market leader in these<br />
conditions would price above marginal cost in such a way to compensate for<br />
the fixed costs <strong>of</strong> investment <strong>and</strong> obtain a pr<strong>of</strong>it margin (over the average<br />
costs <strong>of</strong> production) thanks to the economies <strong>of</strong> scale derived from the large<br />
(worldwide in the case <strong>of</strong> Micros<strong>of</strong>t) scale <strong>of</strong> production. Its (quality adjusted)<br />
price should be below that <strong>of</strong> its immediate competitors, or just low enough<br />
to avoid that they can exploit pr<strong>of</strong>itable opportunities in the market. 28 The<br />
low price <strong>of</strong> Windows induced by competitive pressure <strong>and</strong> network effects<br />
explains its large market share. As Posner (2001, p. 278) has pointed out<br />
acutely, in such a market “a firm may have a monopoly market share only<br />
because it is not charging a monopoly price.”<br />
The significant preference that customers attribute to Windows even in<br />
the presence <strong>of</strong> good alternative products, some <strong>of</strong> which are supplied at no<br />
charge (!), suggests that Micros<strong>of</strong>t is still providing the package with the<br />
best quality-price ratio in the s<strong>of</strong>tware market, at least if we believe in the<br />
rationality <strong>of</strong> consumers. 29<br />
6.3.2 Does Micros<strong>of</strong>t Stifle <strong>Innovation</strong>?<br />
It is also important to look at competition in the s<strong>of</strong>tware market in a dynamic<br />
sense, that is competition for the market, as opposed to the competition<br />
in the market examined until now. As we have emphasized many times,<br />
high-tech sectors can be seen as races to develop new products before others<br />
<strong>and</strong> conquer large market shares with the new products. On the basis <strong>of</strong> the<br />
so-called Arrow effect, we know that incumbent monopolists that do not face<br />
endogenous pressure in the competition for the market have small incentives<br />
to invest in R&D because, by innovating, they only obtain the difference between<br />
the value <strong>of</strong> the next technology <strong>and</strong> that <strong>of</strong> their current technology,<br />
28 IBM initially priced its OS/2 at $ 325 against $ 149 for Windows 3.0. It would<br />
be more interesting to compare the price <strong>of</strong> Windows <strong>and</strong> Mac OS, but the latter<br />
is integrated in Apple computers. However, as Evans et al. (2006) notice, “there<br />
is a clue: the 1990 upgrade to Windows 3.0 was $ 50, about half the price ($<br />
99) <strong>of</strong> a 1991 upgrade to Apple’s System 7.0. Another useful clue comes from a<br />
comparison between computers with similar hardware: in this same period the<br />
average price <strong>of</strong> an Apple PC was over $ 200 more than the average price <strong>of</strong><br />
a similarly equipped <strong>and</strong> powerful Compaq PC sold with Micros<strong>of</strong>t operating<br />
systems.”<br />
29 Nevertheless, it is sometimes claimed, also between economists, that there are<br />
better OSs that provide better services, <strong>and</strong> that a lack <strong>of</strong> information <strong>and</strong> myopic<br />
behavior induce collective hysteresis in the choice <strong>of</strong> Windows. It is hard to<br />
imagine how irrational choices could last so long, <strong>and</strong> it is even more surprising<br />
that economists, always ready to assume rational behavior under the most extreme<br />
circumstances, can believe that consumers suddenly become irrational or<br />
myopic when choosing s<strong>of</strong>tware.
6.3 Is Micros<strong>of</strong>t a Monopolist? 229<br />
while outsiders obtain the full value <strong>of</strong> replacing the incumbents. However,<br />
the theory <strong>of</strong> market leaders developed in Chapter 4 has shown that, when<br />
leaders face competitive pressure, they are induced to invest more in R&D<br />
than any other competitor, with the incentive <strong>of</strong> defending their leadership<br />
from a rapid replacement.<br />
Let us look at the incentives to invest in the s<strong>of</strong>tware market. Of course,<br />
the overall expected value <strong>of</strong> Windows Vista for Micros<strong>of</strong>t can be quite high,<br />
but the net value <strong>of</strong> replacing Windows XP with Vista has been only a<br />
small percentage <strong>of</strong> that value, especially if we take into account that the<br />
real price is not likely to increase <strong>and</strong> that the introduction <strong>of</strong> Vista is only<br />
gradual (<strong>and</strong> associated to the change <strong>of</strong> hardware for most customers). At<br />
the same time, the value <strong>of</strong> developing a successful OS for a competitor<br />
<strong>of</strong> Micros<strong>of</strong>t is incomparably higher. The Arrow effect would suggest that<br />
Micros<strong>of</strong>t has lower incentives to invest in R&D than the other active firms if<br />
further entry in the competition for the OS market is not possible. However,<br />
the theory <strong>of</strong> market leaders replies that this is not the case when entry is<br />
endogenous. Accordingly, this supports the idea that only strong pressure<br />
in the competition for the market could have led Micros<strong>of</strong>t to undertake<br />
an unprecedented investment to rewrite from scratch, develop <strong>and</strong> release a<br />
br<strong>and</strong> new OS as Windows Vista. This pressure, we conjecture, comes mainly<br />
from the new actors in the s<strong>of</strong>tware market, the open source community <strong>and</strong><br />
the commercial companies that are active around this community. At the<br />
same time, it is reasonable to conjecture that the wide <strong>and</strong> fast growing open<br />
source community <strong>and</strong> the commercial companies behind it are investing so<br />
much in R&D exactly because they envision the possibility <strong>of</strong> replacing the<br />
leadership <strong>of</strong> Micros<strong>of</strong>t. In light <strong>of</strong> this, the s<strong>of</strong>tware market appears as a<br />
dynamic sector characterized by strong competition for the market <strong>and</strong> by a<br />
leader that is both a source <strong>and</strong> a cause <strong>of</strong> innovation, quite the opposite <strong>of</strong><br />
how it is sometimes depicted.<br />
Similar ideas appear behind the words <strong>of</strong> the leading scholar <strong>of</strong> the<br />
Chicago school on the s<strong>of</strong>tware industry:<br />
“We have seen all manner <strong>of</strong> firms rise <strong>and</strong> fall in this industry–<br />
falling sometimes from what had seemed a secure monopoly position.<br />
The gale <strong>of</strong> creative destruction that Schumpeter described, in<br />
which a sequence <strong>of</strong> temporary monopolies operates to maximize innovation<br />
that confers social benefits far in excess <strong>of</strong> the social costs<br />
<strong>of</strong> the short-lived monopoly prices that the process also gives rise<br />
to, may be the reality <strong>of</strong> the new economy. This is especially likely<br />
because quality competition tends to dominate price competition in<br />
the s<strong>of</strong>tware market industry. The quality-adjusted price <strong>of</strong> s<strong>of</strong>tware<br />
has fallen steadily simply because quality improvements have vastly<br />
outrun price increases” (Posner, 2001, pp 249-50).<br />
In spring 2007 Micros<strong>of</strong>t unveiled a revolutionary new product called Micros<strong>of</strong>t<br />
Surface, a combination <strong>of</strong> ground breaking s<strong>of</strong>tware <strong>and</strong> hardware
230 6. Micros<strong>of</strong>t Economics<br />
technology that will change the way we interact with digital content. 30 Sooner<br />
or later the technological leadership <strong>of</strong> Micros<strong>of</strong>t in the s<strong>of</strong>tware market will<br />
end, but other companies will have to be more innovative to replace its products.<br />
6.4 Bundling<br />
Virtually any product is a bundle, since it combines multiple basic products<br />
which could be or are sold separately: drugs bundle different molecules,<br />
shoes bundle shoes without laces <strong>and</strong> shoelaces, a car bundles many separate<br />
components, a computer bundles hardware, an operating system <strong>and</strong><br />
basic s<strong>of</strong>tware applications <strong>of</strong> general interest. In some cases, bundling is<br />
just a contractual restriction used to force customers to purchase an ancillary<br />
product in an aftermarket for goods or services, while in other cases<br />
bundling improves a finished product by integrating new components or features<br />
into it: <strong>of</strong> course, only the first situation should be subject to antitrust<br />
investigation. In all <strong>of</strong> its main antitrust cases, Micros<strong>of</strong>t has been accused <strong>of</strong><br />
abusive leveraging through bundling strategies, first between Windows <strong>and</strong><br />
the browser Internet Explorer, <strong>and</strong> then between Windows <strong>and</strong> the Windows<br />
Media Player s<strong>of</strong>tware.<br />
As we noticed in the previous chapter, there are contrasting views on<br />
bundling. The Chicago school has advanced efficiency rationales in its favor<br />
with positive, or at worst ambiguous, consequences on welfare, including<br />
production or distribution cost savings, reduction in transaction costs for<br />
customers, protection <strong>of</strong> intellectual property, product improvements, quality<br />
assurance <strong>and</strong> legitimate price responses. In the case <strong>of</strong> bundling <strong>of</strong> s<strong>of</strong>tware<br />
applications in an OS, in particular, there are efficiencies in internal s<strong>of</strong>tware<br />
design due to componentization <strong>and</strong> code sharing (which facilitate product<br />
development, design <strong>and</strong> testing), there are network externalities made available<br />
encouraging <strong>and</strong> facilitating the development <strong>of</strong> applications that rely on<br />
a bundled functionality <strong>and</strong> there are reductions in customer support costs<br />
which ultimately lead to cost savings for final end-users, whose experience is<br />
also largely simplified by bundling. 31<br />
Moreover, according to the Chicago view only efficiency purposes can<br />
motivate bundling because a firm cannot monopolize another market by<br />
bundling two products: according to the so-called “single monopoly pr<strong>of</strong>it<br />
30 Micros<strong>of</strong>t Surface is a display in a table-like form that is easy for individuals or<br />
small groups to interact with in a way that feels familiar. It can recognize dozens<br />
<strong>of</strong> movements such as multiple touches, gestures <strong>and</strong> actual unique objects that<br />
have identification tags similar to bar codes, it eliminates the need for a mouse<br />
<strong>and</strong> a keyboard <strong>and</strong> allows multiple users to directly interact through the screen.<br />
31 See Davis et al. (2002).
6.4 Bundling 231<br />
theorem”, as long as the secondary market is competitive, not even a monopolist<br />
in a separate market can increase its pr<strong>of</strong>its in the former by tying the<br />
two products. Actually, in the presence <strong>of</strong> complementarities, it can only gain<br />
from having competition <strong>and</strong> high sales in the secondary market to enhance<br />
dem<strong>and</strong> in its monopolistic market. A similar idea has been advanced at a<br />
theoretical level by Davis <strong>and</strong> Murphy (2000) to explain the tying strategies<br />
<strong>of</strong> Micros<strong>of</strong>t: they also emphasize a well known basic principle, for which<br />
a monopolist will choose lower prices for two complement goods than the<br />
prices chosen by two separate monopolists, which suggests that the bundling<br />
strategies <strong>of</strong> Micros<strong>of</strong>t reduce permanently the prices <strong>of</strong> both Windows <strong>and</strong><br />
the bundled products. Motta (2004) adds that a positive impact on prices<br />
emerges also for independent products when there are network effects, because<br />
in the secondary market “it might be very difficult to leapfrog the<br />
current leader, <strong>and</strong> a firm that can rely on important R&D, marketing <strong>and</strong><br />
financial assets might manage to achieve what a small firm might not have.”<br />
With particular reference to the US case, Economides (2001) notes that<br />
Micros<strong>of</strong>t could not have been interested in the browser market when this<br />
was perfectly competitive, but only when this market became dominated by<br />
Netscape for two main reasons. “First, Netscape had a dominant position in<br />
the browser market, thereby taking away from Micros<strong>of</strong>t’s operating system<br />
pr<strong>of</strong>its to the extent that Windows was used together with the Navigator.<br />
Second, as the markets for Internet applications <strong>and</strong> electronic commerce<br />
exploded, the potential loss to Micros<strong>of</strong>t from not having a top browser increased<br />
significantly... Clearly, Micros<strong>of</strong>t had a pro-competitive incentive to<br />
freely distribute IE since that would stimulate dem<strong>and</strong> for the Windows platform.”<br />
The very same point could be made for the more recent bundling <strong>of</strong> Media<br />
Player with Windows <strong>of</strong>fers a very low price. 32 Actually, this motivation for<br />
bundling Windows <strong>and</strong> Media Player (aimed at increasing the attractiveness<br />
<strong>of</strong> the former <strong>and</strong> promoting applications for the latter) appears the main<br />
direct driver in the European case. Today there are no serious threats on<br />
Windows that could come from an alternative media player s<strong>of</strong>tware (while<br />
browsers could represent a potential threat for Windows in the mid 90s).<br />
However, there is a different class <strong>of</strong> motivations for bundling that we need<br />
to examine: these are the indirect or strategic motivations.<br />
32 Notice that, since at the time <strong>of</strong> launch we were in front <strong>of</strong> cases <strong>of</strong> pure (<strong>and</strong><br />
not mixed) bundling, we do not know the implicit prices <strong>of</strong> IE <strong>and</strong> MediaPlayer<br />
in the bundled versions <strong>of</strong> Windows. However, they must have been quite low or<br />
close to zero because there were not apparent increases in the price <strong>of</strong> the new<br />
versions <strong>of</strong> Windows. Moreover, the recent unbundled version (without Media<br />
Player) is sold at the same price as the bundled version.
232 6. Micros<strong>of</strong>t Economics<br />
6.4.1 Strategic Bundling<br />
The post-Chicago approach has shown that, when the bundling firm has some<br />
market power in the primary market, commitment to bundling can only be<br />
used for exclusionary purposes since it enhances competition in the secondary<br />
market <strong>and</strong> increases the pr<strong>of</strong>its <strong>of</strong> the leader only if it excludes rivals from<br />
this secondary market (Whinston, 1990). Nevertheless, even the same proponent<br />
<strong>of</strong> this theory has expressed doubts on its applicability to the case <strong>of</strong><br />
Micros<strong>of</strong>t: evaluating the tying <strong>of</strong> Windows <strong>and</strong> IE, Whinston (2001) notes<br />
that “Micros<strong>of</strong>t seems to have introduced relatively little incompatibility with<br />
other browsers. Since marginal cost is essentially zero, bundling could exclude<br />
Netscape only if consumers, or computer manufacturers for them, faced other<br />
constraints on adding Navigator to their system”, which did not appear to be<br />
the case. The same holds in the case <strong>of</strong> Windows Media Player. It is true that<br />
Micros<strong>of</strong>t bundling in both markets reduced the average prices <strong>of</strong> browsers 33<br />
<strong>and</strong> media players, but this did not lead to the deterrence <strong>of</strong> entry (new<br />
successful browsers such as Firefox have appeared).<br />
As we have formally shown (in Section 2.10), the theory <strong>of</strong> market leaders<br />
emphasizes that when entry in the secondary market is endogenous, an<br />
incumbent can gain from bundling exactly because this creates a sort <strong>of</strong> commitment<br />
to apply a low price to the bundle as a whole, which may end up<br />
increasing the overall pr<strong>of</strong>its for the leader (compared to those obtained only<br />
in the primary market without bundling). Nevertheless, the price <strong>of</strong> the two<br />
goods together would be reduced <strong>and</strong> entry <strong>of</strong> alternative secondary products<br />
would be still viable. This kind <strong>of</strong> rationale for bundling is more likely<br />
to emerge when there are some complementarities between the products,<br />
or there are unexploited network effects (which can be enhanced through<br />
bundling); ultimately, even sales <strong>and</strong> pr<strong>of</strong>its in the primary market may increase.<br />
What matters for our purposes is that bundling is not an extreme<br />
strategy adopted by an incumbent firm to deter entry, but a st<strong>and</strong>ard aggressive<br />
strategy that, by reducing the final prices, may indeed reduce entry<br />
<strong>of</strong> followers, without excluding entry overall. As a matter <strong>of</strong> fact, under some<br />
level <strong>of</strong> product differentiation, the impact on the competitors is quite limited<br />
<strong>and</strong> only marginal firms <strong>of</strong> the secondary market would be driven out <strong>of</strong><br />
it. Hence, in a world <strong>of</strong> price competition, it appears hard to conclude that<br />
bundling could be used as a predatory strategy when it does not even lead<br />
to the exit <strong>of</strong> all the competitors, but just to a permanent reduction <strong>of</strong> the<br />
price level.<br />
33 Netscape charged many private <strong>and</strong> corporate users for its browser until started<br />
facingsubstantialcompetition.Pricesrangedbetween$39in1995to$79in<br />
1996 for a premium version. This was quite a high price if we think that the<br />
average price <strong>of</strong> the entire OS Windows was in this same range during those<br />
years.
6.4 Bundling 233<br />
To sum up our general point, when approaching a bundling case we suggest<br />
verifying the entry conditions <strong>of</strong> the secondary market. If there is a<br />
dominant firm in this market as well, the main problem is not the bundling<br />
strategy, but the lack <strong>of</strong> competition in the secondary market, <strong>and</strong> it should<br />
be addressed within that same market: punishing the bundling strategy would<br />
just guarantee the monopolistic (or duopolistic) rents <strong>of</strong> the dominant firm<br />
in the secondary market. However, things are different when the secondary<br />
market is not monopolized but open to endogenous entry (even if it is not perfectly<br />
competitive, in the sense that firms do not price at marginal cost). In<br />
such a case, <strong>and</strong> especially in the presence <strong>of</strong> product differentiation, bundling<br />
is a pro-competitive strategy <strong>and</strong> punishing it would hurt consumers.<br />
In the case <strong>of</strong> Micros<strong>of</strong>t, in both bundling situations, that <strong>of</strong> Windows<br />
with Internet Explorer <strong>and</strong> that <strong>of</strong> Windows with Media Player, the tied<br />
market was characterized by endogenous entry. Paradoxically, this became<br />
even more evident since Micros<strong>of</strong>t entered in these markets: just think <strong>of</strong> new<br />
successful browsers as Mozilla Firefox, Netscape, Safari, Opera, Konqueror<br />
<strong>and</strong> media player s<strong>of</strong>tware as RealPlayer, Apple’s QuickTime, Adobe’s Flash,<br />
MusicMatch <strong>and</strong> many others. 34 Consequently the bundling strategy <strong>of</strong> Micros<strong>of</strong>t<br />
could be simply seen as an aggressive <strong>and</strong> competitive strategy <strong>of</strong> a<br />
market leader active in a secondary market where entry is indeed endogenous.<br />
Moreover, in these markets the st<strong>and</strong>ard strategy is to provide free s<strong>of</strong>tware<br />
to enhance network effects <strong>and</strong> earn from externalities associated with the<br />
use <strong>of</strong> the s<strong>of</strong>tware (a typical strategy in multisided markets, as we have<br />
seen). For instance, in the case <strong>of</strong> digital media platforms, Micros<strong>of</strong>t looks<br />
for network effects on licenses <strong>of</strong> its OS, Real earns from content subscriptions,<br />
Apple from sales <strong>of</strong> digital audio <strong>and</strong> video devices (the iPod <strong>and</strong>, in<br />
perspective, the iPhone) 35 <strong>and</strong> Adobe from Flash server sales. Even if these<br />
companies adopt very different business models, competition is quite intense<br />
especially because multi-homing is common practice: end users typically use<br />
multiple media players (that are characterized by a certain degree <strong>of</strong> hori-<br />
34 Entry in the market for s<strong>of</strong>tware applications is definitely easier today, since most<br />
s<strong>of</strong>tware is installed through on line downloading in a few seconds. Since most<br />
PCs are endowed with a browser (which, ironically, is largely due to Micros<strong>of</strong>t<br />
bundling IE with Windows), the market for s<strong>of</strong>tware applications has become<br />
one <strong>of</strong> the most transparent <strong>and</strong> competitive.<br />
35 This implies different levels <strong>of</strong> platform integration <strong>and</strong> interoperability with<br />
different platforms. As Evans et al. (2006) noticed: “At one extreme is Apple. Its<br />
iPod/iTunes platform is integrated into the hardware <strong>and</strong> content-provider sides<br />
<strong>of</strong> the media platform, <strong>and</strong> its doesn’t interoperate with any other platform. At<br />
the other extreme is Micros<strong>of</strong>t, whose media platform is integrated into neither<br />
hardware nor content <strong>and</strong> which interoperates with all other media platforms<br />
that allow it to do so. In the middle are vendors like RealNetworks, which limit<br />
interoperability - but not completely - <strong>and</strong> integrate - but only partially - into<br />
the content provider side.”
234 6. Micros<strong>of</strong>t Economics<br />
zontal differentiation), <strong>and</strong> also PC manufacturers typically install multiple<br />
competing mediaplayers at their will (while this is not the case for digital<br />
music devices <strong>and</strong> mobile phones). Finally, multi-homing is a clear symptom<br />
that media players are horizontally differentiated (some are better for music<br />
content, others for videos, others for storing files, <strong>and</strong> so on): from our previous<br />
discussion on bundling for secondary markets with endogenous entry<br />
<strong>and</strong> product differentiation, it follows that these are precisely the conditions<br />
under which bundling assumes a competitive nature rather than a predatory<br />
one.<br />
6.4.2 Technological Bundling<br />
Beyond the debate on the nature <strong>of</strong> the strategic reasons for which firms may<br />
engage in bundling strategies, there are technological reasons why bundling<br />
mayemerge.Thesehavetypicallybeenatthebase<strong>of</strong>Micros<strong>of</strong>tdefenseinits<br />
antitrust cases. In dynamic markets like the s<strong>of</strong>tware market, the same concept<br />
<strong>of</strong> a good is changing over time, since both dem<strong>and</strong> <strong>and</strong> supply change.<br />
If dem<strong>and</strong> by PC users for media player functionality was limited just a few<br />
years ago, now it appears that these functionalities are an essential component<br />
<strong>of</strong> an OS. Because <strong>of</strong> this, an increasing number <strong>of</strong> s<strong>of</strong>tware applications<br />
<strong>and</strong> on line services are associated with media player functionalities, so that<br />
dem<strong>and</strong> is strengthened by network effects. If supply <strong>of</strong> media player functionalities<br />
was inefficient through bundling a few years ago, <strong>and</strong> it was mostly<br />
left to specific add ons, improvements in hardware processing power, in the<br />
cost <strong>of</strong> hard disk storage <strong>and</strong> r<strong>and</strong>om access memory, <strong>and</strong> in the streaming<br />
technology made it simple <strong>and</strong> efficient to bundle media player functionality<br />
within current OSs. As a consequence <strong>of</strong> this, bundling has a natural technological<br />
rationale <strong>and</strong> should emerge endogenously when the size <strong>of</strong> dem<strong>and</strong><br />
is large enough <strong>and</strong> the cost <strong>of</strong> supply is low enough. In other words, while<br />
a few years ago an OS <strong>and</strong> a media player could be regarded as separate<br />
goods whose union could be associated with a bundling strategy, nowadays<br />
an OS must incorporate media player functionalities (as it must incorporate<br />
a browser) so that we cannot even talk <strong>of</strong> a traditional form <strong>of</strong> bundling. 36<br />
36 This is common for s<strong>of</strong>tware. For instance, word processors <strong>and</strong> spell checkers<br />
were in separate markets many years ago, not today. Voice recognition is separate<br />
today, but we can expect that it will be integrated in word processors at some<br />
point. Navigators for cars are still optional tools in the automobile industry,<br />
but are likely to become essential components in the near future <strong>and</strong> bundle<br />
new applications <strong>and</strong> new content. At the same time, videogame consoles <strong>and</strong><br />
PCs are likely to enter in the television ecosystem <strong>and</strong> bundle new features <strong>and</strong><br />
capabilities in the attempt to gain the socalled “control <strong>of</strong> the living room”. In<br />
a sense, bundling creates new industries <strong>and</strong> is a source <strong>of</strong> competition for the<br />
market in itself.
6.5 Intellectual Property Rights 235<br />
In this perspective, attempts <strong>of</strong> antitrust authorities to stop or delay the<br />
evolution <strong>of</strong> OS through additional features, as browsers <strong>and</strong> media players,<br />
appear quite dangerous: while it is difficult to verify in which moment it would<br />
be optimal to bundle secondary products in an evolving primary product, it<br />
is not clear why antitrust authorities should have a better guess than market<br />
driven firms.<br />
Notice that since the 2004 Commission’s decision, Micros<strong>of</strong>t had to prepare<br />
<strong>and</strong> commercialize a version <strong>of</strong> Windows without Media Player in Europe.<br />
37 Dem<strong>and</strong> for the version <strong>of</strong> Windows without Media Player has been<br />
virtually zero in Europe, a likely sign that Micros<strong>of</strong>t bundling strategy was<br />
at least not hurting consumers.<br />
6.5 Intellectual Property Rights<br />
In the previous chapters we discussed the role <strong>of</strong> market leaders in innovative<br />
markets <strong>and</strong> the importance <strong>of</strong> the protection <strong>of</strong> IPRs in stimulating investment<br />
in R&D <strong>and</strong> technological progress. Both aspects are quite relevant in<br />
the underst<strong>and</strong>ing <strong>of</strong> the dynamics <strong>of</strong> the s<strong>of</strong>tware market <strong>and</strong> the Micros<strong>of</strong>t<br />
case.<br />
The s<strong>of</strong>tware market is a major example <strong>of</strong> an industry where competition<br />
is mainly for the market, <strong>and</strong> in such a case, as we have seen, large<br />
market shares by single firms are a typical outcome. The counterpart <strong>of</strong> this,<br />
<strong>of</strong> course, is that these industries can exhibit catastrophic entry where innovators<br />
can replace current leaders quite quickly. As we noticed in Chapter 4,<br />
in such an environment, it is exactly when competition is open that leaders<br />
have incentives to invest deeply to retain their leadership. On the contrary,<br />
when competition for the market is limited, technological leaders are able to<br />
have a quiet life, invest less in R&D <strong>and</strong> accept the risk that someone will<br />
come up with a better product. When competition for the market is open,<br />
this same risk is too high <strong>and</strong> incumbents prefer to accept the challenge <strong>and</strong><br />
try to innovate first: this leads to a more persistent leadership.<br />
When entry is endogenous, innovation by leaders creates a virtuous circle<br />
that also has important implications for the way we can evaluate such a<br />
market (see Section 4.3). The endogenous persistence <strong>of</strong> the technological<br />
leadership has a value that creates incentives for all firms to invest even<br />
more, which in turn strengthens the same incentives <strong>of</strong> the leader to invest<br />
<strong>and</strong> retain its leadership, <strong>and</strong> so on. In other words, persistence <strong>of</strong> leadership<br />
is a source <strong>of</strong> strong competition for the market (through investments in R&D<br />
to replace the current leader), <strong>and</strong>, given that leaders have higher incentives<br />
to invest as long as the race to innovate is open, we can also conclude that<br />
37 The version <strong>of</strong> Windows XP without Media Player was called Windows XP N,<br />
a choice <strong>of</strong> the European Commission between nine potential names submitted<br />
by Micros<strong>of</strong>t.
236 6. Micros<strong>of</strong>t Economics<br />
strong competition for the market is a source <strong>of</strong> persistence <strong>of</strong> leadership.<br />
This circular argument may appear paradoxical, but is the fruit <strong>of</strong> a radical<br />
distinction between static <strong>and</strong> dynamic competition: once again, there is no<br />
consistent correlation between market shares <strong>and</strong> market power in dynamic<br />
markets.<br />
The endogenous multiplicative effect <strong>of</strong> the value <strong>of</strong> leadership that we<br />
have just summarized implies that in dynamic markets the rents <strong>of</strong> a leader<br />
may be spectacularly larger than those <strong>of</strong> its competitors, <strong>and</strong> the market<br />
value <strong>of</strong> a leadership may be extremely large even if the market is perfectly<br />
competitive in a dynamic sense (see Segerstrom, 2007, for a related point). In<br />
our view, this is something not too far from what we can see in the s<strong>of</strong>tware<br />
market <strong>and</strong> in the leadership <strong>of</strong> Micros<strong>of</strong>t, but also in many other high-tech<br />
sectors.<br />
6.5.1 Patents, Trade Secrets <strong>and</strong> Interoperability<br />
Thesource<strong>of</strong>thevalue<strong>of</strong>innovation,thestartingpoint<strong>of</strong>thechain<strong>of</strong>value<br />
that we just described, must be a fundamental rent associated with innovations<br />
<strong>and</strong> protected through IPRs. Hence, all forms <strong>of</strong> IPRs are the ultimate<br />
source <strong>of</strong> leadership, innovation <strong>and</strong> technological progress. As we already<br />
noticed, the role <strong>of</strong> patent legislation is exactly to trade <strong>of</strong>f the benefits <strong>of</strong><br />
patents in terms <strong>of</strong> incentives to innovate with the costs related to temporary<br />
monopolistic pricing. In our opinion, there is no reason why antitrust<br />
authorities should interfere with this legislation when patent protection appears<br />
inconsistent with other goals. And even if these goals are legitimate<br />
<strong>and</strong> relevant, introducing a discretionary evaluation <strong>of</strong> IPRs would create<br />
uncertainty <strong>and</strong> jeopardize the investment, which, after all, goes against the<br />
ultimate objective <strong>of</strong> the same antitrust authorities.<br />
Nevertheless, in the Micros<strong>of</strong>t case the EU Commission has taken this<br />
dangerous direction, asking Micros<strong>of</strong>t to disclose a wide amount <strong>of</strong> technologies.<br />
38 More recently (Statement <strong>of</strong> Objections <strong>of</strong> March 1, 2007), the<br />
Commission has asked to make them available royalty free unless they have<br />
38 At the beginning <strong>of</strong> the Appeal on the Micros<strong>of</strong>t case on April 24, 2006, the<br />
author <strong>of</strong> this book expressed a similar concern in an interview for La Libre<br />
Belgique: “Micros<strong>of</strong>t a été forcé de révéler certaines informations pour assurer<br />
l’interopérabilité entre Windows et d’autres systèmes d’exploitation. Mais les dem<strong>and</strong>es<br />
de la Commission Européenne ne sont toujours pas claires; elle ne cesse<br />
de réclamer plus de la part de la firme américaine. Micros<strong>of</strong>t a révélé récemment<br />
le code source de son système Windows: c’est la documentation ultime du<br />
s<strong>of</strong>tware. Que peut-on encore dem<strong>and</strong>er de plus? L’impact de cette décision serat-il<br />
négatif? Le débat économique futur va porter sur les gains qui résultent du<br />
fait que l’on a forcé une entreprise à révéler des secrets technologiques protégés<br />
par copyright... Mais y a-t-il des gains? Devait-on le faire? Sur le long terme,<br />
c’est contre-productif. Qui va investir dans l’innovation et dans la recherche si
6.5 Intellectual Property Rights 237<br />
an innovative nature (meaning that they involve an inventive <strong>and</strong> novel step<br />
compared to the prior art). 39 Finally, it has started questioning the same<br />
innovative nature (<strong>and</strong> with it the license pricing) <strong>of</strong> most technologies that<br />
Micros<strong>of</strong>t was forced to disclose, technologies which are also covered by many<br />
patents approved by US <strong>and</strong> EU patent <strong>of</strong>fices. This creates an even stronger<br />
contradiction between patent law <strong>and</strong> antitrust policy in the EU, <strong>and</strong> also<br />
a substantial divergence between the US approach to IPRs <strong>and</strong> the EU approach,<br />
with the former much more careful in protecting IPRs <strong>and</strong> promoting<br />
R&D.<br />
It is important to add that new ideas, including those underlying Micros<strong>of</strong>t<br />
s<strong>of</strong>tware, are not protected only with patents. Not all inventive <strong>and</strong><br />
innovative activities fall under the scope <strong>of</strong> patentability <strong>and</strong> it is not always<br />
in the interest <strong>of</strong> a firm to patent every single innovation. In most high-tech<br />
sectors, firms adopt a combination <strong>of</strong> patents <strong>and</strong> trade secrets to protect<br />
products that are the result <strong>of</strong> multiple innovations. Defending (intellectual<br />
or material) property rights is one <strong>of</strong> the fundamental conditions for proper<br />
functioning <strong>of</strong> the market economy: defending trade secrets should not play<br />
a minor role in this context.<br />
Some <strong>of</strong> the most famous trade secrets are the formulas <strong>of</strong> Coca-Cola,<br />
Chanel No. 5 <strong>and</strong> Campari. Consider the first example <strong>and</strong> imagine that<br />
Coca-Cola was required to disclose its secret formula: anyone could reproduce<br />
the very same drink, “clone” it under a different name if you like, but it is hard<br />
to believe that this would create large gains for consumers. Close substitutes<br />
to Coke already exist <strong>and</strong> there are small margins to substantially reduce<br />
prices. However, the incentives for any other firminthesameindustryto<br />
invest <strong>and</strong> create new products could be drastically reduced if trade secrets<br />
were not protected. 40<br />
High-tech sectors are more complicated. In these sectors, patents <strong>and</strong><br />
trade secrets <strong>of</strong>ten cover fundamental inventions <strong>and</strong> protecting those inventions<br />
amounts to promoting innovations that today are the main engine <strong>of</strong><br />
growth. In some fields, however, there maybe, at least apparently, a trade-<strong>of</strong>f<br />
between trade secret protection <strong>and</strong> “interoperability” between products -<br />
les droits de propriété intellectuelle sont bafoués? On crée un dangereux précédent<br />
d’autant que l’industrie de haute technologie est souvent caractérisée par<br />
des investissements massifs en recherche et développement” (Contre-productif,<br />
dangereux pour l’innovation et la recherche, by Martin Buxant).<br />
39 Notice that these are features needed for patentability, but not for trade secrets,<br />
which may just protect technologies that are not novel or innovative, but nevertheless<br />
developed with effort <strong>and</strong> costs. Therefore, the most recent approach<br />
<strong>of</strong> the Commission forces disclosure <strong>of</strong> similar technologies, <strong>and</strong> excludes at the<br />
same time that they could be licensed for a positive royalty: this is a way <strong>of</strong><br />
denying de facto the right <strong>of</strong> protection <strong>of</strong> trade secrets. See Kanevid (2007) for<br />
a related analysis <strong>of</strong> compulsory licensing.<br />
40 On the story <strong>of</strong> Coke’s trade secret <strong>and</strong> its implications see Etro (2005,c).
238 6. Micros<strong>of</strong>t Economics<br />
broadly speaking, this is the ability <strong>of</strong> heterogeneous information technology<br />
systems, components <strong>and</strong> services to exchange <strong>and</strong> use information <strong>and</strong> data,<br />
especially in networks. Interoperability is important in the PC industry <strong>and</strong>,<br />
as we have seen in Section 6.1, the level <strong>of</strong> interoperability has strongly increased<br />
in the last decades. Problems arise, however, when interoperability<br />
is confused with “interchangeability” or with a right to clone the innovations<br />
<strong>of</strong> the competitors.<br />
For instance, take in consideration the leading on line search engine in the<br />
world, Google. We may look at Google’s patented innovations, starting with<br />
the 2001 patent on the invention <strong>of</strong> the PageRank by Larry Page (founder<br />
<strong>of</strong> Google with Sergey Brin), 41 but we would need to know its trade secrets<br />
to fully discover the mechanism <strong>of</strong> its precious algorithms. Forcing disclosure<br />
<strong>of</strong> such trade secrets would help many s<strong>of</strong>tware companies <strong>and</strong> websites to<br />
interoperate with Google even better than they already do, as it would allow<br />
other search engines to improve their performances compared to that <strong>of</strong> the<br />
leading search engine. But after that, surely, few companies would invest<br />
huge resources <strong>and</strong> take substantial risks to create a better search engine<br />
or other brilliant ideas like Google when they can just free ride on others’<br />
ideas <strong>and</strong>/or they can’t be sure <strong>of</strong> their return. The same argument would<br />
apply for the trade secrets <strong>of</strong> Micros<strong>of</strong>t or Apple on the source codes <strong>of</strong><br />
their OSs <strong>and</strong> to many other trade secrets <strong>of</strong> innovative leading companies.<br />
Any forced disclosure <strong>of</strong> similar trade secrets represents an expropriation<br />
<strong>of</strong> legitimate investments <strong>and</strong> establishes inappropriate legal st<strong>and</strong>ards with<br />
perverse effects on the incentives to innovate.<br />
6.5.2 Licenses <strong>and</strong> St<strong>and</strong>ards<br />
Fortunately, giving up the precious role <strong>of</strong> IPRs in promoting innovations is<br />
not the only way to solve interoperability challenges. The market can do it<br />
much better: valuable ideas can be selectively commercialized on a voluntary<br />
basis through licenses, for instance under RAND (reasonable <strong>and</strong> non discriminatory)<br />
terms, a type <strong>of</strong> licensing typically used during st<strong>and</strong>ardization<br />
processes to promote the rapid adoption <strong>of</strong> st<strong>and</strong>ards <strong>and</strong> new technologies<br />
41 The abstract <strong>of</strong> US patent 6285999 (filed in 1998) for a method for node ranking<br />
in a linked database, reads as follows: “A method assigns importance ranks to<br />
nodes in a linked database, such as any database <strong>of</strong> documents containing citations,<br />
the world wide web or any other hypermedia database. The rank assigned<br />
to a document is calculated from the ranks <strong>of</strong> documents citing it. In addition,<br />
the rank <strong>of</strong> a document is calculated from a constant representing the probability<br />
that a browser through the database will r<strong>and</strong>omly jump to the document.<br />
The method is particularly useful in enhancing the performance <strong>of</strong> search engine<br />
results for hypermedia databases, such as the world wide web, whose documents<br />
have a large variation in quality.” Of course, by now, this is just the beginning<br />
<strong>of</strong> Google’s ranking mechanism.
6.5 Intellectual Property Rights 239<br />
<strong>and</strong> to encourage entry. The RAND terms include a definition <strong>of</strong> reasonable<br />
royalties, <strong>and</strong> can include further restrictions as field-<strong>of</strong>-use clauses (that allow<br />
licensees to utilize a patented technology in a use that is directly related<br />
to the implementation <strong>of</strong> the st<strong>and</strong>ard), reciprocity clauses, or limits to sublicensing.<br />
42<br />
Coase (1960) has clarified that whenever there is social value to generate,<br />
the market will properly allocate all the property rights. This is also true<br />
for the intellectual property rights: market mechanism can allocate them efficiently,<br />
insure the accessibility <strong>of</strong> the information that fuels interoperability<br />
<strong>and</strong> acknowledge legitimate ownership rights <strong>of</strong> the innovators, so as to enhance<br />
R&D investments. 43 Suppose firm A invests, innovates <strong>and</strong> obtains a<br />
patent, <strong>and</strong> firm B has a new idea to improve firm A’s innovation, but this<br />
idea cannot be used without infringing the patent. Of course, forced interoperability<br />
would lead firm B to implement its idea. However, in such a case<br />
firm A will not invest to start with, <strong>and</strong> no idea will be actually implemented.<br />
Consider now private agreements between the two firms. First, firm A could<br />
license its patent to firm B for a price between the expected pr<strong>of</strong>its that A<br />
<strong>and</strong> B can respectively obtain from marketing alone their respective ideas.<br />
The price depends on the respective bargaining power, the best idea is implemented,<br />
<strong>and</strong> firm A has all the interest to invest ex ante.Second,firm B could<br />
sell its idea to firmAatapriceatmostequaltothedifference in expected<br />
pr<strong>of</strong>its that firm A can obtain respectively with firm B’s idea <strong>and</strong> without<br />
it, with the price again depending on the bargaining power. Also in this case<br />
the incentives for firm A to invest ex ante would be preserved. This suggests<br />
that it may <strong>of</strong>ten be a unique firm to buy others’ innovations (especially if<br />
this firm has developed a comparative advantage in marketing products), <strong>and</strong><br />
it may <strong>of</strong>ten happen that only outsiders take the initiative to invest in new<br />
fields with the aim <strong>of</strong> reselling their innovations. 44 This is indeed the way<br />
technological progress evolves in many industries under protection <strong>of</strong> IPRs.<br />
Finally, in the presence <strong>of</strong> network effects, dynamic market forces can do<br />
even more: as long as IPRs are well protected <strong>and</strong> firms can invest with the<br />
safe confidence that successful innovations will be rewarded, market forces<br />
can select the best st<strong>and</strong>ard when multiple st<strong>and</strong>ards are available <strong>and</strong> interoperability<br />
is only partial. Liebowitz <strong>and</strong> Margolis (1999) have shown that<br />
42 Notice that extreme open source licenses can create frictions with RAND terms<br />
associated with other licenses, so as to jeopardize useful innovative activity -<br />
this is the case <strong>of</strong> the GNU General Public License, which is incompatible with<br />
technologies licensed with any positive royalty, field-<strong>of</strong>-use limitations or other<br />
st<strong>and</strong>ard restrictions.<br />
43 See Scotchmer (2004, Ch. 6) on an interesting discussion on licensing, R&D joint<br />
ventures <strong>and</strong> antitrust policy.<br />
44 This also suggests that in the presence <strong>of</strong> sequential innovations, primary innovations<br />
deserve stronger patent protection than later ones, since their social value<br />
is higher.
240 6. Micros<strong>of</strong>t Economics<br />
this is the case in many episodes. For instance, in the adoption <strong>of</strong> the common<br />
QWERTY keyboard for PCs (so-called from the first five letters on the<br />
top left): for years it has been claimed that the allocation <strong>of</strong> letters <strong>of</strong> this<br />
keyboard was an inefficient st<strong>and</strong>ard, while these researchers found that evidence<br />
suggests that the Qwerty keyboard, somehow selected by the market,<br />
is not worse than any other alternative. 45<br />
In conclusion, also in this field, markets can properly balance the short<br />
run <strong>and</strong> long run interests <strong>of</strong> consumers better than policymakers: promote<br />
innovation, enable an efficient degree <strong>of</strong> interoperability <strong>and</strong> select the best<br />
st<strong>and</strong>ards. It would be better to leave the ruling <strong>of</strong> intellectual property<br />
protection <strong>and</strong> <strong>of</strong> its limits to the legislative level rather than creating an<br />
important precedent for which antitrust authorities could force firms to reveal<br />
their IPRs.<br />
Much <strong>of</strong> the residual contrast between Micros<strong>of</strong>t <strong>and</strong> the European Commission<br />
depends on the approach to interoperability <strong>and</strong> on its ambiguity.<br />
The Commission’s 2004 antitrust decision m<strong>and</strong>ated the licensing <strong>of</strong> intellectual<br />
property to enable full interoperability between Windows PCs <strong>and</strong> work<br />
group servers <strong>and</strong> competitor products. This m<strong>and</strong>ate has turned out to be<br />
the most problematic in the case. The picture that is emerging is <strong>of</strong> a Commission<br />
that has continued to extend the scope <strong>of</strong> the information required,<br />
<strong>and</strong> more recently has also tried to control Micros<strong>of</strong>t pricing (a tool <strong>of</strong> regulatory<br />
authorities, not <strong>of</strong> antitrust ones), while not spelling out exactly what<br />
would constitute compliance with the remedy. Micros<strong>of</strong>t has been forced to<br />
licence more than a hundred technologies, <strong>and</strong> it has even made available to<br />
its competitors selective access to the source code <strong>of</strong> Windows. Nevertheless<br />
in Europe (differently from the US), not one <strong>of</strong> its competitors has taken out<br />
a license, a likely sign that the existing level <strong>of</strong> interoperability is not as low<br />
as it was depicted. 46<br />
6.6 Conclusions<br />
In this chapter we have focused our attention on the New Economy, which<br />
was developed in the last decades around the PC industry <strong>and</strong> the Internet.<br />
The New Economy has spread rapidly all over the world thanks to what<br />
45 Another example is VHS winning out over Betamax for home video recording.<br />
46 One could read the facts in a more negative way. The long effort <strong>of</strong> the EU<br />
Commission to force Micros<strong>of</strong>t to reveal its technologies at better terms may<br />
prevent European firms from licensing any technology at the current terms until<br />
the case will be solved. We believe that an antitrust authority that decides the<br />
name <strong>of</strong> the products <strong>of</strong> a private company (as for Windows XP N), forces public<br />
disclosure<strong>of</strong>itsIPRs,<strong>and</strong>triestodecideitspricesaswell,iswellbeyondthe<br />
limits between antitrust policy <strong>and</strong> regulatory policy. See Mastrantonio (2005)<br />
for an interesting analysis from the law & economics point <strong>of</strong> view.
6.6 Conclusions 241<br />
we are used to call globalization. <strong>Market</strong>s in the New Economy work in a<br />
radically different way from markets in the Old Economy. First <strong>of</strong> all, while<br />
traditional sectors are <strong>of</strong>ten characterized by competition in the market with<br />
substantial product differentiation <strong>and</strong> U-shaped cost functions, many markets<br />
<strong>of</strong> the New Economy are <strong>of</strong>ten driven by competition for the market<br />
taking place through high fixed costs <strong>of</strong> investment in R&D, <strong>and</strong> production<br />
is typically characterized by small <strong>and</strong> constant marginal costs. Beyond this,<br />
many markets <strong>of</strong> the New Economy exhibit network effects <strong>and</strong> are <strong>of</strong>ten<br />
multi-sided, in the sense that firms act as platforms for different types <strong>of</strong><br />
customers with complex network effects between them.<br />
These strong differences require a new approach to the analysis <strong>of</strong> markets<br />
<strong>and</strong> <strong>of</strong> the behavior <strong>of</strong> their leaders. In the absence <strong>of</strong> such a new approach,<br />
it is not surprising that in the last years the attention <strong>of</strong> antitrust authorities<br />
around the world has been <strong>of</strong>ten biased against market leaders in the<br />
sectors <strong>of</strong> the New Economy. These dynamic sectors are certainly not less<br />
competitive than others, but are <strong>of</strong>ten characterized by large market shares<br />
for their leaders <strong>and</strong> aggressive strategies which are the symptom <strong>of</strong> heavy<br />
competition. Leaders might enjoy high market shares yet be subject to massive<br />
competitive pressure to constantly create better products at lower prices<br />
due to threats from innovative competitors <strong>and</strong> potential entrants. Following<br />
our theoretical analysis, in this chapter we tried to argue that the behavior<br />
<strong>of</strong> leaders as Micros<strong>of</strong>t <strong>and</strong> other firms <strong>of</strong> the New Economy can be better<br />
interpreted through the concept <strong>of</strong> Stackelberg competition with endogenous<br />
entry.
7. Epilogue<br />
The objective <strong>of</strong> this book is to develop a theory <strong>of</strong> market leadership <strong>and</strong> endogenous<br />
entry. In the previous chapters we analyzed the choice <strong>of</strong> strategic<br />
investments before competition takes place <strong>and</strong> analyzed first mover advantages<br />
under competition in the market where quantities or prices are the<br />
strategic variables, <strong>and</strong> under competition for the market where investments<br />
in innovations are the strategic variables. We compared the results in the<br />
presence <strong>of</strong> exogenous <strong>and</strong> endogenous entry, <strong>and</strong> we used this theoretical<br />
framework to derive normative implications for antitrust policy.<br />
In this final chapter we will mainly focus our attention on the descriptive<br />
implications <strong>of</strong> the theory <strong>of</strong> market leadership <strong>and</strong> endogenous entry. In Section<br />
7.1 we point out the main empirical predictions <strong>of</strong> our theory that need<br />
future empirical investigations, in Section 7.2 we emphasize a few principles<br />
<strong>of</strong> business administration emerging from our analysis, <strong>and</strong> in Section 7.3 we<br />
suggest directions for future theoretical research. In Section 7.4 we conclude.<br />
7.1 Empirical Predictions <strong>of</strong> the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong><br />
Leaders<br />
The primary empirical implications <strong>of</strong> the theory <strong>of</strong> market leaders concern<br />
the discrimination between alternative strategies adopted by market leaders<br />
facing an exogenous or an endogenous number <strong>of</strong> competitors. 1 Therefore<br />
any empirical investigation <strong>of</strong> our results should be based on a non-trivial<br />
analysis <strong>of</strong> the entry conditions. 2 Some markets are clearly characterized by<br />
1 A good introduction to empirical studies <strong>of</strong> industrial organization can be found<br />
in Martin (2002).<br />
2 Most <strong>of</strong> the empirical work on the reaction <strong>of</strong> incumbents to entry takes entry<br />
as given. The problem <strong>of</strong> endogeneity <strong>of</strong> entry is briefly discussed in Thomas<br />
(1999), who examines the reactions <strong>of</strong> incumbents in the US ready-to-eat cereal<br />
industry. He finds that incumbents are accommodating between themselves, but<br />
they adopt aggressive pricing to face new entrants. This result may be due to the<br />
typical behavior <strong>of</strong> market leaders facing endogenous entry: while price competition<br />
would lead leaders to be accommodating when facing an exogenous number<br />
<strong>of</strong> firms, an aggressive pricing strategy is forced by endogenous entry.
244 7. Epilogue<br />
exogenous constraints on the number <strong>of</strong> firms: for instance, when there are<br />
legal barriers to entry, when only a restricted number <strong>of</strong> firms have licenses,<br />
patents or other essential inputs needed to produce a certain good or service,<br />
or when a certain activity is confined to a predetermined number <strong>of</strong> subjects<br />
with special permission, we are in front <strong>of</strong> a market where the number <strong>of</strong><br />
competitors is exogenous. Some other markets are clearly characterized by<br />
entry open to domestic <strong>and</strong> international firms that changes over time, reacts<br />
rapidly to variations in dem<strong>and</strong> <strong>and</strong> supply conditions, <strong>and</strong> reduces to zero<br />
the supra-normal pr<strong>of</strong>its <strong>of</strong> the marginal entrants: when this is the case, we<br />
are in front <strong>of</strong> a market where the number <strong>of</strong> competitors can be regarded as<br />
endogenous. In other markets the situation is not so clear, therefore we need<br />
to add a few remarks to clarify how one could approach the concept <strong>of</strong> entry<br />
in an empirical investigation aimed at testing the theory <strong>of</strong> market leaders.<br />
First, there are markets in which processes <strong>of</strong> liberalization or deregulation<br />
have radically changed the entry conditions, from a situation with a<br />
fixed number <strong>of</strong> competitors to one with endogenous entry: these shocks may<br />
represent interesting natural experiments for a test <strong>of</strong> our theory. 3 Other<br />
exogenous shocks leading to entry <strong>of</strong> new firms may create interesting natural<br />
experiments. 4 A related situation emerges in markets with IPRs: when a<br />
3 Spiller <strong>and</strong> Favaro (1984) have studied the behavior <strong>of</strong> market leaders in the<br />
process <strong>of</strong> deregulation <strong>of</strong> the commercial banking sector (with data on the<br />
Uruguay experience in the late 70s). The “results are consistent with a von<br />
Stackelberg type <strong>of</strong> industry where the degree <strong>of</strong> oligopolistic interaction among<br />
the leading firms is reduced as a consequence <strong>of</strong> the relaxation <strong>of</strong> the legal entry<br />
barriers.” In recent times, it would be interesting to verify the impact <strong>of</strong> online<br />
banking, which has dramatically increased entry (also <strong>of</strong> international banks)<br />
<strong>and</strong> competition in the banking sector <strong>of</strong> many countries: in such a case, the<br />
theory <strong>of</strong> market leaders would imply the emergence <strong>of</strong> leaders <strong>of</strong>fering better<br />
conditions on savings accounts (think <strong>of</strong> the Orange Savings Account by ING<br />
Direct).<br />
4 Goolsbee <strong>and</strong> Syverson (2006) have examined how incumbents respond to the<br />
threat <strong>of</strong> entry <strong>of</strong> competitors. They use a case study from the American passenger<br />
airline industry, namely the evolution <strong>of</strong> Southwest Airlines’ route network<br />
between 1993 <strong>and</strong> 2004, to identify routes where the probability <strong>of</strong> future entry<br />
rises suddenly for major US carriers as American, Continental, Delta, Northwest,<br />
TWA, United <strong>and</strong> US Airways. Notice that this is a market characterized by a<br />
limited degree <strong>of</strong> product differentiation (mostly driven by frequent flyer miles<br />
programs), by U-shaped cost functions, <strong>and</strong> by competition in prices between airlines<br />
active on each route. When Southwest begins operating in airports on both<br />
sides <strong>of</strong> a route but not the route itself, the probability that it will start flying<br />
that route in the near future increases. Examining the pricing <strong>of</strong> the incumbents<br />
on threatened routes in the period surrounding these events, <strong>and</strong> controlling for<br />
a number <strong>of</strong> airport-specific operating costs, it emerges that incumbents cut fares<br />
significantly when they have faced an exogenous number <strong>of</strong> competitors in the
7.1 Empirical Predictions <strong>of</strong> the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders 245<br />
patent or a copyright expire, endogenous entry suddenly takes place, <strong>and</strong> the<br />
effect on the behavior <strong>of</strong> the incumbents could be used to test our results. 5 Another<br />
interesting situation that could be used for empirical purposes emerges<br />
in markets that, after a period <strong>of</strong> protection from international competition,<br />
are opened to entry <strong>of</strong> foreign firms: this represents another experiment in<br />
past, but expect endogenous entry in the future. More exactly, 3 to 4 quarters<br />
before Southwest starts its operations on a new route, the fares <strong>of</strong> the market<br />
leader on that route have fallen about 7 %, <strong>and</strong> by 1 to 2 quarters prior, they<br />
have fallen 10 %, while when Southwest actually starts operating, prices are almost<br />
12 % lower, <strong>and</strong> after entry the total drop in fares is about 26%. However,<br />
price cuts (in the run up to Southwest starting operations) are absent in lowconcentration<br />
routes, that is in the routes where, most likely, entry was already<br />
free.<br />
Furthermore, the empirical analysis <strong>of</strong> Goolsbee <strong>and</strong> Syverson (2006) reveals<br />
a switch toward the aggressive behavior <strong>of</strong> market leaders facing endogenous<br />
entry <strong>and</strong> without exclusionary purposes. They test whether there are differences<br />
between the reactions <strong>of</strong> incumbents when pre-emptive deterrence is possible<br />
(Southwest’s entry is likely after starting operations on both sides <strong>of</strong> a route, but<br />
could be avoided through price cuts) <strong>and</strong> when it is not (Southwest’s entry in the<br />
route is announced simultaneously with its start <strong>of</strong> operations in the airport).<br />
The pricing strategies <strong>of</strong> the incumbent are quite similar in the two samples,<br />
<strong>and</strong> the conclusion is that “even on routes where deterrence is impossible, the<br />
incumbents engage in the same pre-emptive price cutting behavior. Thus the<br />
behavior cannot be motivated as seeking to deter entry.” Following the traditional<br />
theory <strong>of</strong> price leadership which generates an accommodating behavior <strong>of</strong> the<br />
leaders, Goolsbee <strong>and</strong> Syverson (2006) are forced to conclude that “the firms are<br />
instead accommodating entry”, which can be quite misleading since these leaders<br />
are radically reducing their prices rather than increasing them. The paradox<br />
disappearsoncewerealizethatweareinfront <strong>of</strong> price leaders facing endogenous<br />
entry, <strong>and</strong> that our theory tells us that these leaders should be aggressive <strong>and</strong><br />
also reduce their prices when they are not trying to deter entry.<br />
5 A similar experiment, which could be re-interpreted in the terms <strong>of</strong> our theory,<br />
is in Ellison <strong>and</strong> Ellison (2007). They examine the behavior <strong>of</strong> market leaders<br />
in the pharmaceutical industry in the periods around the expiration <strong>of</strong> patent<br />
protection for their patented drugs. Advertising by incumbents declines before<br />
entry occurs. Drug prices always decline when entry occurs, <strong>and</strong> also before the<br />
expiration <strong>of</strong> the patent, but only if the probability <strong>of</strong> entry is high. Again, these<br />
preliminary results are consistent with an aggressive strategy by the leaders,<br />
which is induced by endogenous entry. Bergman <strong>and</strong> Rudholm (2003) examine<br />
the Swedish pharmaceutical market where the commitment to a low price is<br />
enforced by a particular regulation (for which, if a price is reduced, it is impossible<br />
to increase it again). They show that the prices <strong>of</strong> the incumbent leaders fall at<br />
thetime<strong>of</strong>thepatentexpiration(evenbeforeactualentryoccurs)by5-8%for<br />
products with small sales volumes.
246 7. Epilogue<br />
which endogenous entry suddenly takes place. 6 In all <strong>of</strong> these examples, one<br />
can compare the behavior <strong>of</strong> market leaders relative to the behavior <strong>of</strong> the<br />
followers before <strong>and</strong> after endogenous entry takes place. Ideally, any empirical<br />
methodology should control for the differences between the leader <strong>and</strong> all <strong>of</strong><br />
the other firms (our basic testable predictions refer to the behavior <strong>of</strong> leaders<br />
facing competition from equally efficient firms). 7<br />
Second, there are intermediate situations in which entry can be regarded<br />
as exogenous in the short run, but endogenous only in the medium-long run<br />
simply because entry takes time. This time can be different in different sectors:<br />
rather than being a limit to the testability <strong>of</strong> the theory <strong>of</strong> market<br />
leaders, this variability in the degree <strong>of</strong> reactivity <strong>of</strong> entry to pr<strong>of</strong>it opportunities<br />
could be exploited as a useful control variable, especially if one has<br />
good instruments available to identify the entry conditions. 8<br />
Third, one has to take into consideration entry in the competition in the<br />
market but also entry in the competition for the market: the former is visible<br />
<strong>and</strong> active in the same market, while the latter is <strong>of</strong>ten not visible because<br />
firms may be effectively competing for a market <strong>and</strong> investing in R&D, but<br />
they will not enter in the market until they actually develop a successful<br />
product.<br />
Fourth, one has to distinguish between effective entry <strong>and</strong> potential entry:<br />
while the former is visible <strong>and</strong> the latter is not, the existence <strong>of</strong> potential entry<br />
is the essential element <strong>of</strong> a market in which entry is endogenous compared<br />
to a market in which the number <strong>of</strong> competitors is exogenous.<br />
Another important preliminary issue concerns the form <strong>of</strong> competition<br />
in the market. It is well known that the difference between competition in<br />
prices <strong>and</strong> in quantities is more a theoretical abstraction than a clear-cut<br />
element <strong>of</strong> differentiation between sectors. However, there are some markets<br />
in which price choices are an essential component <strong>of</strong> competition, <strong>and</strong> others<br />
where production decisions determine, to a large extent, the equilibrium<br />
6 In a similar vein, Scherer <strong>and</strong> Keun (1992) looked at the increase in high-tech<br />
imports in US <strong>and</strong> found that incumbents in sectors without barriers to entry<br />
react more aggressively to endogenous entry, increasing R&D/sales more than<br />
other firms.<br />
7 Thomas (1999) has studied the behavior <strong>of</strong> incumbents in the ready-to-eat cereal<br />
industry, which is characterized by competition in prices, product differentiation<br />
<strong>and</strong> large advertising. The main result is that “incumbent firms accommodate<br />
one another on price but respond aggressively using advertising. Entrants on the<br />
other h<strong>and</strong> are more likely to be met with an aggressive price response.” The<br />
difference in the behavior <strong>of</strong> market leaders may indicate a switch in strategy<br />
from a situation with an exogenous number <strong>of</strong> competitors (the incumbents) <strong>and</strong><br />
a situation where entry (<strong>of</strong> new firms) is endogenous.<br />
8 In this perspective, a good empirical strategy to measure the entry conditions,<br />
would involve measuring the likelihood <strong>of</strong> entry in a market <strong>and</strong> estimating its<br />
determinants.
7.1 Empirical Predictions <strong>of</strong> the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders 247<br />
price: markets for highly differentiated goods typically belong to the first<br />
group, while markets for homogenous goods belong more <strong>of</strong>ten to the second<br />
group. These broad differences should be kept in mind when comparing results<br />
from different markets. This is particularly important because, as we<br />
have seen repeatedly, entry conditions can fundamentally change the behavior<br />
<strong>of</strong> market leaders under competition in prices. Furthermore, when firms compete<br />
in multiple strategies, it is important to underst<strong>and</strong> which preliminary<br />
investments or commitments can substantially affect competition: the different<br />
behavior <strong>of</strong> market leaders in undertaking strategic investments compared<br />
to other firms is a crucial element <strong>of</strong> the theory <strong>of</strong> market leaders. 9<br />
Finally, our predictions refer to the behavior <strong>of</strong> market leaders versus<br />
the behavior <strong>of</strong> their followers, <strong>and</strong> the definition <strong>of</strong> leaders <strong>and</strong> followers<br />
requires some additional specifications. In this case, market shares can be<br />
useful because it is normal to associate first mover advantages to the leading<br />
firm in terms <strong>of</strong> market share. One may consider more than one firm as a<br />
leader according to the sector under consideration: our analysis has shown<br />
that multiple leaders would tend to replicate the behavior <strong>of</strong> a single leader.<br />
Of course, there can be differences between firms that are beyond the strategic<br />
advantages: for instance costs differences, differences in product quality<br />
or locational differences. Since our results refer to symmetric firms from a<br />
technological point <strong>of</strong> view, these exogenous differences should be used as<br />
control variables in the analysis.<br />
Given these short but important methodological premises, in what follows<br />
we will list some <strong>of</strong> the empirical predictions <strong>of</strong> the theory <strong>of</strong> market leaders<br />
that distinguish between markets with an exogenous number <strong>of</strong> firms <strong>and</strong><br />
markets with endogenous entry.<br />
In Chapters 1 <strong>and</strong> 3 we have seen that, with competition in the market,<br />
endogenous entry turns market leaders into more aggressive players comparedtoasituationinwhichallfirms<br />
(leaders <strong>and</strong> followers) do not face<br />
entry threats. In particular, our analysis allows one to discriminate a radical<br />
change <strong>of</strong> strategy under competition in prices: when the number <strong>of</strong> firms<br />
is exogenous, market leaders should choose higher prices than the followers,<br />
9 Röller <strong>and</strong> Sickles (2000) have performed the first empirical study <strong>of</strong> a twostage<br />
competition with preliminary investment in cost reducing capacity. They<br />
considered the European airline industry in the period 1976-1990, before the<br />
recent liberalization efforts. On the basis <strong>of</strong> a panel <strong>of</strong> the largest carriers (Air<br />
France, Alitalia, British Airways, Iberia, KLM, Lufthansa, SABENA <strong>and</strong> SAS)<br />
<strong>and</strong> a large dataset on cost, network <strong>and</strong> dem<strong>and</strong> data, they have shown that<br />
airline companies behaved as puppy dogs: underinvesting in capacity to keep<br />
high prices. It would be interesting to compare that situation with the current<br />
situation in which EU liberalization is promoting the entry <strong>and</strong> competition:<br />
according to our the theory <strong>of</strong> market leaders, we would expect leading carriers<br />
to turn into top dogs <strong>and</strong> overinvest in capacity to reduce their relative marginal<br />
costs.
248 7. Epilogue<br />
when entry is endogenous they should choose lower prices. This strong implication<br />
does not necessarily hold when firms compete in quantities, but in<br />
all cases we would expect that the price <strong>of</strong> the leaders decreases compared to<br />
the price <strong>of</strong> the followers when endogenous entry occurs. Therefore, our first<br />
testable implication is a weak one <strong>and</strong> can be expressed as follows:<br />
P.1a : The gap between the price <strong>of</strong> the leaders <strong>and</strong> the average price <strong>of</strong><br />
the followers decreases with entry.<br />
When this prediction is satisfied in the data, one can look at the stronger<br />
result, which is supposed to hold for markets with competition in prices, <strong>and</strong><br />
test the following implication:<br />
P.1b: <strong>Market</strong> leaders facing exogenous entry choose higher prices than the<br />
followers; market leaders facing endogenous entry choose lower prices than<br />
the followers.<br />
Of course, the stronger hypothesis P.1b implies the weaker hypothesis<br />
P.1a, while the opposite is not true. Eventually, one could test further predictions<br />
<strong>of</strong> the basic model <strong>of</strong> Stackelberg competition with endogenous entry.<br />
For instance, in the presence <strong>of</strong> homogenous goods, increasing marginal costs,<br />
<strong>and</strong> competition in quantities we would expect that the equilibrium price corresponds<br />
to the marginal cost <strong>of</strong> the leader but it is higher than its average<br />
cost, while the same price is above the marginal cost <strong>of</strong> the marginal entrant<br />
but just enough to match its average cost. This is consistent with positive<br />
pr<strong>of</strong>its for the leader <strong>and</strong> endogenous entry. When one introduces product<br />
differentiation, also the equilibrium price for the leader is above its marginal<br />
cost according to a mark up which increases in the degree <strong>of</strong> product differentiation,<br />
but the equilibrium price <strong>of</strong> the followers is still equal to their average<br />
cost. These predictions could be tested against other hypotheses using the<br />
tools <strong>of</strong> the new empirical industrial organization, 10 <strong>and</strong> are summarized as<br />
follows:<br />
P.2: In a sector with homogenous goods <strong>and</strong> increasing marginal costs, the<br />
equilibrium price <strong>of</strong> a market leader facing endogenous entry is equal to its<br />
marginal cost <strong>and</strong> above its average cost, <strong>and</strong> the equilibrium price for the<br />
marginal entrant is above its marginal cost <strong>and</strong> equal to its average cost; an<br />
increase in product differentiation increases the equilibrium price above the<br />
marginal cost <strong>of</strong> the leader.<br />
Let us move to the case <strong>of</strong> strategic investments by market leaders. In<br />
Chapter 2 we have seen that, with competition in the market, entry conditions<br />
affect the way leaders undertake preliminary investments. In particular, using<br />
the classic taxonomy <strong>of</strong> business strategies, we have seen that leaders facing<br />
endogenous entry always act as top dogs or with a lean <strong>and</strong> hungry look,<br />
10 See Bresnahan (1987) <strong>and</strong> Berry et al. (2004) on the US automobile market, <strong>and</strong><br />
Kadiyali (1996) with particular reference to entry deterrence <strong>and</strong> accommodation<br />
in the US consumer market for photographic film in the period 1970-1990, when<br />
Kodak was the leader <strong>and</strong> Fuji the follower.
7.1 Empirical Predictions <strong>of</strong> the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders 249<br />
<strong>and</strong> never as puppy dogs or fat cats: ultimately, they are always aggressive<br />
compared to the entrants in the competition in the market, which is in line<br />
with the previous results. Between the many commitments we analyzed, some<br />
can be particularly interesting for empirical investigations. For instance, we<br />
analyzed cost reducing <strong>and</strong> dem<strong>and</strong> enhancing investments. From the first<br />
category we obtained neat predictions (leaders invest more in cost reducing<br />
activities when facing endogenous entry), <strong>and</strong> later we will revisit them again<br />
when dealing with more general forms <strong>of</strong> investments in R&D. From the<br />
second category we obtained results that depend crucially on the kind <strong>of</strong><br />
dem<strong>and</strong> enhancing investments under consideration.<br />
Consider product quality. Here, our focus will be on the implications <strong>of</strong><br />
the theory <strong>of</strong> market leaders in the presence <strong>of</strong> a double choice on both the<br />
quality <strong>and</strong> the price <strong>of</strong> the products. Summarizing the strategies with the<br />
quality-price ratio, a strong prediction deriving from our characterization<br />
(based on Prop. 3.9) would be the following:<br />
P.3: <strong>Market</strong> leaders facing an exogenous number <strong>of</strong> firms choose a lower<br />
quality-price ratio than the followers; market leaders facing endogenous entry<br />
choose a higher quality-price ratio than the followers.<br />
Given the complex strategic interactions emerging in a situation where<br />
firms choose multiple variables, it could be reasonable to limit the analysis to<br />
a weaker implication like the following: the quality-price ratio <strong>of</strong> the leaders<br />
increases with entry compared to the quality-price ratio <strong>of</strong> the followers.<br />
Another form <strong>of</strong> dem<strong>and</strong> enhancing investment is the expenditure on nonprice<br />
advertising aimed at increasing dem<strong>and</strong>. If we focus on markets with<br />
product differentiation <strong>and</strong> competition in prices, our theory (Prop. 2.5) implied<br />
the following strong testable prediction:<br />
P.4: <strong>Market</strong> leaders spend more than the followers in nonprice advertising<br />
(as a percentage <strong>of</strong> turnover) when the number <strong>of</strong> firmsisexogenous,<strong>and</strong><br />
less when entry is endogenous.<br />
Finally, after pointing out empirical implications for the policies concerning<br />
price, product <strong>and</strong> promotion, we emphasize an implication for the<br />
last strategic investment that characterizes the marketing mix <strong>of</strong> a firm (the<br />
fourth P), place which st<strong>and</strong>s for distribution. From our analysis on the choice<br />
<strong>of</strong> wholesale prices to retailers in the presence <strong>of</strong> downstream distribution<br />
channels (Prop. 2.9), we have the following prediction:<br />
P.5: <strong>Market</strong> leaders set higher wholesale prices for their retailers than<br />
their competitors when the number <strong>of</strong> firmsisexogenous,whiletheysetlower<br />
prices when entry is endogenous.<br />
Concerning financial issues, we need to take care <strong>of</strong> a more subtle differentiation<br />
on the source <strong>of</strong> uncertainty in the market, which can be used as an
250 7. Epilogue<br />
additional control variable. 11 Then, on the basis <strong>of</strong> our analysis (Prop. 2.6),<br />
we have the following prediction based on the hypothesis <strong>of</strong> competition in<br />
prices:<br />
P.6: The financial structure <strong>of</strong> market leaders is biased toward debt financing<br />
compared to the financial structure <strong>of</strong> the followers when the number<br />
<strong>of</strong> firms is exogenous, while it is biased toward equity financing when entry<br />
is endogenous, as long as uncertainty is mainly on the dem<strong>and</strong> side (while<br />
uncertainty on costs pushes the predictions in the opposite direction).<br />
Notice that under competition in quantities our model always implies<br />
abiastowarddebtfinancing for the leader, therefore, once again, we can<br />
distinguish a weaker hypothesis from the strong one stated above: the debtequity<br />
ratio <strong>of</strong> the market leaders should increase with entry.<br />
As we noticed earlier, investments in cost reductions aimed at reducing<br />
the price <strong>of</strong> a good give rise to neat predictions under competition in prices:<br />
in particular, market leaders should spend less than the other firms in R&D<br />
investments in cost reductions when the number <strong>of</strong> firms is exogenous, <strong>and</strong><br />
they should spend more when entry is endogenous. One should always keep<br />
in mind that this hypothesis holds under competition in prices, while under<br />
competition in quantities the leader would generally spend more than the<br />
followers in cost reductions under both entry conditions. However, we can<br />
generalize our result under general forms <strong>of</strong> competition for the market. In<br />
Chapter 4 we have seen that the theory <strong>of</strong> market leaders provides radical<br />
predictions concerning the incentives to invest in R&D by the firms already<br />
present in a market with the leading products. We can express the main<br />
implications in different ways. We start from the weakest possible prediction,<br />
which is already in contrast with the traditional result <strong>of</strong> the theory <strong>of</strong><br />
innovation: 12<br />
P.7: Incumbent market leaders facing endogenous entry in the competition<br />
for the market invest in R&D.<br />
11 An interesting related analysis on the effect <strong>of</strong> debt on prices is in Chevalier<br />
(1995), but that treatment does not take into account the source <strong>of</strong> uncertainty<br />
<strong>and</strong> the endogeneity <strong>of</strong> entry.<br />
12 See Malerba <strong>and</strong> Orsenigo (1999), Blundell et al. (1999), Czarnitzki <strong>and</strong> Kraft<br />
(2007a) <strong>and</strong> Hughes (2007) on evidence on the high investment in R&D by<br />
market leaders. The empirical study by Blundell et al. (1999) witnesses a positive<br />
relationship between market power <strong>and</strong> innovation activity, which is consistent<br />
with strategic investment in R&D by the leaders. This result holds in a panel<br />
data with many sectors, but especially in the pharmaceutical sector, which is a<br />
sector with a high R&D-sales ratio, strong patent protection <strong>and</strong> where firms<br />
typically recognize that they are in races to develop innovations (in this sector<br />
a few drug companies on their own undertake truly innovative research, <strong>and</strong> a<br />
number <strong>of</strong> mergers in the mid-90s were even motivated to enhance leadership in<br />
the patent races).
7.1 Empirical Predictions <strong>of</strong> the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders 251<br />
Of course the theory <strong>of</strong> market leaders suggests more than this. First <strong>of</strong><br />
all, we saw that leaders invest more than the followers when entry is endogenous.<br />
This was true in all <strong>of</strong> our models <strong>of</strong> competition for the market,<br />
independently from the kind <strong>of</strong> strategic interaction between firms (remember<br />
that investments could be strategic substitutes or complements in alternative<br />
models). Therefore, we state this as an intermediate hypothesis:<br />
P.8: The investment rate in R&D <strong>of</strong> market leaders is higher than the average<br />
investment rate in R&D <strong>of</strong> the followers when entry in the competition<br />
for the market is endogenous.<br />
We also have a radically strong hypothesis that derives from our favorite<br />
model in which the investments <strong>of</strong> the firms are strategic complements: 13<br />
P.9: <strong>Market</strong> leaders invest less in R&D (as a percentage <strong>of</strong> turnover) than<br />
the followers when the competition for the market is between an exogenous<br />
number <strong>of</strong> firms, they invest more when entry is endogenous.<br />
Finally, our theory <strong>of</strong> sequential innovation by leaders suggests a way<br />
to discriminate between different degrees <strong>of</strong> persistence <strong>of</strong> leadership in innovative<br />
sectors. When entry <strong>of</strong> firms in the competition for the market is<br />
endogenous we should expect that technological leaders invest a lot <strong>and</strong> their<br />
persistence is more likely. Of course, when there is no competition for the<br />
market we should expect that the monopolistic leadership is also persistent.<br />
However, when the degree <strong>of</strong> competition for the market is intermediate (entry<br />
is not free but more than one firm invests), we should expect that the<br />
incumbent does not invest much in R&D <strong>and</strong> that its leadership is more likely<br />
to be replaced. This suggests our last prediction:<br />
P. 10: The degree <strong>of</strong> persistence <strong>of</strong> leadership should follow <strong>and</strong> inverted<br />
U relation with the degree <strong>of</strong> entry in the competition for the market.<br />
These testable implications could be brought to the data in future research.<br />
Of course, the analysis <strong>of</strong> this book was limited to the issues that<br />
we considered interesting to underst<strong>and</strong> the behavior <strong>of</strong> market leaders <strong>and</strong><br />
to derive implications for antitrust policy. Many other issues could be studied<br />
through the market leaders approach <strong>and</strong>, accordingly, other empirical<br />
implications could be derived <strong>and</strong> eventually tested.<br />
13 Tournaments for team sports are an ideal context to test the theory <strong>of</strong> innovation<br />
by leaders since entry is definitely endogenous in these contests. Casual evidence<br />
from Formula 1 racing, the America’s Cup, or the European soccer Champions<br />
League suggests that leading teams do invest more than the followers <strong>and</strong> their<br />
leadership is partially persistent. Sport economics may investigate further the<br />
issue.
252 7. Epilogue<br />
7.2 Implications for Business Administration<br />
In this book we looked at the behavior <strong>of</strong> market leaders from a descriptive<br />
point <strong>of</strong> view. The attempt was to underst<strong>and</strong> how leaders behave in different<br />
competitive scenarios. There is another way to read the results <strong>of</strong> this book.<br />
This is the point <strong>of</strong> view <strong>of</strong> business administration: our results may suggest a<br />
rule <strong>of</strong> behavior for the management <strong>of</strong> the leading firms. A vast literature on<br />
marketing (see Kotler, 1999) <strong>and</strong> business strategy (see Porter, 1985) exists<br />
which questions what should be the optimal marketing mix <strong>and</strong> the optimal<br />
strategic investments. Here we have emphasized that the right answer may<br />
depend on the entry conditions in a crucial way.<br />
Consider a generic market where products are highly differentiated <strong>and</strong><br />
firms compete in prices. When entry in the market is limited to a predetermined<br />
number <strong>of</strong> firms <strong>and</strong> there are pr<strong>of</strong>itable opportunities for all <strong>of</strong><br />
these firms, the management should always follow an accommodating philosophy.<br />
This requires high prices, high investments in advertising, delegation<br />
<strong>of</strong> the distribution to downstream sellers with high wholesale prices, limited<br />
investments in R&D <strong>and</strong> expansion through horizontal mergers. In such a<br />
situation, an aggressive management is counterproductive because it induces<br />
retaliation by the rivals <strong>and</strong> reduces pr<strong>of</strong>its in the long run.<br />
The optimal management rule changes radically when the market is characterized<br />
by high reactivity <strong>of</strong> entry to the pr<strong>of</strong>itable conditions. In such a<br />
case, entry is pervasive <strong>and</strong> can reduce pr<strong>of</strong>its at low levels, but a firm can acquire<br />
a leadership <strong>and</strong> preserve high pr<strong>of</strong>its by adopting a correct marketing<br />
philosophy. This must be an aggressive philosophy, which requires the exact<br />
opposite <strong>of</strong> what we have seen before: low prices, low investments in advertising,<br />
delegation <strong>of</strong> the distribution with wholesale prices below cost <strong>and</strong> high<br />
fees, high investments in R&D <strong>and</strong> expansion through internal growth.<br />
The general principle is that the management should follow an aggressive<br />
or an accommodating philosophy according to the entry conditions, be able<br />
to monitor these conditions <strong>and</strong> adopt the right marketing mix <strong>and</strong> strategies<br />
accordingly.<br />
7.3 Implications for Economic <strong>Theory</strong><br />
In this section we would like to emphasize a few areas where further theoretical<br />
research on the theory <strong>of</strong> market leaders could be fruitful. The model <strong>of</strong><br />
Stackelberg competition with endogenous entry can be generalized in many<br />
other dimensions <strong>and</strong> applied to other specific forms <strong>of</strong> market structures.<br />
The same holds for the model <strong>of</strong> strategic investment with endogenous entry.<br />
Our discussions <strong>of</strong> the investments in cost reducing activities, in persuasive<br />
advertising <strong>and</strong> our analysis <strong>of</strong> the optimal financial structure related to the<br />
competition in the market were purely introductory <strong>and</strong> need further investigations.<br />
The literature on multi-sided markets is just taking <strong>of</strong>f, therefore we
7.3 Implications for Economic <strong>Theory</strong> 253<br />
look forward to further applications. Concerning bundling, vertical restraints<br />
for interbr<strong>and</strong> competition, price discrimination <strong>and</strong> horizontal mergers we<br />
limited our results to basic attempts to approach these issues within a new<br />
perspective: further studies should generalize the outcomes <strong>and</strong>, most <strong>of</strong> all,<br />
verify their applicability for antitrust purposes. Growing literature on innovation<br />
by leaders already exists, but further theoretical work is needed in<br />
this field as well, especially for the underst<strong>and</strong>ing <strong>of</strong> the relationship between<br />
competitionforthemarket<strong>and</strong>inthemarket.Finally,wehaveseenthatthe<br />
theory <strong>of</strong> endogenous entry has some general consequences for the approach<br />
to antitrust policy, <strong>and</strong> may limit the validity <strong>of</strong> some <strong>of</strong> the implications <strong>of</strong><br />
the post-Chicago approach. Hopefully, our results will be useful in improving<br />
our underst<strong>and</strong>ing <strong>of</strong> the behavior <strong>of</strong> market leaders under different entry<br />
conditions, <strong>and</strong> in deriving a unified economic approach to antitrust issues.<br />
Interesting results could be developed in the field <strong>of</strong> government policy<br />
aimed at helping domestic firms in international markets: we briefly analyzed<br />
the choice <strong>of</strong> the optimal state aids <strong>and</strong> subsidies for exporting firms<br />
<strong>and</strong> some issues concerning privatizations, but many other issues could be<br />
investigated. One could extend the analysis to more general forms <strong>of</strong> export<br />
promoting policies, as general forms <strong>of</strong> strategic trade policy, competitive<br />
devaluations in partial equilibrium <strong>and</strong> R&D subsidies in an international<br />
competition for the market. 14 Therole<strong>of</strong>non-pr<strong>of</strong>it firms could be investigated<br />
in an analogous way to our analysis <strong>of</strong> public firms. Furthermore, it<br />
would be interesting to analyze further the Schumpeterian model sketched<br />
in our analysis on sequential innovations with leaders driving technological<br />
progress <strong>and</strong> growth.<br />
The analysis <strong>of</strong> full-fledged models <strong>of</strong> competition for the market with<br />
endogenous entry, <strong>and</strong> eventually with asymmetries between leaders <strong>and</strong><br />
followers, leads naturally to other applications in macroeconomics. Recent<br />
newkeynesian research has been limited to the analysis <strong>of</strong> competition in<br />
prices between an exogenous number <strong>of</strong> firms (or in situations where the<br />
number <strong>of</strong> firms was indeterminate or irrelevant), <strong>and</strong> has introduced sticky<br />
prices in this environment. We hope that this book contributes to suggest<br />
the relevance <strong>of</strong> endogenous entry under both competition in prices <strong>and</strong> in<br />
quantities, <strong>and</strong> the relevance <strong>of</strong> asymmetries between leaders <strong>and</strong> followers<br />
which can have important consequences over the business cycles. Moreover,<br />
we believe that the study <strong>of</strong> sticky entry in these markets can have interesting<br />
consequences for the same underst<strong>and</strong>ing <strong>of</strong> the business cycle (maybe more<br />
than the usual study <strong>of</strong> sticky prices). For instance, we have seen that entry<br />
heavily affectsprices<strong>and</strong>markupsinmarketswithrelevantfixed costs <strong>of</strong><br />
production, <strong>and</strong> endogenous entry affects the pricing behavior <strong>of</strong> the existing<br />
14 While we neglected general equilibrium considerations, a deeper analysis <strong>of</strong> endogenous<br />
entry with market leaders in a 2 × 2 × 2 model could be fruitful.
254 7. Epilogue<br />
firms (which can be seen as leaders in such a case). These factors could be<br />
fruitfully introduced in the general equilibrium macroeconomic analysis. 15<br />
Potentially, one could also apply the principles <strong>of</strong> the behavior <strong>of</strong> market<br />
leaders in other contexts <strong>of</strong> economic theory. The simple contest where<br />
a leader <strong>and</strong> its followers exert effort to obtain a compensation can be introduced<br />
in a principal-agent framework. A principal could choose the compensation<br />
for the agent that achieves the desired result, <strong>and</strong> could also choose the<br />
appropriate hierarchical structure between agents. Assigning a leadership to<br />
one <strong>of</strong> the agents may reduce total effort <strong>and</strong> the probability <strong>of</strong> achieving the<br />
desired result, but it may also avoid the waste in fixed costs <strong>of</strong> participation<br />
associated with multiple agents.<br />
The model <strong>of</strong> competition for a prize can also be re-interpreted in terms <strong>of</strong><br />
a rent-seeking contest, <strong>and</strong> it foresees that incumbent lobbyists should invest<br />
more than entrants when there are no barriers to rent seeking.<br />
Political leadership can be analyzed in a related way as a function <strong>of</strong> the<br />
entry conditions in the electoral competition. One could think <strong>of</strong> incumbent<br />
politicians running for new elections in a parallel way to our incumbent monopolists<br />
that are leaders in the competition for the market. The strategies<br />
<strong>of</strong> competing politicians would affect the incentives <strong>of</strong> the political leaders to<br />
engage in the electoral campaign, spend resources <strong>and</strong> effort in fund raising,<br />
promote the endorsement <strong>of</strong> opinion makers, <strong>and</strong> finally (if possible) commit<br />
to challenging policies <strong>and</strong> promises that benefit the citizens. In the case<br />
<strong>of</strong> an electoral competition between two predetermined parties or coalitions<br />
where there is no space for the entry <strong>of</strong> other parties we could expect a systematic<br />
leapfrogging <strong>of</strong> the political opposition on the incumbent party. In<br />
the presence <strong>of</strong> electoral systems allowing for endogenous entry <strong>of</strong> c<strong>and</strong>idates<br />
(when there are chances to replace the political contestants) we could expect<br />
the incumbent politicians to engage in more aggressive political campaigns<br />
to preserve political power. 16<br />
Finally, we cannot exclude that the role <strong>of</strong> entry conditions in inducing<br />
aggressive or accommodating behavior extends to more general interactions<br />
between persons, like those emerging in small communities, clubs, or circle <strong>of</strong><br />
friends where leaders <strong>and</strong> followers interact to achieve personal satisfaction.<br />
The wide development <strong>of</strong> behavioral <strong>and</strong> psychological economics in the last<br />
years, may find related results in the study <strong>of</strong> social interactions. These could<br />
be the subject <strong>of</strong> further empirical investigations in experimental economics.<br />
15 See Bilbiie et al. (2007) for an important macroeconomic analysis <strong>of</strong> endogenous<br />
entry in the competition in the market, <strong>and</strong> Etro (2007,a) on endogenous entry<br />
also in the competition for the market.<br />
16 Of course, this parallel is limited by the ideological component which is present<br />
in political competitions, <strong>and</strong> it is confined to democratic political systems.
7.4 Conclusions 255<br />
7.4 Conclusions<br />
Entry manages to discipline competition more than any government policy.<br />
In particular, endogenous entry forces market leaders to act in an aggressive<br />
or pro-competitive way that creates benefits for the consumers, avoids welfare<br />
reducing mergers <strong>and</strong> can even reduce the effectiveness <strong>of</strong> collusive cartels.<br />
These results are quite close to those <strong>of</strong> the Chicago school, but have found a<br />
game theoretic formalization in this book <strong>and</strong> in the emerging literature on<br />
endogenous entry.<br />
On this basis we believe that industrial policy should be aimed at preserving<br />
free entry conditions <strong>and</strong> promoting competition in <strong>and</strong> for all the<br />
markets. More specifically, antitrust policy should intervene only in markets<br />
where entry is exogenously blocked or in which a firm attempts to build artificial<br />
barriers to entry. Other than that, we believe that the invisible h<strong>and</strong><br />
<strong>of</strong> endogenous entry drives markets better than any other exogenous intervention.<br />
With these considerations we have arrived to the end <strong>of</strong> our book on<br />
market leaders <strong>and</strong> endogenous entry. As Jerry Seinfeld once said, “the big<br />
advantage <strong>of</strong> a book is it’s very easy to rewind: close it <strong>and</strong> you’re right back<br />
at the beginning.”
8. References<br />
Abreu, Dilip, 1986, Extremal Equilibria <strong>of</strong> Oligopoly Supergames, Journal <strong>of</strong><br />
Economic <strong>Theory</strong>, 39, 191-225<br />
Acemoglu, Daron <strong>and</strong> Joshua Linn, 2004, <strong>Market</strong> Size in <strong>Innovation</strong>: <strong>Theory</strong><br />
<strong>and</strong> Evidence from the Pharmaceutical Industry, Quarterly Journal <strong>of</strong><br />
Economics, 119, 3, 1049-90<br />
Aerts, Kris <strong>and</strong> Tobias Schmidt, 2007, Two for the Price <strong>of</strong> One? On Additionality<br />
Effects <strong>of</strong> R&D Subsidies: a Comparison between Fl<strong>and</strong>ers <strong>and</strong><br />
Germany, mimeo, Katholieke Universiteit Leuven, presented at the 2007<br />
<strong>Intertic</strong> Conference<br />
Aghion, Philippe, Nick Bloom, Richard Blundell, Rachel Griffith <strong>and</strong> Peter<br />
Howitt, 2005, <strong>Competition</strong> <strong>and</strong> <strong>Innovation</strong>: an Inverted-U Relationship,<br />
Quarterly Journal <strong>of</strong> Economics, 2, 701-28<br />
Aghion, Philippe, Richard Blundell, Rachel Griffith, Peter Howitt <strong>and</strong> Susanne<br />
Prantl, 2006, The Effects <strong>of</strong> Entry on Incumbent <strong>Innovation</strong> <strong>and</strong><br />
Productivity, NBER wp 12027<br />
Aghion, Philippe <strong>and</strong> Griffith, Rachel, 2005, <strong>Competition</strong> <strong>and</strong> Growth. Reconciling<br />
<strong>Theory</strong> <strong>and</strong> Evidence, Cambridge, The MIT Press<br />
Aghion, Philippe, Christopher Harris, Peter Howitt <strong>and</strong> John Vickers, 2001,<br />
<strong>Competition</strong>, Imitation, <strong>and</strong> Growth with Step-by-Step <strong>Innovation</strong>, Review<br />
<strong>of</strong> Economic Studies, 68, 467-92<br />
Aghion, Philippe <strong>and</strong> Peter Howitt, 1992, A Model <strong>of</strong> Growth Through Creative<br />
Destruction, Econometrica, 60, 2, 323-51<br />
Aghion, Philippe <strong>and</strong> Peter Howitt, 1998, Endogenous Growth <strong>Theory</strong>, Cambridge,<br />
The MIT Press<br />
Aghion, Philippe <strong>and</strong> Salvatore Modica, 2006, Open-source without Free Riding,<br />
mimeo, Harvard University, presented at the 2007 <strong>Intertic</strong> Conference<br />
Ahlborn, Christian, David Evans <strong>and</strong> Jorge Padilla, 2004, The <strong>Antitrust</strong> Economics<br />
<strong>of</strong> Tying: a Farewell to per se Illegality, The <strong>Antitrust</strong> Bulletin,<br />
Spring-Summer, 287-341<br />
Akerl<strong>of</strong>, George, 1970, The <strong>Market</strong> for Lemons: Quality Uncertainty <strong>and</strong> the<br />
<strong>Market</strong> Mechanism, Quarterly Journal <strong>of</strong> Economics 89, 488-500<br />
Ambrosanio, Flavia, Massimo Bordignon <strong>and</strong> Federico Etro, 2004, Investments<br />
in Infrastructures, Cifrel, Milan
258 8. References<br />
Amir, Rabah, 2005, <strong>Market</strong> Structure, Scale Economies <strong>and</strong> Industry Performance,<br />
mimeo, University <strong>of</strong> Arizona, presented at the 2007 <strong>Intertic</strong><br />
Conference<br />
Amir, Rabah <strong>and</strong> Eugen Lambson, 2000, On the Effects <strong>of</strong> Entry in Cournot<br />
<strong>Market</strong>s, Review <strong>of</strong> Economic Studies, 67, 235-54<br />
Amir, Rabah <strong>and</strong> Natalia Lazzati, 2007, Cournot Oligopoly with Network<br />
Effects, “Stackelberg Lecture” presented at the 2007 <strong>Intertic</strong> Conference<br />
Anderson, Simon <strong>and</strong> Andrè de Palma, 1992, The Logit as a Model <strong>of</strong> Product<br />
Differentiation, Oxford Economic Papers, 44, 1, 51-67<br />
Anderson, Simon, Andrè de Palma <strong>and</strong> Jean Francois Thisse, 1988, A Representative<br />
Consumer <strong>Theory</strong> <strong>of</strong> the Logit Model, International Economic<br />
Review, 29, 461-66<br />
Anderson, Simon, Andrè de Palma <strong>and</strong> Jean Francois Thisse, 1992, Discrete<br />
Choice <strong>Theory</strong> <strong>of</strong> Product Differentiation, Cambridge, The MIT Press<br />
Anderson, Simon, Andrè de Palma <strong>and</strong> Jean Francois Thisse, 1997, Privatization<br />
<strong>and</strong> Efficiency in a Differentiated Industry, European Economic<br />
Review, 41, 1635-54<br />
Anderson, Simon <strong>and</strong> Maxim Engers, 1994, Strategic Investment <strong>and</strong> Timing<br />
<strong>of</strong> Entry, International Economic Review, 35, 4, 833-53<br />
Areeda, Phillip <strong>and</strong> Donald Turner, 1975, Predatory Pricing <strong>and</strong> Related<br />
Practices under Section 2 <strong>of</strong> the Sherman Act, Harvard Law Review, 88,<br />
637-733<br />
Armstrong, Mark, 2006, <strong>Competition</strong> in Two-Sided <strong>Market</strong>s, The RAND<br />
Journal <strong>of</strong> Economics, inpress<br />
Arrow, Kenneth, 1962, Economic Welfare <strong>and</strong> the Allocation <strong>of</strong> Resources<br />
for Invention, in The Rate <strong>and</strong> Direction <strong>of</strong> Innovative Activity, Editedby<br />
R. Nelson, Princeton University Press, Princeton<br />
Aumann, Robert <strong>and</strong> Lloyd Shapley, 1976, Long Term <strong>Competition</strong>-A Game<br />
Theoretic Analysis, mimeo, UCLA<br />
Baarsma, Barbara, Joost Ooort, Coen Teulings <strong>and</strong> Michiel de Nooij, 2004,<br />
Private Investment in New Infrastructures, European Union, Bruxelles<br />
Bain, Joe, 1956, Barriers to New <strong>Competition</strong>: their character <strong>and</strong> consequences<br />
in manufacturing industry, Harvard University Press, Cambridge<br />
Barone, Guglielmo <strong>and</strong> Federico Cingano, 2007, Service Regulation <strong>and</strong><br />
Growth: Evidence from Cross-Country Data, mimeo, Bank <strong>of</strong> Italy<br />
Barro, Robert, 1998, Why the <strong>Antitrust</strong> Cops should Lay <strong>of</strong>f Hight Tech,<br />
Business Week, August 17, 20<br />
Barro, Robert <strong>and</strong> Xavier Sala-i-Martin, 1995, Economic Growth, McGraw<br />
Hill, New York<br />
Baumol, William, John Panzar <strong>and</strong> Robert Willig, 1982, Contestable <strong>Market</strong>s<br />
<strong>and</strong> the <strong>Theory</strong> <strong>of</strong> Industry Structure, San Diego, Harcourt Brace<br />
Jovanovich<br />
Baxter, William, 1983, Bank Interchange <strong>of</strong> Transactional Paper: Legal <strong>and</strong><br />
Economic Perspectives, Journal <strong>of</strong> Law & Economics, 26, 3, 541-88
8. References 259<br />
Beath, John, Yannis Katsoulakos <strong>and</strong> David Ulph, 1989, Strategic R&D Policy,<br />
TheEconomicJournal, 99, 395, 74-83<br />
Becker, Gary <strong>and</strong> Kevin Murphy, 1993, A Simple <strong>Theory</strong> <strong>of</strong> Advertising as a<br />
Good or Bad, Quarterly Journal <strong>of</strong> Economics, 108, 4, 941-64<br />
Behrens, Kristian <strong>and</strong> Yasusada Murata, 2007, General Equilibrium Models<br />
<strong>of</strong> Monopolistic <strong>Competition</strong>: A New Approach, Journal <strong>of</strong> Economic<br />
<strong>Theory</strong>, inpress<br />
Bergman Mats <strong>and</strong> Niklas Rudholm, 2003, The Relative Importance <strong>of</strong> Actual<br />
<strong>and</strong> Potential <strong>Competition</strong>: Empirical Evidence from the Pharmaceuticals<br />
<strong>Market</strong>, Journal <strong>of</strong> Industrial Economics, 51, 4, 455-67<br />
Berry, Stephen, James Levinsohn <strong>and</strong> Ariel Pakes, 2004, Differentiated Products<br />
Dem<strong>and</strong> Systems From a Combination <strong>of</strong> Micro <strong>and</strong> Macro Data: the<br />
New Car <strong>Market</strong>, Journal <strong>of</strong> Political Economy, 112, 1, 68-105<br />
Bertr<strong>and</strong>, Joseph, 1883, Théorie des Richesses: revue de Théories mathématiques<br />
de la richesse sociale par Léon Walras et Recherches sur les principes<br />
mathématiques de la théorie des richesses par Augustin Cournot, Journal<br />
des Savants, 499-508<br />
Bessen, Jim <strong>and</strong> Eric Maskin, 2002, Sequential <strong>Innovation</strong>, Patents <strong>and</strong> Imitation,mimeo,MIT<br />
Bilbiie, Florin, Fabio Ghironi <strong>and</strong> Marc Melitz, 2007, Endogenous Entry,<br />
Product Variety, <strong>and</strong> Business Cycles, mimeo, Princeton University<br />
Blundell, Richard, Rachel Griffith <strong>and</strong> John Van Reenen, 1999, <strong>Market</strong> Share,<br />
<strong>Market</strong> Value <strong>and</strong> <strong>Innovation</strong> in a Panel <strong>of</strong> British Manufacturing Firms,<br />
Review <strong>of</strong> Economic Studies, 66, 529-54<br />
Blundell, Richard, Rachel Griffith <strong>and</strong> Frank Windmeijer, 2002, Individual<br />
Effects <strong>and</strong> Dynamics in Count Data Models, Journal <strong>of</strong> Econometrics,<br />
108, 1, 113-31<br />
Boldrin, Michele <strong>and</strong> David Levine, 1998, Growth under Perfect <strong>Competition</strong>,<br />
mimeo, U.C.L.A.<br />
Boldrin, Michele <strong>and</strong> David Levine, 2005, Against Intellectual Monopoly,<br />
manuscript, U.C.L.A. <strong>and</strong> University <strong>of</strong> Minnesota<br />
Bonanno, Giacomo <strong>and</strong> John Vickers, 1988, Vertical Separation, Journal <strong>of</strong><br />
Industrial Economics, 36, 257-265<br />
Boone, Jan, Delia Ionaşcu <strong>and</strong> Krešimir Žigić, 2006, Trade Policy, <strong>Market</strong><br />
Leaders <strong>and</strong> Endogenous <strong>Competition</strong> Intensity, mimeo, <strong>Intertic</strong><br />
Bork, Robert, 1993, The <strong>Antitrust</strong> Paradox. A Policy at War with Itself, The<br />
Free Press, New York<br />
Bowley, A., 1924, The Mathematical Groundwork <strong>of</strong> Economics, Oxford,Oxford<br />
University Press<br />
Boycko, Maxim, Andrei Shleifer <strong>and</strong> Robert Vishny, 1997, Privatizing Russia,<br />
Cambridge, The MIT Press<br />
Br<strong>and</strong>er, James <strong>and</strong> Tracy Lewis, 1986, Oligopoly <strong>and</strong> Financial Structure:<br />
The Limited Liability Effect, The American Economic Review, 76, 5, 956-<br />
70
260 8. References<br />
Br<strong>and</strong>er, James <strong>and</strong> Tracy Lewis, 1988, Bankruptcy Costs <strong>and</strong> the <strong>Theory</strong><br />
<strong>of</strong> Oligopoly, Canadian Journal <strong>of</strong> Economics, 21, 2, 221-43<br />
Br<strong>and</strong>er, Jim <strong>and</strong> Barbara Spencer, 1985, Export Subsidies <strong>and</strong> International<br />
<strong>Market</strong> Share Rivalry, Journal <strong>of</strong> International Economics, 16, 83-100<br />
Brealey, Richard <strong>and</strong> Stewart Myers, 2002, Principles <strong>of</strong> Corporate Finance,<br />
New York, Mc Graw Hill<br />
Bresnahan, Timothy, 1987, <strong>Competition</strong> <strong>and</strong> Collusion in the American Automobile<br />
Industry: The 1995 Price War, Journal <strong>of</strong> Industrial Economics,<br />
35, 457-82<br />
Bresnahan, Timothy, 2001, The Economics <strong>of</strong> the Micros<strong>of</strong>t Case, mimeo,<br />
Stanford University<br />
Bulow, Jeremy, John Geanakoplos <strong>and</strong> Paul Klemperer, 1985, Multimarket<br />
Oligopoly: Strategic Substitutes <strong>and</strong> Complements, Journal <strong>of</strong> Political<br />
Economy, 93, 3, 488-511<br />
Cable, John <strong>and</strong> Dennis Mueller, 2006, Testing for Persistence <strong>of</strong> Pr<strong>of</strong>its’<br />
Differences Across Firms, mimeo, University <strong>of</strong> Vienna<br />
Caillaud, Bernard <strong>and</strong> Bruno Jullien, 2003, Chicken <strong>and</strong> Egg: <strong>Competition</strong><br />
among Intermediation Service Providers, The RAND Journal <strong>of</strong> Economics,<br />
34, 2, 521-52<br />
Cambini, Carlo <strong>and</strong> Tommaso Valletti, 2007, Information Exchange <strong>and</strong><br />
<strong>Competition</strong> in Communications Networks, mimeo, Imperial College, London,<br />
presented at the 2007 <strong>Intertic</strong> Conference<br />
Carlton, Dennis <strong>and</strong> Jeffrey Perl<strong>of</strong>f, 2004, Modern Industrial Organization,<br />
Pearson, Addison Wesley<br />
Cawley John <strong>and</strong> Tomas Philipson, 1999, An Empirical Examination <strong>of</strong> Information<br />
Barriers to Trade in Insurance, American Economic Review 89,<br />
4, 827-46<br />
Cenciarini, Renzo, Maurizio Dallocchio, Alberto Dell’Acqua <strong>and</strong> Leonardo<br />
Etro, Eds, 2006, A Bridge for Growth. Firms, banks <strong>and</strong> corporate finance<br />
for the future <strong>of</strong> the Italian system, Gangemi Ed., Rome<br />
Chamberlin, Edward, 1933, The <strong>Theory</strong> <strong>of</strong> Monopolistic <strong>Competition</strong>, Cambridge,<br />
Harvard University Press<br />
Ch<strong>and</strong>ler, Alfred,1990, Scale <strong>and</strong> Scope: The Dynamics <strong>of</strong> Industrial Capitalism,<br />
Cambridge, Harvard University Press<br />
Chaudhuri, Anindya <strong>and</strong> Kenneth Flamm, 2004, Price, <strong>Competition</strong> <strong>and</strong> Regulating<br />
the Net, mimeo, University <strong>of</strong> Texas at Austin<br />
Chaudhuri, Anindya <strong>and</strong> Kenneth Flamm, 2005, The <strong>Market</strong> Structure <strong>of</strong><br />
Internet Service Provision, mimeo, University <strong>of</strong> Texas at Austin<br />
Chevalier, Judith, 1995, Do LBO Supermarkets Charge More? An Empirical<br />
Analysis <strong>of</strong> the Effects <strong>of</strong> LBOs on Supermarket Pricing, Journal <strong>of</strong><br />
Finance, 50, 4, 1095-1112<br />
Chiappori, Pierre-Andre’ <strong>and</strong> Bernard Salanie’, 2000, Testing for Asymmetric<br />
InformationinInsurance<strong>Market</strong>s,Journal <strong>of</strong> Political Economy 108, 58-78
8. References 261<br />
Coase, Ronald, 1960, The Problem <strong>of</strong> Social Cost, Journal <strong>of</strong> Law <strong>and</strong> Economics,<br />
3,1-44<br />
Cohen, Wesley <strong>and</strong> Steven Klepper, 1996, A Reprise <strong>of</strong> Size <strong>and</strong> R&D, The<br />
Economic Journal, 106, 437, 925-51<br />
Cournot, Augustin, 1838, Researches sur les Principles Mathematiques de la<br />
Theorie des Richesses, Paris<br />
Cowell, Frank, 2004, Carrots <strong>and</strong> Sticks in Enforcement, in The Crisis in Tax<br />
Administration, Edited by H. Aaron, J. Slemrod, The Brookings Institution,<br />
pp. 230-275<br />
Cozzi, Guido, 2007, The Arrow Effect under Competitive R&D, The B.E.<br />
Journal <strong>of</strong> Macroeconomics, 7, 1 (Contributions), Art. 2<br />
Cozzi, Guido <strong>and</strong> Silvia Galli, 2007, Changing the Research Patenting Regime:<br />
Schumpeterian Explanation, mimeo, University <strong>of</strong> Rome, Tor Vergata, presented<br />
at the 2007 <strong>Intertic</strong> Conference<br />
Cremer, Jacques, Patrick Rey, <strong>and</strong> Jean Tirole, 2000, Connectivity in the<br />
Commercial Internet, Journal <strong>of</strong> Industrial Economics, 48, 433-72<br />
Czarnitzki, Dirk <strong>and</strong> Kornelius Kraft, 2004, An Empirical Test <strong>of</strong> the Asymmetric<br />
Models on Innovative Activity: Who Invests more into R&D, the<br />
Incumbent or the Challenger?, Journal <strong>of</strong> Economic Behavior & Organization,<br />
54, 2, 153-73<br />
Czarnitzki, Dirk <strong>and</strong> Kornelius Kraft, 2007a, License Expenditures <strong>of</strong> Incumbents<br />
<strong>and</strong> Potential Entrants: An Empirical Analysis <strong>of</strong> Firm Behavior,<br />
mimeo, University <strong>of</strong> Dortmund, presented at the 2007 <strong>Intertic</strong> Conference<br />
Czarnitzki, Dirk <strong>and</strong> Kornelius Kraft, 2007b, The Incentives <strong>of</strong> Oligopolists<br />
<strong>and</strong> Challengers to Acquire a New Technology, mimeo, University <strong>of</strong> Dortmund,<br />
presented at the 2007 <strong>Intertic</strong> Conference<br />
Dallocchio, Maurizio, Alberto Dell’Acqua, Leonardo Etro, Giulia Ligresti <strong>and</strong><br />
Barbara Rovetta, Italian Br<strong>and</strong>s, EGEA, Milan<br />
Dasgupta, Partha <strong>and</strong> Joseph E. Stiglitz, 1980, Uncertainty, Industrial Structure<br />
<strong>and</strong> the Speed <strong>of</strong> R&D, Bell Journal <strong>of</strong> Economics, 1-28<br />
D’Aspremont, Claude, Jean Gabszewicz, <strong>and</strong> Jean Francois Thisse, 1979, On<br />
Hotelling’s Stability in <strong>Competition</strong>, Econometrica, 47, 1145-50<br />
Davidson, Carl <strong>and</strong> Arijit, Mukherjee, 2007, Horizontal Mergers with Free<br />
Entry, International Journal <strong>of</strong> Industrial Organization, 25, 1, 157-172<br />
Davis, Stephen, Jack MacCrisken <strong>and</strong> Kevin Murphy, 2002, Economic Perspectives<br />
on S<strong>of</strong>tware Design: PC Operating Systems <strong>and</strong> Platforms, in<br />
Micros<strong>of</strong>t, <strong>Antitrust</strong> <strong>and</strong> the New Economy: Selected Essays, Editedby<br />
David Evans, Norwell, MA, Kluwer Academic Publishers, 361-419<br />
Davis, Stephen <strong>and</strong> Kevin Murphy, 2000, A Competitive Perspective on Internet<br />
Explorer, The American Economic Review, Papers <strong>and</strong> Procedeengs,<br />
90, 2, 184-87<br />
De Bondt, Raymond <strong>and</strong> Jan V<strong>and</strong>ekerckhove, 2007, <strong>Innovation</strong> by Leaders<br />
without Wninner-take-all, mimeo, Catholic University <strong>of</strong> Leuven
262 8. References<br />
Degryse, Hans <strong>and</strong> Steven Onega, 2004, The Impact <strong>of</strong> <strong>Competition</strong> on Bank<br />
Orientation <strong>and</strong> Specialization, mimeo, Tilburg University<br />
Delipalla, Sophia <strong>and</strong> Michael Keen, 1992, The Comparison between Ad Valorem<br />
<strong>and</strong> Specific Taxation under Imperfect <strong>Competition</strong>, Journal <strong>of</strong> Public<br />
Economics, 49, 351-67<br />
Demstez, Harold, 1973, Industry Structure, <strong>Market</strong> Rivalry, <strong>and</strong> Public Policy,<br />
Journal <strong>of</strong> Law & Economics, 16, 1, 1-9<br />
Deneckere, Raymond <strong>and</strong> Carl Davidson, 1985, Incentives to Form Coalitions<br />
with Bertr<strong>and</strong> <strong>Competition</strong>, The RAND Journal <strong>of</strong> Economics, 16, 4, 473-<br />
86,<br />
Denicolò, Vincenzo, 2001, Growth with Non-Drastic <strong>Innovation</strong>s <strong>and</strong> the Persistence<br />
<strong>of</strong> Leadership, European Economic Review, 45, 8, 1399-1414<br />
Denicolò, Vincenzo, 2007, Do Patents Over-compensate Innovators?, mimeo,<br />
University <strong>of</strong> Bologna, presented at the 2007 <strong>Intertic</strong> Conference<br />
Denicolò, Vincenzo <strong>and</strong> Piercarlo Zanchettin, 2006, <strong>Competition</strong>, Darwinian<br />
Selection <strong>and</strong> Growth, mimeo, University <strong>of</strong> Bologna, presented at the 2007<br />
<strong>Intertic</strong> Conference<br />
Dixit, Avinash, 1979, A Model <strong>of</strong> Duopoly Suggesting a <strong>Theory</strong> <strong>of</strong> Entry<br />
Barriers, The Bell Journal <strong>of</strong> Economics, 10, 1, 20-32<br />
Dixit, Avinash, 1980, The Role <strong>of</strong> Investment in Entry-Deterrence, The Economic<br />
Journal, 90, 95-106<br />
Dixit, Avinash, 1987, Strategic Behavior in Contests, The American Economic<br />
Review, 77, 5, 891-98<br />
Dixit, Avinash <strong>and</strong> Joseph Stiglitz, 1977, Monopolistic <strong>Competition</strong> <strong>and</strong> Optimum<br />
Product Diversity, The American Economic Review, 67, 297-308<br />
Dorfman, Robert <strong>and</strong> Peter Steiner, 1954, Optimal Advertising <strong>and</strong> Optimal<br />
Quality, The American Economic Review, 44, 826-36<br />
Dosi, Giovanni, Alfonso Gambardella, Marco Grazzi <strong>and</strong> Luigi Orsenigo,<br />
2007, Technological Revolutions <strong>and</strong> the Evolution <strong>of</strong> Industrial Structures.<br />
Assessing the impact <strong>of</strong> new technologies upon size, pattern <strong>of</strong><br />
growth <strong>and</strong> boundaries <strong>of</strong> the firms, presented at the 2007 <strong>Intertic</strong> Conference<br />
Eaton, Jonathan <strong>and</strong> Gene Grossman, 1986, Optimal Trade <strong>and</strong> Industrial<br />
Policy under Oligopoly, Quarterly Journal <strong>of</strong> Economics, 101, 386-406<br />
Economides, Nicholas, 1993, Quantity Leadership <strong>and</strong> Social Inefficiency, International<br />
Journal <strong>of</strong> Industrial Organization, 11, 2, 219-37<br />
Economides, Nicholas, 2001, The Micros<strong>of</strong>t Anti-trust Case, Journal <strong>of</strong> Industry,<br />
<strong>Competition</strong> <strong>and</strong> Trade, 1,7-39<br />
Ellison, Sara <strong>and</strong> Glenn Ellison, 2007, Strategic Entry Deterrence <strong>and</strong> Behavior<br />
<strong>of</strong> Pharmaceutical Incumbents Prior to Patent Expiration, NBER<br />
wp 13069<br />
Elzinga, Kenneth, David Evans <strong>and</strong> Albert Nichols, 2002, U.S. v. Micros<strong>of</strong>t<br />
Corp.: Remedy or Malady?, in Micros<strong>of</strong>t, <strong>Antitrust</strong> <strong>and</strong> the New Economy:
8. References 263<br />
Selected Essays, Edited by David Evans, Norwell, MA, Kluwer Academic<br />
Publishers, 127-91<br />
Encaoua, David, 2006, <strong>Antitrust</strong> Policy, the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders<br />
<strong>and</strong> the New Economy: Comments, in Economics <strong>of</strong> Information <strong>and</strong><br />
Communication Technologies, Edited by P. Waelbroeck, Ecole Nationale<br />
Supérieure des Télécommunications, Paris, 244-50<br />
Erkal, Nisvan, 2005, The Decision to Patent, Cumulative <strong>Innovation</strong>, <strong>and</strong><br />
Optimal Policy, International Journal <strong>of</strong> Industrial Organization, 23, 535-<br />
562<br />
Erkal, Nisvan <strong>and</strong> Daniel Piccinin, 2007a, Horizontal Mergers with Free Entry<br />
in Differentiated Oligopolies, mimeo, University <strong>of</strong> Melbourne, presented<br />
at the 2007 <strong>Intertic</strong> Conference<br />
Erkal, Nisvan <strong>and</strong> Daniel Piccinin, 2007b, Cooperative R&D under Uncertainty<br />
with Free Entry, mimeo, University <strong>of</strong> Melbourne, presented at the<br />
2007 <strong>Intertic</strong> Conference<br />
Etro, Federico, 1997, Optimal Fiscal Systems with Income <strong>and</strong> Corporate Tax<br />
Evasion, Dissertation Thesis, U.C.S.C.<br />
Etro, Federico, 1998a, Tax Incidence <strong>and</strong> Ramsey Rules with <strong>Market</strong> Power<br />
<strong>and</strong> Tax Evasion, Rivista di Politica Economica, July, 27-61<br />
Etro, Federico, 1998b, Tax Evasion by Firms, DP 321, University <strong>of</strong> Bologna<br />
Etro, Federico, 1999, A Note on an Inexistent Theorem by Rothschild <strong>and</strong><br />
Stiglitz, mimeo, U.C.L.A.<br />
Etro, Federico, 2000, Does Asymmetric Information Matter in Competitive<br />
Insurance <strong>Market</strong>s?, mimeo, U.C.L.A., published on Rivista Internazionale<br />
di Scienze Sociali, 2004, 112, 1, 27-46<br />
Etro, Federico, 2001, <strong>Innovation</strong> by Leaders II: Schumpeterian Growth with<br />
Endogenous Persistence <strong>of</strong> Monopoly, mimeo, Harvard University, published<br />
on Rivista Internazionale di Scienze Sociali, 2007, 115, 1, 81-108<br />
Etro, Federico, 2002, Stackelberg <strong>Competition</strong> with Free Entry, mimeo, Harvard<br />
University<br />
Etro, Federico, 2004, <strong>Innovation</strong> by Leaders, TheEconomicJournal, 114, 4,<br />
495, 281-303<br />
Etro, Federico, 2005a, The Role <strong>of</strong> IPRs in Promoting <strong>Innovation</strong>, in Intellectual<br />
Property Frontiers Exp<strong>and</strong>ing the Borders <strong>of</strong> Discussion, Editedby<br />
A. Jensen <strong>and</strong> M. Pugatch, Stockholm Network Press, London, 6-7<br />
Etro, Federico, 2005b, The Future for CII Patents in the EU, Managing Intellectual<br />
Property, September, 58-59<br />
Etro, Federico, 2005c, IPRs, Trade Secrets <strong>and</strong> Interoperability: from Coca-<br />
Cola to the New Economy, Stockholm Network Monthly Bulletin, 10(December),<br />
5-7<br />
Etro, Federico, 2006a, Aggressive Leaders, The RAND Journal <strong>of</strong> Economics,<br />
37, 1, 146-54<br />
Etro, Federico, 2006b, <strong>Competition</strong> Policy: Toward a New Approach, European<br />
<strong>Competition</strong> Journal, 2, 1, 29-55
264 8. References<br />
Etro, Federico, 2006c, <strong>Market</strong> Dominance: An Economic Perspective on Art.<br />
82 EU, in <strong>Competition</strong>. Economic Approach to Article 82, EditedbyM.<br />
Stober, European Enterprise Institute Press, Bruxelles, 6-12<br />
Etro, Federico, 2006d, <strong>Antitrust</strong> Policy, the <strong>Theory</strong> <strong>of</strong> <strong>Market</strong> Leaders <strong>and</strong><br />
the New Economy, in Economics <strong>of</strong> Information <strong>and</strong> Communication Technologies,<br />
Edited by P. Waelbroeck, Ecole Nationale Supérieure des Télécommunications,<br />
Paris, 251-285<br />
Etro, Federico, 2006e, <strong>Market</strong> Leaders <strong>and</strong> Industrial Policy, WP 103, University<br />
<strong>of</strong> Milan, Bicocca, Dept. <strong>of</strong> Economics<br />
Etro, Federico, 2006f, Strategic Export Promotion, WP 101, University <strong>of</strong><br />
Milan, Bicocca, Dept. <strong>of</strong> Economics<br />
Etro, Federico, 2007a, Growth Leaders, Journal <strong>of</strong> Macroeconomics, inpress<br />
Etro, Federico, 2007b, The Economics <strong>of</strong> <strong>Competition</strong> Policy <strong>and</strong> Dominant<br />
<strong>Market</strong> Position, The Stockholm Network Experts’ Series on Intellectual<br />
Property <strong>and</strong> <strong>Competition</strong>, Stockholm Network Press, London<br />
Etro, Federico, 2007c, <strong>Innovation</strong> <strong>and</strong> Patents for the S<strong>of</strong>tware Industry: an<br />
Empirical Analysis <strong>of</strong> the EU <strong>and</strong> US Cases, Chapter 4 in S<strong>of</strong>tware Patents:<br />
Protections <strong>and</strong> Licensing, Edited by A.V.N. Rao, ICFAI University Press,<br />
Hyderabad, India, 83-122<br />
Etro, Federico, 2007d, <strong>Competition</strong> Policy for the New Economy, in Abuse <strong>of</strong><br />
Dominance, EditedbyY.Katsoulacos,AthensUniversityPress,Athens,<br />
46-83<br />
Etro, Federico, 2007e, Rather than Democratising <strong>Innovation</strong>, We Should be<br />
Protecting it, Financial Times, July16<br />
Etro, Federico, 2008, Stackelberg <strong>Competition</strong> with Endogenous Entry, The<br />
Economic Journal, forthcoming<br />
European Commission, 2005, DG <strong>Competition</strong> Discussion Paper on the Application<br />
<strong>of</strong> Article 82 <strong>of</strong> the Treaty to Exclusionary Abuses, Brussels, Commission<br />
<strong>of</strong> the European Communities<br />
Evans, David (Editor), 2002, Micros<strong>of</strong>t, <strong>Antitrust</strong> <strong>and</strong> the New Economy:<br />
Selected Essays, Norwell, MA, Kluwer Academic Publishers<br />
Evans, David, 2003a, Some Empirical Aspects <strong>of</strong> Multi-sided Platform Industries,<br />
Review <strong>of</strong> Network Economics, 2, 3, 191-209<br />
Evans, David, 2003b, The <strong>Antitrust</strong> Economics <strong>of</strong> Multi-Sided Platform <strong>Market</strong>s,<br />
Yale Journal <strong>of</strong> Regulation, 20, 325-81<br />
Evans, David, Andrei Hagiu <strong>and</strong> Richard Schmalensee, 2006, Invisible Engines.<br />
How S<strong>of</strong>tware Platforms Drive <strong>Innovation</strong> <strong>and</strong> Transform Industries,<br />
Cambridge, The MIT Press<br />
Evans, David, Albert Nichols <strong>and</strong> Richard Schmalensee, 2005, U.S. v Micros<strong>of</strong>t:<br />
Did Consumers Win?, NBER wp N. 11727<br />
Faini, Riccardo, Jonathan Haskel, Giorgio Barba Navaretti, Carlo Scarpa <strong>and</strong><br />
Christian Wey, 2006, Contrasting Europe’s Decline: Do Product <strong>Market</strong><br />
Reforms Help?, inStructural Reforms Without Prejudices, EditedbyT.<br />
Boeri, M. Castanheira, R. Faini <strong>and</strong> V. Galasso, Oxford University Press
8. References 265<br />
Farrel, Joseph <strong>and</strong> Carl Shapiro, 1990, Horizontal Mergers: An Equilibrium<br />
Analysis, The American Economic Review, 80, 107-25<br />
Fisher, Franklin <strong>and</strong> Daniel Rubinfeld, 2001, United States v. Micros<strong>of</strong>t: An<br />
Economic Analysis, The <strong>Antitrust</strong> Bulletin, 1-69<br />
Foncel, Jerome <strong>and</strong> Marc Ivaldi, 2005, Operating System Prices in the Home<br />
PC <strong>Market</strong>, Journal <strong>of</strong> Industrial Economics, 53, 2, 265-97<br />
Freixas, Xavier <strong>and</strong> Jean-Charles Rochet, 1997, Microeconomics <strong>of</strong> Banking,<br />
Cambridge, The MIT Press<br />
Friedman, James, 1971, A Non-cooperative Equilibrium for Supergames, Review<br />
<strong>of</strong> Economic Studies, 38, 113, 1-12<br />
Fudenberg, Drew, Richard Gilbert, Joseph Stiglitz <strong>and</strong> Jean Tirole, 1983,<br />
Preemption, Leapfrogging <strong>and</strong> <strong>Competition</strong> in Patent Races, European<br />
Economic Review, 22, 1, 3-31<br />
Fudenberg, Drew <strong>and</strong> David Levine, 1998, The <strong>Theory</strong> <strong>of</strong> Learning in Games,<br />
Cambridge, The MIT Press<br />
Fudenberg, Drew <strong>and</strong> Eric Maskin, 1986, The Folk Theorem in Repeated<br />
Games with Discounting or with Incomplete Information, Econometrica,<br />
54, 522-54<br />
Fudenberg, Drew <strong>and</strong> Jean Tirole, 1984, The Fat Cat Effect, the Puppy Dog<br />
Ploy <strong>and</strong> the Lean <strong>and</strong> Hungry Look, The American Economic Review,<br />
Papers <strong>and</strong> Procedeengs, 74, 2, 361-68<br />
Fudenberg, Drew <strong>and</strong> Jean Tirole, 1990, Game <strong>Theory</strong>, Cambridge, The MIT<br />
Press<br />
Fudenberg, Drew <strong>and</strong> Jean Tirole, 2000, Pricing a Network Good to Deter<br />
Entry, Journal<strong>of</strong>IndustrialEconomics, 48, 4, 373-90<br />
Galbraith, John K., 1952, American Capitalism: the Concept <strong>of</strong> Countervailing<br />
Power, Boston, Houghton-Miffin<br />
Gaskins, Darius, 1971, Dynamic Limit Pricing: Optimal Pricing under Threat<br />
<strong>of</strong> Entry, Journal <strong>of</strong> Economic <strong>Theory</strong>, 3, 3, 306-22<br />
Gilbert, Richard <strong>and</strong> David Newbery, 1982, Preemptive Patenting <strong>and</strong> the<br />
Persistence <strong>of</strong> Monopoly, The American Economic Review, 72, 514-26<br />
Gilbert; Richard <strong>and</strong> David Newbery, 1984, Uncertain <strong>Innovation</strong> <strong>and</strong> the<br />
Persistence <strong>of</strong> Monopoly: Comment, The American Economic Review, 74,<br />
1, 238-42<br />
Gilbert, Richard <strong>and</strong> Xavier Vives, 1986, Entry Deterrence <strong>and</strong> the Free Rider<br />
Problem, Review <strong>of</strong> Economic Studies, 53, 1, 71-83<br />
Gilbert, Richard <strong>and</strong> Michael Katz, 2001, An Economist’s Guide to U.S. v.<br />
Micros<strong>of</strong>t, Journal <strong>of</strong> Economic Perspectives, 15, 2, 25-44<br />
Goldfain, Katerina, 2007, Organization <strong>of</strong> R&D with Two Agents <strong>and</strong> a Principal,<br />
mimeo, University <strong>of</strong> Bonn, presented at the 2007 <strong>Intertic</strong> Conference<br />
Goldfain, Katerina <strong>and</strong> Eugen Kováč, 2007, On Compatibility in Two-Sided<br />
<strong>Market</strong>s, mimeo, University <strong>of</strong> Bonn, presented at the 2007 <strong>Intertic</strong> Conference
266 8. References<br />
Goldman, S. <strong>and</strong> Hir<strong>of</strong>umi Uzawa, 1964, A Note on Separability in Dem<strong>and</strong><br />
Analysis, Econometrica, 32, 387-398<br />
Goolsbee, Austan <strong>and</strong> Chad Syverson, 2006, How do Incumbents Respond to<br />
the Threat <strong>of</strong> Entry? Evidence from the Major Airlines, mimeo, University<br />
<strong>of</strong> Chicago<br />
Grieben, Wolf-Heimo, 2004, Schumpeterian Growth <strong>and</strong> the Political Economy<br />
<strong>of</strong> Employment Protection, Journal <strong>of</strong> Economics, 10, 77-118<br />
Griliches, Zvi, 1989, Patents: Recent Trends <strong>and</strong> Puzzles, Brookings Papers<br />
on Economic Activity Microeconomics, 291-330<br />
Griliches, Zvi, 1994, Productivity, R&D, <strong>and</strong> the Data Constraint, The American<br />
Economic Review, 84, 1, 1-23<br />
Hall, Bronwin, Zvi Griliches <strong>and</strong> Jerry Hausman, 1986, Patents <strong>and</strong> R&D: Is<br />
There a Lag?, International Economic Review, 27, 2, 265-83<br />
Hall, Chris <strong>and</strong> Robert Hall, 2000, Toward a Quantification <strong>of</strong> the Effects <strong>of</strong><br />
Micros<strong>of</strong>t’s Conduct, The American Economic Review, Papers <strong>and</strong> Procedeengs,<br />
90, 2, 188-191<br />
Hamilton, Jonathan <strong>and</strong> Steven Slutsky, 1990, Endogenous Timing in Duopoly<br />
Games: Stackelberg or Cournot Equilibria, Games <strong>and</strong> Economic Behavior,<br />
2, 1, 29-46<br />
Harris, Christopher <strong>and</strong> John Vickers, 1985, Patent Races <strong>and</strong> the Persistence<br />
<strong>of</strong> Monopoly, The Journal <strong>of</strong> Industrial Economics, 33, 4, 461-81<br />
Harsanyi, John, 1967-68, Games with Incomplete Information Played by<br />
Bayesian Players, Management Science, 14, 159-82, 320-34 <strong>and</strong> 486-502<br />
Hart, Oliver, 1995, Firms, Contracts, <strong>and</strong> Financial Structure, Clarendon<br />
Press, Oxford<br />
Hausman, Jerry, Bronwin Hall <strong>and</strong> Zvi Griliches, 1984, Econometric Models<br />
for Count Data with an Application to the Patents-R&D Relationship,<br />
Econometrica, 52, 909-38<br />
Helpman Elhanan <strong>and</strong> Paul Krugman, 1989, Trade Policy <strong>and</strong> <strong>Market</strong> Structure,<br />
Cambridge, The MIT Press<br />
Hirshleifer, Jack <strong>and</strong> John Riley, 1992, The Analytics <strong>of</strong> Uncertainty <strong>and</strong><br />
Information, Cambridge, Cambridge University Press<br />
Holmstrom, Bengt <strong>and</strong> Jean Tirole, 1997, Financial Intermediation, Loanable<br />
Funds, <strong>and</strong> the Real Sector, Quarterly Journal <strong>of</strong> Economics, 112, 3, 663-<br />
91<br />
Hotelling, Harold, 1929, Stability in <strong>Competition</strong>, The Economic Journal, 39,<br />
153, 41-57<br />
Howitt, Peter, 2000, Endogenous Growth <strong>and</strong> Cross-Country Income Differences<br />
, The American Economic Review, 90, 4, 829-46<br />
Hughes, Danny, 2007, <strong>Innovation</strong>, Firm Efficiency, <strong>and</strong> <strong>Market</strong> Structure,<br />
mimeo, University <strong>of</strong> South Alabama<br />
ICC, 2006, ICC Comments on the European Commission Discussion Paper<br />
on the Application <strong>of</strong> Article 82 <strong>of</strong> the Treaty to Exclusionary Abuses,<br />
International Chamber <strong>of</strong> Commerce, Paris
8. References 267<br />
ICC, 2007, Single-Firm Conduct as Related to <strong>Competition</strong>, Submission to<br />
the US Federal Trade Commission <strong>and</strong> Department <strong>of</strong> Justice Hearings on<br />
Section 2 <strong>of</strong> the Shearman Act, International Chamber <strong>of</strong> Commerce, Paris<br />
Impullitti, Giammario, 2006a, International <strong>Competition</strong>, Growth, <strong>and</strong> Optimal<br />
R&D Subsidies, mimeo, New York University<br />
Impullitti, Giammario, 2006b, International <strong>Competition</strong> <strong>and</strong> Defensive R&D<br />
Subsidies in Growing Economies, mimeo, New York University<br />
Impullitti, Giammario, 2007, Foreign Technological <strong>Competition</strong> <strong>and</strong> the U.S.<br />
R&D Subsidies Response in the 1970s <strong>and</strong> 1980s: a Quantitative Welfare<br />
Analysis, mimeo, EUI, presented at the 2007 <strong>Intertic</strong> Conference<br />
Johansen, Klas <strong>and</strong> Niklas Damm, 2004, Research <strong>and</strong> Development in Transition.<br />
The Effects <strong>of</strong> <strong>Market</strong> Libaralization, mimeo, Stockholm School <strong>of</strong><br />
Economics<br />
Kahai, Simran, David Kaserman <strong>and</strong> John Mayo, 1996, Is the Dominant Firm<br />
Dominant? An Empirical Analysis <strong>of</strong> AT&T’s <strong>Market</strong> Power, Journal <strong>of</strong><br />
Law <strong>and</strong> Economics, 39, 499-517<br />
Kanevid, Malin, 2007, Compulsory Licensing <strong>of</strong> Intellectual Property Rights,<br />
mimeo, University <strong>of</strong> Lund<br />
Katsoulacos, Yannis, 2006, Exploitative Practices in Article 82, in <strong>Competition</strong>.<br />
Economic Approach to Article 82, EditedbyM.Stober,European<br />
Enterprise Institute Press, Bruxelles, 26-33<br />
Katsoulacos, Yannis <strong>and</strong> David Ulph, 2007, Optimal Design <strong>of</strong> <strong>Competition</strong><br />
Policy Procedures, mimeo, University <strong>of</strong> St. Andrews, presented at the<br />
2007 <strong>Intertic</strong> Conference<br />
Katz, Michael <strong>and</strong> Carl Shapiro, 1985, Network Externalities, <strong>Competition</strong><br />
<strong>and</strong> Compatibility, The American Economic Review, 75, 424-40<br />
Kadiyali, Vrinda, 1996, Entry, Its Deterrence, <strong>and</strong> Its Accommodation: A<br />
Study <strong>of</strong> the U.S. Photographic Film Industry, The RAND Journal <strong>of</strong><br />
Economics, 27, 3, 452-78<br />
Klein, Benjamin, 2001, The Micros<strong>of</strong>t Case: What Can a Dominant Firm Do<br />
to Defend its <strong>Market</strong> Position, Journal <strong>of</strong> Economic Perspectives, 15, 2,<br />
45-62<br />
Kornprobsty, Wolfgang, 2006, Das Romer-Modell mit Qualitatsverbesserndem<br />
Technischen Fortschritt, Regensburger Diskussionsbeitrage zur Wirtschaftswissenschaft,<br />
# 412<br />
Kortum, Samuel, 1993, Equilibrium R&D <strong>and</strong> the Patent-R&D Ratio: U.S.<br />
Evidence, The American Economic Review, Papers & Proceedings, 83, 2,<br />
450-7<br />
Koski, Heli, 2007, Private-Collective S<strong>of</strong>tware Business Models: Coordination<br />
<strong>and</strong> Commercialization via Licensing, Review <strong>of</strong> Economic Research on<br />
Copyright Issues, 4, 1, 47-61, presented at the 2007 <strong>Intertic</strong> Conference<br />
Kotler, Philip, 1999, <strong>Market</strong>ing Management: Analysis, Planning, Implementation,<br />
<strong>and</strong> Control, PrenticeHall
268 8. References<br />
Koulovatianos, Christos, 2005, Oligopolistic Cost <strong>Innovation</strong>, Stock <strong>Market</strong>s,<br />
<strong>and</strong> Macroeconomic Development, mimeo, University <strong>of</strong> Vienna<br />
Kováč, Eugen <strong>and</strong> Krešimir Žigić, 2006, International <strong>Competition</strong> in Vertically<br />
Differentiated <strong>Market</strong>s with <strong>Innovation</strong> <strong>and</strong> Imitation: Impacts <strong>of</strong><br />
Trade Policy , mimeo, <strong>Intertic</strong><br />
Kreps, David <strong>and</strong> Wilson, Robert, 1982, Reputation <strong>and</strong> Imperfect Information,<br />
Journal <strong>of</strong> Economic <strong>Theory</strong>, 7, 253-79<br />
Krugman, Paul, 2000a, Reckonings: Micros<strong>of</strong>t: What’s Next?, The New York<br />
Times, April26<br />
Krugman, Paul, 2000b, The Last Refuge, The New York Times, June 11<br />
Laffont, Jean-Jacques, Patrick Rey <strong>and</strong> Jean Tirole, 1998, Network <strong>Competition</strong>,<br />
The RAND Journal <strong>of</strong> Economics, 29, 1, 1-37<br />
Laffont, Jean-Jacques <strong>and</strong> Jean Tirole, 1993, A <strong>Theory</strong> <strong>of</strong> Incentives in Procurement<br />
<strong>and</strong> Regulation, TheMITPress,Cambridge,MA<br />
Leiner-Killinger, Nadine, Víctor López Pérez, Roger Stiegert <strong>and</strong> Giovanni<br />
Vitale, 2007, Structural Reforms in EMU <strong>and</strong> the Role <strong>of</strong> Monetary Policy,<br />
ECB N. 66<br />
Lambin, Jean-Jacques, 1970, Optimal Allocation <strong>of</strong> Competitive <strong>Market</strong>ing<br />
Efforts: An Empirical Study, The Journal <strong>of</strong> Business, 43, 4, 468-84<br />
Lee, Chang-Yang <strong>and</strong> Taeyoon Sung, 2004, <strong>Innovation</strong> by Leaders: Comments<br />
<strong>and</strong> New Results, mimeo, Korea Advanced Institute <strong>of</strong> Science <strong>and</strong><br />
Technology<br />
Lee, Tom <strong>and</strong> Louis Wilde, 1980, <strong>Market</strong> Structure <strong>and</strong> <strong>Innovation</strong>: a Reformulation,<br />
Quarterly Journal <strong>of</strong> Economics, 3, 395-410<br />
Lerner, Josh, 2001, Did Micros<strong>of</strong>t Deter S<strong>of</strong>tware <strong>Innovation</strong>?, mimeo, Harvard<br />
University<br />
Lerner, Josh <strong>and</strong> Jean Tirole, 2002, Some Simple Economics <strong>of</strong> Open Source,<br />
Journal <strong>of</strong> Industrial Economics, 50, 2, 197-234<br />
Lerner, Josh <strong>and</strong> Jean Tirole, 2006, The Dynamics <strong>of</strong> Opens-Source Contributors,<br />
The American Economic Review, Papers & Proceedings, 96, 2,<br />
114-118<br />
Liebowitz, Stan <strong>and</strong> Stephen Margolis, 1999, Winners, Losers & Micros<strong>of</strong>t:<br />
<strong>Competition</strong> <strong>and</strong> <strong>Antitrust</strong> in High Technology, The Independent Institute<br />
Loury, Glenn, 1979, <strong>Market</strong> Structure <strong>and</strong> <strong>Innovation</strong>, Quarterly Journal <strong>of</strong><br />
Economics, March, 429-36<br />
Maggi, Giovanni, 1996, Strategic Trade Policies with Endogenous Mode <strong>of</strong><br />
<strong>Competition</strong>,The American Economic Review, 86, 1, 237-58<br />
Mailath, George <strong>and</strong> Larry Samuelson, 2006, Repeated Games <strong>and</strong> Reputations.<br />
Long.run Relationships, Oxford,OxfordUniversityPress<br />
Malerba, Franco, <strong>and</strong> Luigi Orsenigo, 1999, Technology Entry, Exit <strong>and</strong> Survival:<br />
an Empirical Analysis <strong>of</strong> Patent Data, Research Policy, 28, 643-60<br />
Malerba, Franco, Luigi Orsenigo <strong>and</strong> Pietro Peretto, 1997, Persistence <strong>of</strong><br />
Innovative Activities, Sectoral Patterns <strong>of</strong> <strong>Innovation</strong> <strong>and</strong> International
8. References 269<br />
Technological Specialization, International Journal <strong>of</strong> Industrial Organization,<br />
15, 801-26<br />
Mankiw, Gregory <strong>and</strong> Michael Whinston, 1986, Free Entry <strong>and</strong> Social Inefficiency,<br />
The RAND Journal <strong>of</strong> Economics, 17, 1, 48-58<br />
Mann, Ronald, 2005, Do Patents Facilitate Financing in the S<strong>of</strong>yware Industry?,<br />
Texas Law Review, 83, 4, 961-1030<br />
Manne, Ge<strong>of</strong>frey <strong>and</strong> Marcellus Williamson, 2005, Hot Docs vs. Cold Economics:<br />
The Use <strong>and</strong> Misuse <strong>of</strong> Business Documents in <strong>Antitrust</strong> Enforcement<br />
<strong>and</strong> Adjudication, Arizona Law Review, 47, 3, 609-59<br />
Marchese, Carla, 2006, Rewarding the Consumer for Curbing the Evasion <strong>of</strong><br />
Commodity Taxes?, WP 79 POLIS<br />
Marshall, Alfred, 1890, Principles <strong>of</strong> Economics, London, Macmillan <strong>and</strong> Co.<br />
Martin, Stephen, 2002, Advanced Industrial Economics, Blackwell Publishers<br />
Inc.<br />
Mas-Colell, Andreu, Michael Whinston <strong>and</strong> Jerry Green, 1995, Microeconomic<br />
<strong>Theory</strong>, Oxford University Press, New York <strong>and</strong> Oxford<br />
Maskin, Eric <strong>and</strong> John Riley, 1984, Monopoly with Incomplete Information,<br />
The RAND Journal <strong>of</strong> Economics, 15, 2, 171-96<br />
Mastrantonio, Giuseppe, 2005, A Story <strong>of</strong> Interfaces at the Intersection Between<br />
Mithology <strong>and</strong> <strong>Antitrust</strong>, mimeo, Luiss University, Rome, presented<br />
at the 2007 <strong>Intertic</strong> Conference<br />
McGee, John, 1958, Predatory Price Cutting: the St<strong>and</strong>ard Oil (N.J.) Case,<br />
Journal <strong>of</strong> Law <strong>and</strong> Economics, 1, 137-69<br />
McFadden, Daniel, 1974, Conditional Logit Analysis <strong>of</strong> Qualitative Choice<br />
Behavior, in Frontiers in Econometrics, Edited by P. Zarembka, Academic<br />
Press, New York, 105—42<br />
McKenzie, Richard, 2001, Trust on Trial: How the Micros<strong>of</strong>t Case is Reframing<br />
the Rules <strong>of</strong> <strong>Competition</strong>, Cambridge, Perseus Publishing<br />
Merges, Robert, 2006, Patents, Entry <strong>and</strong> Growth in the S<strong>of</strong>tware Industry,<br />
mimeo, University <strong>of</strong> California, Berkeley<br />
Milgrom, Paul <strong>and</strong> Roberts, John, 1982, Limit Pricing <strong>and</strong> Entry Under<br />
Incomplete Information: an Equilibrium Analysis, Econometrica, 50, 2,<br />
443-59<br />
Milgrom, Paul <strong>and</strong> Roberts, John, 1992, Economics, Organization <strong>and</strong> Management,<br />
Prentice Hall International Editions<br />
Minniti, Antonio, 2006, Multi-Product Firms, R&D, <strong>and</strong> Growth, The B.E.<br />
Journal <strong>of</strong> Macroeconomics, 6,3(Topics),Art.4<br />
Modigliani, Franco, 1958, New Developments on the Oligopoly Front, Journal<br />
<strong>of</strong> Political Economy, 66, 3 (June), 215-32<br />
Modigliani, Franco <strong>and</strong> Marcus Miller, 1958, The Cost <strong>of</strong> Capital, Corporation<br />
Finance <strong>and</strong> the <strong>Theory</strong> <strong>of</strong> Investment, The American Economic<br />
Review, 3, 261-97<br />
Motta, Massimo, 2004, <strong>Competition</strong> Policy. <strong>Theory</strong> <strong>and</strong> Practice, Cambridge,<br />
Cambridge University Press
270 8. References<br />
Myles, Gareth, 1995, Public Economics, Cambridge University Press<br />
Myerson, Roger, 1991, Game<strong>Theory</strong>.Analysis<strong>of</strong>Conflict, HarvardUniversity<br />
Press<br />
Nash, John, 1950, Equilibrium Points in n-Person Games, Proceedings <strong>of</strong> the<br />
National Academy <strong>of</strong> Sciences, 36, 48-9<br />
Nicholas, Tom, 2003, Why Schumpeter was Right: <strong>Innovation</strong>, <strong>Market</strong> power,<br />
<strong>and</strong> Creative Destruction in 1929s America, Journal <strong>of</strong> Economic History,<br />
63, 4, 1023-58<br />
Novshek, William, 1980, Cournot Equilibrium with Free Entry, Review <strong>of</strong><br />
Economic Studies, 47, 3, April, 473-86<br />
Ogilvie, Sheilagh, 2004a, Guilds, Efficiency, <strong>and</strong> Social Capital: Evidence from<br />
German Proto-Industry, Economic History Review, 57, 2, 286-333<br />
Ogilvie, Sheilagh, 2004b, The Use <strong>and</strong> Abuse <strong>of</strong> Trust: the Deployment <strong>of</strong> Social<br />
Capital by Early modern Guilds, Jahrbuch fur Wirtschaftsgeschichte,<br />
4<br />
Pakes, Ariel <strong>and</strong> Zvi Griliches, 1980, Patents <strong>and</strong> R&D at the Firm Level: A<br />
First Look, Economics Letters, 5, 4, 377-81<br />
Parasc<strong>and</strong>olo, Paola <strong>and</strong> Grazia Sgarra, 2006, Crescita e Produttività: gli<br />
Effetti Economici della Regolazione, Confindustria, Rome<br />
Peretto, Pietro <strong>and</strong> Michelle Connolly, 2005, The Manhattan Metaphor,<br />
mimeo, Duke University<br />
Peretto, Pietro, 2007, Energy Taxes <strong>and</strong> Endogenous Technological Change,<br />
mimeo, Duke University<br />
Porter, Michael, 1985, The Competitive Advantage: Creating <strong>and</strong> Sustaining<br />
Superior Performance, New York, Free Press<br />
Posner, Richard, 2001, <strong>Antitrust</strong> Law, University <strong>of</strong> Chicago Press, Chicago<br />
Prescott, Edward <strong>and</strong> Michael Visscher, 1977, Sequential Location among<br />
Firms with Foresight, The Bell Journal <strong>of</strong> Economics, 8, 2, 378-93<br />
Reinganum, Jennifer, 1982, A Dynamic Game <strong>of</strong> R&D: Patent Protection<br />
<strong>and</strong> Competitive Behaviour, Econometrica, 50, 671-88<br />
Reinganum, Jennifer, 1983, Uncertain <strong>Innovation</strong>s <strong>and</strong> the Persistence <strong>of</strong><br />
Monopoly, The American Economic Review, 73, 741-8<br />
Reinganum, Jennifer, 1985a, <strong>Innovation</strong> <strong>and</strong> Industry Evolution, Quarterly<br />
Journal <strong>of</strong> Economics, February, 81-100<br />
Reinganum, Jennifer , 1985b, A Two-Stage Model <strong>of</strong> Research <strong>and</strong> Development<br />
with Endogenous Second-Mover Advantages, International Journal<br />
<strong>of</strong> Industrial Organization, 3, 275-92.<br />
Reksulak, Michael, William Shughart II <strong>and</strong> Robert Tollison, 2006, <strong>Innovation</strong><br />
<strong>and</strong> the Opportunity Cost <strong>of</strong> Monopoly, mimeo, Clemson University<br />
Rey, Patrick <strong>and</strong> Joseph Stiglitz, 1988, Vertical Restraints <strong>and</strong> Producers’<br />
<strong>Competition</strong>, European Economic Review, 32, 561-68<br />
Rey, Patrick, Jordi Gual, Martin Hellwig, Anne Perrot, Michele Polo, Klaus<br />
Schmidt <strong>and</strong> Rune Stenbacka, 2005, An Economic Analysis to Article 82,<br />
Report <strong>of</strong> the EAGCP, European Commission
8. References 271<br />
Rhee, Ki-Eun, 2006, Reevaluating Merger Guidelines for the New Economy,<br />
KDI wp 06-18, Seoul<br />
Riela, Stefano, 2005, La Politica della Concorrenza e la Competitività, in La<br />
competitività dell’Unione Europea dopo Lisbona, EditedbyF.Borghese,<br />
M. P. Caruso <strong>and</strong> S. Riela, Rubbettino, 41-54<br />
Rochet, Jean-Charles <strong>and</strong> Jean Tirole, 2003, Platform <strong>Competition</strong> in Two-<br />
Sided <strong>Market</strong>s, Journal <strong>of</strong> the European Economic Association, 1, 4, 990-<br />
1029<br />
Rochet, Jean-Charles <strong>and</strong> Jean Tirole, 2006, Two Sided <strong>Market</strong>s: A Progress<br />
Report, The RAND Journal <strong>of</strong> Economics, inpress<br />
Röller, Lars-Hendrick <strong>and</strong> Robin Sickles, 2000, Capacity <strong>and</strong> Product <strong>Market</strong><br />
<strong>Competition</strong>: Measuring <strong>Market</strong> Power in a Puppy-dog Industry, International<br />
Journal <strong>of</strong> Industrial Organization, 18, 6, 845-65<br />
Romer, Paul, 1990, Endogenous Technological Change, Journal <strong>of</strong> Political<br />
Economy, 98, 5, S71-102<br />
Rothschild Michael <strong>and</strong> Joseph Stiglitz, 1976, Equilibrium in Competitive<br />
Insurance <strong>Market</strong>s: An Essay on the Economics <strong>of</strong> Imperfect Information,<br />
Quarterly Journal <strong>of</strong> Economics, 90, 4, 629-49<br />
Saarikoski, Ville, 2006, The Odyssey <strong>of</strong> the Mobile Internet. The emergence<br />
<strong>of</strong> a networking attribute in a multidisciplinary study, Taloustieto Oy,<br />
Helsinki<br />
Salant, Stephen, Sheldon Switzer <strong>and</strong> Robert Reynolds, 1983, Losses from<br />
Horizontal Merger: The Effects <strong>of</strong> an Exogenous Change in Industry Structure<br />
on Cournot-Nash Equilibrium, Quarterly Journal <strong>of</strong> Economics, 98,<br />
2, 185-99<br />
Salant, Stephen, 1984, Preemptive Patenting <strong>and</strong> the Persistence <strong>of</strong> Monopoly:<br />
Comment, The American Economic Review, Papers <strong>and</strong> Procedeengs, 74,<br />
1, 247-51<br />
Schelling, Thomas, 1960, The Strategy <strong>of</strong> Conflict, Harvard University Press,<br />
Cambridge<br />
Scherer, Frederic <strong>and</strong> Huh, Keun, 1992, R & D Reactions to High-Technology<br />
Import <strong>Competition</strong>, Review <strong>of</strong> Economics <strong>and</strong> Statistics, 74, 2, 202-12<br />
Schmalensee, Richard, 1982, Product Differentiation Advantages <strong>of</strong> Pioneering<br />
Br<strong>and</strong>s, The American Economic Review, 72, 349-65<br />
Schmalensee, Richard, 2000, <strong>Antitrust</strong> Issues in Schumpeterian Industries,<br />
The American Economic Review, Papers <strong>and</strong> Procedeengs, 90, 2, 192-96<br />
Schumpeter, Joseph, 1942, Capitalism, Socialism <strong>and</strong> Democracy, Harper&<br />
Row, Publishers, Inc, New York<br />
Scotchmer, Suzanne, 2004, <strong>Innovation</strong> <strong>and</strong> Incentives, TheMITPress,Cambridge<br />
Segerstrom, Paul, 2007, Intel Economics, International Economic Review, 48,<br />
1, 247-80
272 8. References<br />
Segerstrom, Paul, T. Anant <strong>and</strong> Elias Dinopoulos, 1990, A Schumpeterian<br />
Model <strong>of</strong> the Product Life Cycle, The American Economic Review, 80,<br />
1077-92<br />
Selten, Reinhard, 1965, Spieltheoretische Beh<strong>and</strong>lung eines Oligopolmodells<br />
mit Nachfrageträgheit, Zeitschrift für die Gesamte Staatswissenschaft, 121,<br />
301-24 <strong>and</strong> 667-89<br />
Showalter, Dean, 1995, Oligopoly <strong>and</strong> Financial Structure: Comment, The<br />
American Economic Review, 2, 647-53<br />
Showalter, Dean, 1999, Debt as an Entry Deterrent under Bertr<strong>and</strong> Price<br />
<strong>Competition</strong>, Canadian Journal <strong>of</strong> Economics, 32, 4, 1069-81<br />
Shubik, Martin with Richard Levitan, 1980, <strong>Market</strong> Structure <strong>and</strong> Behavior,<br />
Cambridge, Harvard University Press<br />
Shy, Oz, 1995, Industrial Organization. <strong>Theory</strong> <strong>and</strong> Applications, Cambridge,<br />
The MIT Press<br />
Singh, Satwinder, Michael Utton <strong>and</strong> Michael Watterson, 1998, Strategic Behavior<br />
<strong>of</strong> Incumbent Firms in the UK, International Journal <strong>of</strong> Industrial<br />
Organization, 16, 229-51<br />
Spence, Michael, 1974, <strong>Market</strong> Signaling, Cambridge, Harvard University<br />
Press<br />
Spence, Michael, 1976, Product Selection, Fixed Costs <strong>and</strong> Monopolistic<br />
<strong>Competition</strong>, Review <strong>of</strong> Economic Studies, 43, 217-35<br />
Spence, Michael, 1977, Entry, Capacity, Investment <strong>and</strong> Oligopolistic Pricing,<br />
The Bell Journal <strong>of</strong> Economics, 534-44<br />
Spiller, Pablo <strong>and</strong> Edgardo Favero, 1984, The Effects <strong>of</strong> Entry Regulation<br />
on Oligopolistic Interaction: the Uruguayan Banking Sector, The RAND<br />
Journal <strong>of</strong> Economics, 15, 2, 244-54<br />
Stackelberg, Heinrich Von, 1934, Marktform und Gleichgewicht, Vienna <strong>and</strong><br />
Berlin, Springer Verlag<br />
Stigler, George, 1968. The Organization <strong>of</strong> Industry, Homewood, Ill.: Richard<br />
D. Irwin<br />
Stiglitz, Joseph <strong>and</strong> Andrew Weiss, 1981, Credit Rationing in <strong>Market</strong>s with<br />
Imperfect Information, The American Economic Review, 71, 3, 393-410<br />
Sutton, John, 1998, Technology <strong>and</strong> <strong>Market</strong> Structure, Cambridge, The MIT<br />
Press<br />
Sutton, John, 2005, <strong>Market</strong> Structure: The Bounds Approach, in M. Armstrong<br />
<strong>and</strong> R. Porter Eds., H<strong>and</strong>book <strong>of</strong> Industrial Organization, Vol.3,<br />
North Holl<strong>and</strong><br />
Sylos-Labini, Paolo, 1956, Oligopolio e Progresso Tecnico, Giuffrè, Milano<br />
(English Edition: 1962, Cambridge: Harvard University Press)<br />
Tesoriere, Antonio, 2006, Endogenous Timing with Free Entry, CORE DP<br />
93, Louvain, presented at the 2007 <strong>Intertic</strong> Conference<br />
Tirole, Jean, 1988, The <strong>Theory</strong> <strong>of</strong> Industrial Organization, TheMITPress,<br />
Cambridge
8. References 273<br />
Tirole, Jean, 2006, The <strong>Theory</strong> <strong>of</strong> Corporate Finance, Princeton University<br />
Press, Princeton<br />
Thomas, Louis, 1999, Incumbent Firms’ Response to Entry: Price, Advertising,<br />
<strong>and</strong> New Product Introduction, International Journal <strong>of</strong> Industrial<br />
Organization, 17, 527-55<br />
T-Mobile, 2004, T-Mobile (UK)’s response to Ofcom consultation document:<br />
Strategic Review <strong>of</strong> Telecomunications, London<br />
Tullock, Gordon, 1967, The Welfare Costs <strong>of</strong> Tariffs, Monopoly <strong>and</strong> Theft,<br />
Western Economic Journal, 5, 224-32<br />
V<strong>and</strong>ekerckhove, Jan <strong>and</strong> Raymond De Bondt, 2007, Asymmetric Spillovers<br />
<strong>and</strong> Sequential Strategic Investments, Economics <strong>of</strong> <strong>Innovation</strong> <strong>and</strong> New<br />
Technology, in press, presented at the 2007 <strong>Intertic</strong> Conference<br />
Vickers, John, 1986, The Evolution <strong>of</strong> <strong>Market</strong> Structure when there is a<br />
Sequence <strong>of</strong> <strong>Innovation</strong>s, Journal <strong>of</strong> Industrial Economics, 35, 1-12<br />
Vickers, John, 2001, <strong>Competition</strong> Policy <strong>and</strong> <strong>Innovation</strong>, mimeo, Oxford University<br />
Vickrey, William, 1964, Microstatics, Harcourt, Brace <strong>and</strong> World, New York<br />
Viscusi, Kip, Joseph Harrington <strong>and</strong> John Vernon, 2005, Economics <strong>of</strong> Regulation<br />
<strong>and</strong> <strong>Antitrust</strong>, MITPress<br />
Vives, Xavier, 1988, Sequential entry, industry structure <strong>and</strong> welfare, European<br />
Economic Review, 32, 8, 1671-87<br />
Vives, Xavier, 1999, Oligopoly Pricing. Old Ideas <strong>and</strong> New Tools, TheMIT<br />
Press, Cambridge<br />
von Weizsäcker, C.C., 1980, A Welfare Analysis <strong>of</strong> Barriers to Entry, Bell<br />
Journal <strong>of</strong> Economics, 11, 2, 399-420<br />
Wiethaus, Lars, 2006a, Excess Absorptive Capacity <strong>and</strong> the Persistence <strong>of</strong><br />
Monopoly, mimeo, University <strong>of</strong> Hamburg<br />
Wiethaus, Lars, 2006b, Business Strategies Towards the Creation, Absoprption<br />
<strong>and</strong> Dissemination <strong>of</strong> New Technologies, PhD Dissertation, Universität<br />
Hamburg<br />
Whinston, Michael, 1990, Tying, Foreclosure <strong>and</strong> Exclusion, The American<br />
Economic Review, 80, 837-59<br />
Whinston, Michael, 2001, Exclusivity <strong>and</strong> Tying in U.S. v. Micros<strong>of</strong>t: What<br />
We Know, <strong>and</strong> Don’t Know, Journal <strong>of</strong> Economic Perspectives, 15, 2, 63-80<br />
Whinston, Michael, 2006, Lectures on <strong>Antitrust</strong> Economics, TheMITPress,<br />
Cambridge<br />
Williamson, Oliver E., 1968, Economies as an <strong>Antitrust</strong> Defense: The Welfare<br />
Trade-<strong>of</strong>fs, The American Economic Review, 59, 954-59<br />
Zeira, Joseph, 2004, <strong>Innovation</strong>s, Patent Races <strong>and</strong> Endogenous Growth,<br />
mimeo, Harvard University<br />
Žigić, Krešimir, 1998, Intellectual Property Rights Violations <strong>and</strong> Spillovers<br />
in North-South Trade, European Economic Review, 42, 1779-99
274 8. References<br />
Žigić, Krešimir, 2000, Strategic Trade Policy, Intellectual Property Rights<br />
Protection, <strong>and</strong> North-South Trade, Journal <strong>of</strong> Development Economics,<br />
61, 1, 27-60<br />
Žigić, Krešimir, Viatcheslav Vinogradov <strong>and</strong> Eugen Kováč, 2006, Persistence<br />
<strong>of</strong> Monopoly, <strong>Innovation</strong>, <strong>and</strong> R&D Spillovers: Static versus Dynamic<br />
Analysis, mimeo, <strong>Intertic</strong>, presented at the 2007 <strong>Intertic</strong> Conference
Index<br />
Abreu, Dilip, 8<br />
Absorptive capacity, 29<br />
Abuse <strong>of</strong> dominance, 6, 171, 174, 195,<br />
197, 200<br />
Accommodating philosophy, 252<br />
Acemoglu, Daron, 156<br />
Ad valorem tax, 53, 57<br />
Adobe, 210, 233<br />
Adverse selection, 84<br />
Advertising, 63, 70, 78, 249<br />
Aerts, Kris, 159<br />
Aggressive philosophy, 252<br />
Aghion, Philippe, 29, 31, 34, 141, 148,<br />
150, 155, 157, 160, 162, 164<br />
Ahlborn, Christian, 176<br />
Airbus, 66, 120<br />
Aircraft industry, 66, 120<br />
Airline industry, 244, 247<br />
Akerl<strong>of</strong>, George, 84<br />
Allen, Paul, 215<br />
Amazon, 213<br />
America Online, 219<br />
America’s Cup, 251<br />
American Express, 224<br />
Amir, Rabah, 16, 52, 76<br />
Anant, T.C.A., 155<br />
Anderson, Simon, 21, 46, 55, 107, 113,<br />
116, 123, 124<br />
Apple, 131, 209—211, 215, 225, 228, 233<br />
Application Programming Interfaces,<br />
210<br />
Areeda, Phillip, 198<br />
Areeda-Turner rule, 198<br />
Armani, 21<br />
Armstrong, Mark, 78<br />
Arrow’s paradox, 27, 31, 133, 137, 140,<br />
146, 187, 228<br />
Arrow, Kenneth, 28, 31, 133, 218<br />
Article 81 <strong>of</strong> EU Treaty, 172<br />
Article 82 <strong>of</strong> EU Treaty, 172, 195<br />
Asymmetric information, 69, 84, 85,<br />
177<br />
Asymmetries between leaders <strong>and</strong><br />
followers, 109<br />
AT&T, 93, 220, 221<br />
Atari, 215<br />
Aumann, Robert, 8<br />
Automobile industry, 15<br />
Average Avoidable Cost, 200<br />
Average Total Cost, 16, 199<br />
Average Variable Cost, 198, 199, 201<br />
Bain,Joe,3,93,178<br />
Ballmer, Steve, 218<br />
Banking sector, 183, 244<br />
Barriers to entry, 33, 108, 178, 186,<br />
224, 244<br />
Barro, Robert, 29, 155, 157, 223<br />
Baumol, William, 3, 14, 21, 93, 105,<br />
176, 179<br />
Baxter, William, 213<br />
Bayer, 25<br />
Bayesian equilibrium, 2, 69, 84<br />
Beath, John, 133<br />
Becker, Gary, 70<br />
Benetton, 21<br />
Bergman, Mats, 245<br />
Berry, Stephen, 248<br />
Bertr<strong>and</strong> equilibrium, 86
276 Index<br />
Bertr<strong>and</strong>, Joseph, 20, 46, 54, 56, 106,<br />
162<br />
Bessen, Jim, 151, 191<br />
Bilbiie, Florin, 254<br />
BlackBerry, 211, 215<br />
Bloom, Nick, 150<br />
Blundell, Richard, 31, 34, 134, 150, 156,<br />
160<br />
Boeing, 66<br />
Boldrin, Michele, 156, 193, 194<br />
Bonanno, Giacomo, 43, 82—84, 177, 185<br />
Bonus-malus insurance, 85<br />
Boone, Jan, 122<br />
Bork, Robert, 79, 88, 119, 174, 179, 182<br />
Boston Consulting Group, 186<br />
Bowley model, 46<br />
Boycko, Maxim, 123<br />
Br<strong>and</strong>er, James, 43, 72, 74, 120, 177<br />
Brealey, Richard, 72<br />
Bresnahan, Timothy, 219, 248<br />
Brin, Sergey, 238<br />
Bulow, Jeremy, 42, 68, 114, 177<br />
Bundling, 43, 79, 185, 201, 205, 221,<br />
230, 232, 234<br />
Bush, George W., 221<br />
Buxant, Martin, 237<br />
Cable, John, 155<br />
Caillaud, Bernard, 212<br />
Cambini, Carlo, 63<br />
Campari, 237<br />
Capital-labour ratio, 115<br />
Carlton, Dennis, 93<br />
Cartels, 118, 150, 205<br />
Cawley, John, 85<br />
Cellophane fallacy, 200<br />
CES dem<strong>and</strong>, 55, 57, 107, 127<br />
Chamberlin, Edward, 46<br />
Champions League, 251<br />
Ch<strong>and</strong>ler, Alfred, 132, 187<br />
Chanel, 237<br />
Chevalier, Judith, 75, 250<br />
Chiappori, Pierre Andre, 85<br />
Chicago school, 44, 79, 88, 119, 173,<br />
174, 186, 204, 229, 230, 255<br />
Clayton Act, 175<br />
Clinton, Bill, 218<br />
Coase, Ronald, 239<br />
Coca-Cola, 237<br />
Cohen, Wesley, 141<br />
Collusion, 7, 26, 118, 148, 150<br />
<strong>Competition</strong> for the market, 1, 25, 27,<br />
31, 45, 108, 124, 131, 135, 142, 151,<br />
159, 186, 189, 195, 196, 198, 203,<br />
205, 229, 235, 246, 250<br />
<strong>Competition</strong> in prices, 2, 20, 41, 44, 50,<br />
54, 67, 71, 73, 79, 82, 84, 100, 106,<br />
115, 121, 183, 212, 227, 246, 249, 252<br />
<strong>Competition</strong> in quantities, 1, 4, 16, 18,<br />
36, 41, 44, 50, 66, 70, 73, 76, 91, 100,<br />
111, 115, 121, 180, 197, 227, 246, 248<br />
Competitive strategy, 59, 252<br />
Corporate finance, 43, 72<br />
Cost reductions, 66<br />
Cournot duopoly, 5, 111<br />
Cournot, Augustin, 1, 52, 102<br />
Court <strong>of</strong> First Instance, 172, 223<br />
Cowell, Frank, 53<br />
Cozzi, Guido, 154, 194<br />
Credit cards, 78, 224<br />
Cremer, Jacques, 77<br />
Czarnizki, Dirk, 133, 149, 250<br />
D’Aspremont Claude, 46<br />
Darwinian selection effect, 161<br />
Dasgupta, Partha, 133, 136<br />
Davidson, Carl, 87<br />
Davis, Stephen, 231<br />
De Bondt, Raymond, 67, 151<br />
de Palma, Andrè, 21, 46, 55, 107, 116,<br />
123, 124<br />
Debt financing, 72, 150, 250<br />
Deep pocket theory <strong>of</strong> predation, 76<br />
Dell,210,216<br />
Demsetz, Harold, 188<br />
Deneckere, Raymond, 87<br />
Denicolò, Vincenzo, 151, 154, 161, 162,<br />
194<br />
Department <strong>of</strong> Justice, U.S.A., 171,<br />
218, 221
Index 277<br />
Diners Club, 224<br />
Dinopoulos, Elias, 155<br />
Director, Aaron, 174<br />
Discover, 224<br />
Distinct products test for bundling, 202<br />
Dixit, Avinash, 2, 41, 55—57, 60, 91, 93,<br />
127, 177<br />
Dominance, 171, 174, 186, 196, 200<br />
Dominant firm theory, 93<br />
Dorfman, Robert, 71<br />
Dorfman-Steiner condition, 71<br />
Dosi,Giovanni,208<br />
Du Pont, 132, 200<br />
Ducati, 121<br />
Dynamic inefficiency, 159<br />
Eaton, Jonathan, 120, 122<br />
eBay, 213<br />
Economides, Nicholas, 113, 227, 231<br />
Economies <strong>of</strong> scope, 68<br />
Efficiency defense, 196<br />
Ellison, Glenn, 245<br />
Ellison, Sarah, 245<br />
Elzinga, Kenneth, 219<br />
Encaoua, David, 116<br />
Endogenous costs <strong>of</strong> entry, 38<br />
Endogenous entry, 2, 3, 9, 12, 16, 17,<br />
19, 22, 23, 26, 27, 30, 32, 36, 38, 42,<br />
49, 53, 54, 57, 59, 63, 67, 71, 74, 77,<br />
81, 83, 86, 88, 91, 97, 102, 107—109,<br />
119, 121, 134, 138, 144, 146, 150,<br />
178, 199, 224, 228, 232, 236, 243, 252<br />
Endogenous leadership, 113<br />
Engers, Maxim, 113<br />
Entry deterrence, 3, 7, 13, 18, 20, 24,<br />
27, 32, 36, 66, 69, 77, 79, 91, 98, 104,<br />
108, 111, 177, 181, 197<br />
Equity-Debt ratio, 72, 250<br />
Erkal, Nisvan, 87, 104, 116, 117, 150,<br />
151, 194<br />
Erkal-Piccinin model, 88<br />
Escape competition effect, 29, 30, 34,<br />
140, 143, 145, 148, 149, 159<br />
Essential facility, 203<br />
Etro, Federico, 12, 31, 33, 42, 53, 63,<br />
74, 85, 92, 97, 111, 121, 122, 134,<br />
138, 146, 150, 153, 154, 158, 159,<br />
163, 178, 179, 191, 195, 200, 218<br />
European Commission, 172, 195—197,<br />
200, 201, 203, 204, 218, 221, 223,<br />
235, 236, 240<br />
European Court <strong>of</strong> Justice, 172, 223<br />
European Parliament, 191<br />
Evans, David, 176, 202, 207, 209, 212,<br />
217, 219, 223, 228<br />
Excel, 210, 217<br />
Exclusive dealing, 83<br />
Exclusive territories, 83<br />
Export promoting policy, 120<br />
Farrell, Joseph, 6<br />
Fashion industry, 21<br />
Fat cat strategy, 63, 72, 248<br />
Favaro, Edgardo, 244<br />
Federal Trade Commission, U.S.A.,<br />
171, 218<br />
Ferrari, 15<br />
FIAT, 15<br />
Financial predation, 76<br />
Financial structure, 72, 250<br />
Firefox, 192, 233<br />
First degree price discrimination, 84<br />
Fisher, Franklin, 219, 226<br />
Foncel, Jerome, 226, 227<br />
Ford, 15<br />
Formula 1, 251<br />
Franchise fees, 82<br />
Free S<strong>of</strong>tware Movement, 192<br />
Friedman,James,8<br />
Front-loading effect, 161<br />
Fudenberg, Drew, 2, 8, 42, 60, 61, 63,<br />
72, 96, 133, 177, 225<br />
Fudenberg-Tirole taxonomy, 61<br />
Gabszewicz, Jean, 46<br />
Galbraith, John, 132, 187<br />
Galilei, Galileo, 189<br />
Gap, 21<br />
Gates, Bill, 209, 210, 215
278 Index<br />
Geanakoplos, John, 42, 68, 114, 177<br />
General Motors, 15<br />
GeneralPublicLicense,192,239<br />
Ghironi, Fabio, 254<br />
Gilbert, Richard, 111, 133, 140, 148<br />
GlaxoSmithKline, 25<br />
Goldfain, Katerina, 59<br />
Google, 213, 218, 238<br />
Goolsbee, Austan, 244<br />
Green, Jerry, 56<br />
Grieben, Wolf-Heimo, 159<br />
Griffith, Rachel, 29, 31, 34, 134, 148,<br />
150, 156, 160, 162, 164<br />
Griliches, Zvi, 141, 156<br />
Grossman, Gene, 120, 122<br />
Gual, Jordi, 172, 173<br />
Gucci, 21<br />
H&M, 21<br />
Hagiu, Andrei, 209, 217, 228<br />
Hall,Brownin,156<br />
Hall,Chris,227<br />
Hall, Robert, 227<br />
Hamilton, Jonathan, 113<br />
Harley & Davidson, 121<br />
Harrington, Joseph, 93<br />
Harris, Christopher, 133<br />
Harsanyi, John, 2<br />
Hart, Oliver, 76<br />
Hausman, Jerry, 156<br />
Hellwig, Martin, 172, 173<br />
Helpman, Elhanan, 120<br />
Hewlett-Packard, 132, 192, 209, 210,<br />
216<br />
Hirshleifer, Jack, 69<br />
H<strong>of</strong>fmann-La Roche, 25<br />
Holmstrom, Bengt, 76<br />
Homogenous goods, 4, 16, 52, 101<br />
Honda, 121<br />
Horizontal differentiation, 45<br />
Hotelling model, 45, 62<br />
Hotelling, Harold, 45<br />
Howitt, Peter, 141, 150, 155, 157, 160<br />
Hughes, Danny, 188<br />
Hyperbolic dem<strong>and</strong>, 53, 102, 105<br />
IBM, 132, 192, 193, 208, 212, 223, 228<br />
Implications <strong>of</strong> the theory <strong>of</strong> market<br />
leaders for business administration,<br />
252<br />
Impullitti, Giammario, 159<br />
Industrial revolution, 208<br />
Information <strong>and</strong> Communication<br />
Technology, 207<br />
Informative advertising, 70, 79<br />
Insurance market, 84, 85<br />
Intel, 66, 131, 132, 192, 206, 208<br />
Interbr<strong>and</strong> competition, 82, 185, 249<br />
International Chamber <strong>of</strong> Commerce,<br />
195<br />
Internet, 207, 218, 219<br />
Interoperability, 43, 81, 204, 222, 235,<br />
236, 240<br />
Ionascu, Delia, 122<br />
iPhone,131,193,211,215<br />
iPod, 131, 211, 233<br />
IPRs, 121, 151, 164, 187, 189, 192, 194,<br />
203, 204, 223, 235, 236, 238—240, 244<br />
Ivaldi, Mark, 226, 227<br />
Java, 218<br />
Jobs, Steve, 209<br />
Jullien,Bruno,212<br />
Kadiyali, Vrinda, 248<br />
Katsoulacos, Yannis, 133, 172, 186<br />
Katz,Michael,76<br />
Keen, Michael, 53<br />
Keun, Huh, 133, 246<br />
Klein, Benjamin, 219<br />
Klemperer, Paul, 42, 68, 114, 177<br />
Klepper, Steven, 141<br />
Kodak, 132, 248<br />
Kortum, Samuel, 141, 156<br />
Koski, Heli, 192<br />
Kotler,Philip,59,70,252<br />
Koulovatianos, Christos, 159<br />
Kovac, Eugen, 68, 122<br />
Kraft, Kornelius, 133, 149, 250<br />
Kreps, David, 177<br />
Kroes, Neelie, 222
Index 279<br />
Krugman, Paul, 120, 220, 221<br />
Laffont, Jean-Jacques, 62, 69<br />
Lambin, Jean-Jacques, 70<br />
Lambson,Eugen,52<br />
Lazzati, Natalia, 76<br />
Leadership in prices, 23, 106, 107, 115,<br />
122, 183<br />
Leadership in quantities, 10, 12, 17, 19,<br />
100, 102, 111, 115, 121, 180<br />
Lean <strong>and</strong> hungry look, 63, 64, 248<br />
Leapfrogging, 157, 254<br />
Learning by doing, 43, 66<br />
Lee, Tom, 133, 142<br />
Lerner, Josh, 192, 225<br />
Leverage buyouts, 75<br />
Leverage theory <strong>of</strong> tied good sales, 79,<br />
81<br />
Levine,David,156,193,194<br />
Levinsohn, James, 248<br />
Lewis, Tracy, 43, 72, 74, 177<br />
Liberalizations, 123<br />
Licenses, 238<br />
Liebowitz, Stan, 217, 219, 239<br />
Limit pricing, 3, 7, 13, 18, 20, 24, 36,<br />
66, 69, 77, 91, 104, 177, 181, 197, 225<br />
Linn, Joshua, 156<br />
Linux, 192, 210, 216, 224, 225<br />
Logit dem<strong>and</strong>, 21, 54, 57, 107, 116, 127<br />
Long-purse theory <strong>of</strong> predation, 76<br />
Long-run Average Incremental Cost,<br />
201<br />
Loury, Glenn, 59, 133, 136<br />
Maggi, Giovanni, 120<br />
Malerba, Franco, 132, 250<br />
Mankiw,Gregory,103<br />
Mann, Ronald, 191<br />
Margolis, Stephen, 217, 219, 239<br />
<strong>Market</strong>ing, 59, 252<br />
<strong>Market</strong>ing mix 4 P’s model, 59, 249,<br />
252<br />
Marshall equilibrium, 2, 9, 16, 19, 22,<br />
26, 30, 41, 49, 53, 54, 57, 59, 64, 138,<br />
144, 149, 153, 166, 181<br />
Marshall, Alfred, 2<br />
Martin,Stephen,243<br />
Mas-Colell, Andreu, 56<br />
Maskin, Eric, 8, 85, 151, 191<br />
MasterCard, 224<br />
McFadden, Daniel, 21<br />
McGee, John, 175, 199<br />
McKenzie, Richard, 224<br />
Melitz, Marc, 254<br />
Merchant guilds, 134<br />
Merck, 25, 132<br />
Mergers, 6, 43, 87, 171, 205<br />
Merges, Robert, 191<br />
Micros<strong>of</strong>t, 31, 34, 132, 207, 208, 210,<br />
212, 215, 218, 223, 225, 228, 230,<br />
235, 240<br />
Micros<strong>of</strong>t Surface, 193, 229, 230<br />
Micros<strong>of</strong>t vs. EU case, 221<br />
Micros<strong>of</strong>t vs. US case, 218<br />
Milgrom, Paul, 59, 69, 177<br />
Miller, Merton, 72<br />
Minniti, Antonio, 159<br />
Modigliani, Franco, 3, 72, 93<br />
Modigliani-Miller Theorem, 43, 72, 76<br />
Monopolistic competition, 58<br />
Monopoly, 5, 7, 53, 79, 118, 176, 185,<br />
188, 189, 193, 223, 228, 251<br />
Monti, Mario, 221<br />
Moore’s Law, 131<br />
Mosaic, 219<br />
Most-favored-customer clause, 63<br />
Motorola, 132, 210, 211<br />
Motta, Massimo, 87, 171, 231<br />
Mozilla, 192, 233<br />
MP3 players, 131<br />
Mueller, Dennis, 155<br />
Mukherjee, Arijit, 87<br />
Multi-homing, 79, 213, 225, 234<br />
Multi-sided markets, 43, 76, 185, 198,<br />
201, 202, 212, 217, 226<br />
Multimarket competition, 68<br />
Multiple leaders, 110<br />
Multiple strategies, 114<br />
Murphy,Kevin,70,231
280 Index<br />
Myers, Stewart, 72<br />
Myerson, Roger, 2<br />
Myles, Gareth, 53<br />
Nash equilibrium, 1, 8, 16, 19, 22, 25,<br />
29, 41, 48, 52, 54, 57, 59, 61, 138, 143<br />
Nash, John, 1<br />
National Champions, 120<br />
Netscape, 218, 219, 223, 231—233<br />
Network effects, 43, 76, 184, 188, 198,<br />
201, 202, 210, 212, 217, 225, 226,<br />
232, 234, 239<br />
Newbery, David, 140, 148<br />
Nichols, Albert, 219<br />
Nintendo, 215<br />
Nokia, 132, 211<br />
Non-drastic innovations, 148<br />
Nordhaus, WIlliam, 189<br />
Novell, 192, 223, 224<br />
Novshek, William, 2, 52<br />
NTT, 212<br />
Ogilvie, Sheilagh, 134<br />
Open source s<strong>of</strong>tware, 192, 193, 222,<br />
229<br />
Operating System, 192, 209, 210, 213,<br />
215, 218, 221, 223, 225, 228—230, 238<br />
Optimal export subsidy with price<br />
competition, 122<br />
Optimal export subsidy with quantity<br />
competition, 121<br />
Optimal protection <strong>of</strong> IPRs, 189, 190,<br />
194, 196, 203, 235<br />
Oracle, 192, 223<br />
Orsenigo, Luigi, 132, 250<br />
Padilla, Jorge, 176<br />
Page,Larry,238<br />
Pakes, Ariel, 156, 248<br />
Palm OS, 211<br />
Panzar, John, 3, 14, 21, 93, 105, 176<br />
Patent races, 133, 135, 142, 152, 155<br />
Patentability <strong>of</strong> Computer Implemented<br />
Inventions, 191<br />
Patents, 34, 151, 189, 191, 192, 194,<br />
203, 204, 223, 235, 236, 238, 240, 244<br />
PC industry, 207, 208, 210, 225<br />
Peretto, Pietro, 157<br />
Perl<strong>of</strong>f, Jeffrey, 93<br />
Perrot, Anne, 172, 173<br />
Persistence <strong>of</strong> leadership, 27, 29—32, 34,<br />
66, 131, 135, 138, 142—144, 146, 148,<br />
151—153, 155, 157—159, 162, 186, 189,<br />
194, 195, 203, 205, 228, 235, 236,<br />
250, 251<br />
Pfizer, 25, 132<br />
Pharmaceutical sector, 25, 190, 245<br />
Philipson, Tomas, 85<br />
Photographic film industry, 248<br />
Piccinin, Daniel, 87, 104, 116, 117, 150<br />
PlayStation, 210, 215, 218<br />
Poisson process, 136<br />
Political leadership, 254<br />
Polo, Michele, 172, 173<br />
Pooling equilibrium, 69, 84, 85<br />
Porter, Michael, 59, 252<br />
Posner, Richard, 79, 88, 119, 174, 175,<br />
184, 220, 228, 229<br />
post-Chicago approach, 69, 173, 176,<br />
177, 186, 204, 227, 232, 253<br />
PowerPoint, 217<br />
Prantl, Susanne, 160<br />
pre-Chicago approach, 174, 176<br />
Predatory pricing, 7, 69, 175, 177, 183,<br />
197, 205<br />
Prescott, Edward, 113<br />
Price discrimination, 43, 84, 185, 205<br />
Principal-agent models, 59, 254<br />
Privatizations, 123<br />
Product differentiation, 18, 88, 104,<br />
106, 181, 183, 246<br />
Public enterprises, 123<br />
Public production <strong>of</strong> private goods, 123<br />
Public-private partnerships, 190<br />
Puppy dog strategy, 62, 248<br />
Quadratic utility function, 51, 116<br />
Quality-price ratio, 115, 249<br />
Quantity discounts, 82, 84<br />
QWERTY, 240
Index 281<br />
Röller, Lars-Hendrick, 247<br />
R&D Cartels, 26, 151<br />
R&D investment, 25, 27, 31, 66, 131,<br />
135, 142, 151, 159, 186, 189, 194,<br />
203, 228, 235, 249, 250<br />
R&D leadership, 26, 31, 66, 108, 131,<br />
139, 144, 146, 150, 153, 157, 186,<br />
235, 250<br />
R&D subsidies, 26, 121, 159<br />
RAND terms, 222, 238<br />
Ready-to-eat cereal industry, 71, 246<br />
Reagan, Ronald, 176<br />
RealNetworks, 215, 221<br />
Rebates, 82<br />
Red Hat, 192, 224<br />
Refusal to supply, 203<br />
Reinganum, Jennifer, 133, 143, 145, 153<br />
Reksulak, Michael, 190<br />
Rent seeking, 45, 59, 254<br />
Repeated games, 8<br />
Resale price maintenance, 83<br />
Research Joint Ventures, 151<br />
Retailers, 82, 185, 249<br />
Rey, Patrick, 43, 62, 77, 82, 84, 172,<br />
173, 177, 185<br />
Reynolds, Roberts, 87<br />
Rhee, Ki-Eun, 171<br />
Riley, John, 69, 85<br />
Roberts, John, 59, 69, 177<br />
Rochet, Jean-Charles, 78, 212<br />
Rochet-Tirole rule, 78, 201, 214<br />
Rockefeller, 175<br />
Romer, Paul, 155, 220<br />
Rothschild, Michael, 84, 85<br />
Rothschild-Stiglitz model, 84<br />
Rubinfeld, Daniel, 219, 226<br />
Rudholm, Niklas, 245<br />
Rule <strong>of</strong> reason, 175, 198, 201<br />
Sala-i-Martin, Xavier, 155, 157<br />
Salaniè, Bernard, 85<br />
Salant, Stephen, 87, 149<br />
Schelling, Thomas, 2<br />
Scherer, Frederic, 133, 246<br />
Schmalensee, Richard, 72, 207, 209,<br />
217, 226, 228<br />
Schmidt, Klaus, 172, 173<br />
Schmidt, Tobias, 159<br />
Schumpeter, Joseph, 31, 132, 135, 157,<br />
187<br />
Schumpeterian growth, 155, 158, 160,<br />
166<br />
Scotchmer, Suzanne, 151, 156, 159, 188,<br />
239<br />
Second degree price discrimination, 84<br />
Sega, 215<br />
Segerstrom, Paul, 155, 156, 159, 206,<br />
236<br />
Seinfeld, Jerry, 255<br />
Selten, Reinhard, 2<br />
Separating equilibrium, 69, 84, 85<br />
Sequential innovations, 151, 152, 154,<br />
187, 229, 235<br />
Shakespeare, William, 63<br />
Shapiro, Carl, 6, 76, 220<br />
Shapley, Lloyd, 8<br />
Sherman Act, 171<br />
Shleifer, Andrei, 123<br />
Showalter, Dean, 72, 74<br />
Shubik dem<strong>and</strong>, 117<br />
Shubik, Martin, 116<br />
Shughart II, William, 190<br />
Sickles, Robin, 247<br />
Siemens, 211<br />
Signaling, 69<br />
Slutsky, Steven, 113<br />
Smart phones, 131, 211, 212<br />
S<strong>of</strong>tware market, 191, 192, 207, 208,<br />
210, 212, 215, 218, 223, 225, 228,<br />
230, 235, 240<br />
S<strong>of</strong>tware platforms, 212<br />
Sony, 210, 211, 215, 218<br />
Southwest Airlines, 244<br />
Specific tax,53<br />
Spence, Michael, 55, 69, 91<br />
Spencer, Barbara, 120<br />
Spiller, Pablo, 244<br />
SSNIP test, 200
282 Index<br />
Stackelberg equilibrium, 2, 10, 17, 19,<br />
23, 27, 31, 94, 100, 106, 108, 138, 144<br />
Stackelberg equilibrium with endogenous<br />
entry, 3, 12, 17, 19, 21, 23,<br />
27, 32, 36, 38, 91, 94, 97, 102, 105,<br />
107—110, 113, 114, 121, 139, 146, 149,<br />
153, 162, 167, 180, 183, 187, 241,<br />
248, 252<br />
Stackelberg, Heinrich von, 2, 52, 91,<br />
100<br />
Stallman, Richard, 192<br />
St<strong>and</strong>ard Oil Trust, 175<br />
St<strong>and</strong>ards, 238<br />
State aids, 120, 205<br />
Steiner, Peter, 71<br />
Stenbacka, Rune, 172, 173<br />
Sticky entry, 253<br />
Stigler, George, 179<br />
Stiglitz, Joseph, 2, 41, 43, 55—57, 74,<br />
82, 84, 85, 127, 133, 136, 177, 185,<br />
223<br />
Strategic commitments, 23, 41, 42, 59,<br />
61, 63, 118, 120, 123, 150, 184, 232,<br />
249<br />
Strategic complementarity, 42, 47,<br />
49—52, 56, 60—63, 65, 68, 74, 81, 92,<br />
96, 99, 106, 143<br />
Strategic substitutability, 42, 47, 49—51,<br />
59—63, 65, 66, 68, 74, 92, 96, 99, 102,<br />
108<br />
SubGame Perfect Equilibrium, 2, 10,<br />
12, 63, 94, 97, 113<br />
Sun Microsystems, 192, 219, 223<br />
Sunk costs, 38, 179<br />
Supergames, 8<br />
Supermarkets vs retail business, 176<br />
Sutton, John, 38, 66, 179, 209<br />
Switzer, Sheldon, 87<br />
Sylos Labini, Paolo, 3, 93<br />
Symbian, 211, 218<br />
Syverson, Chad, 244<br />
Tax evasion, 53<br />
Tax incidence, 53, 57<br />
Tesoriere, Antonio, 111, 112, 114<br />
Testing the theory <strong>of</strong> market leaders,<br />
243<br />
<strong>Theory</strong> <strong>of</strong> contestable markets, 3, 14,<br />
21, 93, 105, 176<br />
Third degree price discrimination, 85<br />
Thisse, Jean Francois, 21, 46, 55, 107,<br />
116, 123, 124<br />
Thomas, Louis, 243, 246<br />
Tirole, Jean, 2, 42, 60—63, 69, 70, 72,<br />
76—78, 81, 83, 84, 96, 133, 177, 192,<br />
212, 225<br />
Tollison, Robert, 190<br />
Top dog strategy, 62, 64, 79, 248<br />
Torvalds,Linus,192<br />
Toyota, 15<br />
Trade Policy, 120<br />
Trade secrets, 236, 237<br />
Tullock, Gordon, 59<br />
Turner, Donald, 198<br />
Tying, 79, 177, 185, 201, 219, 221, 230<br />
U-shaped cost functions, 15, 16, 103,<br />
181, 198<br />
Ulph, David, 133, 186<br />
US Patent <strong>and</strong> Trademark Office, 191<br />
Valletti, Tommaso, 63<br />
Van Reenen, John, 31, 134<br />
V<strong>and</strong>ekerckhove, Jan, 67, 151<br />
Venture capital financing, 150<br />
Vernon, John, 93<br />
Vertical differentiation, 71, 115<br />
Vertical integration, 82<br />
Vertical restraints, 43, 82, 185, 205, 249<br />
VHS, 240<br />
Vickers, John, 43, 82—84, 133, 163, 177,<br />
185<br />
Vickrey, William, 46<br />
Videogame industry, 210, 215<br />
Vinogradov, Viatcheslav, 68<br />
Visa, 224<br />
Viscusi, Kip, 93<br />
Vishny, Robert, 123<br />
Visscher,Michael,113<br />
Vives, Xavier, 95, 111, 113
Index 283<br />
Vodafone, 212<br />
von Weizsacker, C.C., 2, 17<br />
Webb-Pomerene Act, 120<br />
Weiss, Andrew, 74<br />
Welfare analysis, 14, 104, 108, 126, 127,<br />
141, 148<br />
Whinston, Michael, 43, 56, 79, 81, 83,<br />
103, 177, 185, 232<br />
Wiethaus, Lars, 29<br />
Wikipedia, 193<br />
Wilde, Louis, 133, 142<br />
Williamson, Oliver E., 6<br />
Willig, Robert, 3, 14, 21, 93, 105, 176<br />
Wilson, Robert, 177<br />
Windows, 210, 216, 218, 219, 221, 225,<br />
229, 231<br />
Windows MediaPlayer, 217, 221, 231<br />
Word, 210, 217<br />
World Trade Organization, 120<br />
WorldWideWeb,207,218<br />
Wozniak, Steve, 209<br />
Xbox, 210, 215, 218<br />
Yahoo, 213<br />
Yves Saint Laurent, 21<br />
Zanchettin, Piercarlo, 161, 162<br />
Zara, 21<br />
Zeira, Joseph, 157<br />
Zigic, Kresimir, 67, 68, 122