Hans W. Gottinger, Essays on Technology, Markets and Development
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
ESSAYS ON<br />
TECHNOLOGY,<br />
MARKETS AND<br />
DEVELOPMENT
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
ESSAYS ON<br />
TECHNOLOGY,<br />
MARKETS AND<br />
DEVELOPMENT<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich, Germany<br />
www.stratec-c<strong>on</strong>.net<br />
stratec_c@yahoo.com<br />
gottingerhans@gmail.com
Aggregate Keywords (10): Investment, Technological <strong>Development</strong>, Stochastic Models,<br />
Network Ec<strong>on</strong>omics, Supply Chain Competiti<strong>on</strong>/Coopetiti<strong>on</strong>, Strategic Alliances,<br />
Microec<strong>on</strong>omics of Ec<strong>on</strong>omic Growth, Ec<strong>on</strong>omic Growth <strong>and</strong> Catching-Up, <strong>Development</strong> Paths,<br />
Ec<strong>on</strong>omic Growth under Envir<strong>on</strong>mental C<strong>on</strong>straints
CONTENTS<br />
HANS W. GOTTINGER<br />
CONTENTS<br />
ESSAYS ON TECHNOLOGY, MARKETS<br />
AND DEVELOPMENT<br />
C<strong>on</strong>tents ................................................................................................ I<br />
Foreword. .............................................................................................. III<br />
Preface <strong>and</strong> Introducti<strong>on</strong> ............................................................................... V<br />
1. Innovati<strong>on</strong> <strong>and</strong> <strong>Technology</strong> ............................................................ 1<br />
1.1 Stochastics of Innovati<strong>on</strong> Processes .............................................................. 3<br />
1.2 Stochastic Models for Capital Growth ........................................................... 17<br />
1.3 Modeling stochastic innovati<strong>on</strong> races ........................................................... 33<br />
1.4 Global technological races. ..................................................................... 53<br />
1.5 Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets .......................................... 67<br />
1.6 High speed technology competiti<strong>on</strong>. ............................................................ 85<br />
1.7 Sequential <strong>Technology</strong> Choice <strong>and</strong> R&D Racing. ................................................ 103<br />
1.8 Internet of Things Ec<strong>on</strong>omic <strong>and</strong> Industrial Dimensi<strong>on</strong>s ......................................... 123<br />
2. <strong>Markets</strong> <strong>and</strong> Competiti<strong>on</strong> ........................................................... 133<br />
2.1 Innovati<strong>on</strong>, Dynamics of Competiti<strong>on</strong> <strong>and</strong> Market Dynamics ................................... 135<br />
2.2 Network Ec<strong>on</strong>omics for Open Source Technologies ............................................. 151<br />
2.3 Vertical Competiti<strong>on</strong> <strong>and</strong> Outsourcing ......................................................... 169<br />
2.4 Supply-Chain Coopetiti<strong>on</strong> ..................................................................... 177<br />
2.5 A tale <strong>on</strong> Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> ................................................. 183<br />
2.6 Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> .......................................................... 189<br />
2.7 Competitive Positi<strong>on</strong>ing through Strategic Alliances ............................................ 199<br />
3. Growth <strong>and</strong> <strong>Development</strong> ........................................................... 217<br />
3.1 Microec<strong>on</strong>omic Foundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth (Review) ..................................... 219<br />
3.2 Ec<strong>on</strong>omic Growth, Catching-Up, Falling Behind <strong>and</strong> Getting Ahead ............................. 301<br />
3.3 Catch-Up <strong>and</strong> C<strong>on</strong>vergence. ................................................................... 315<br />
3.4 Escaping from <strong>Development</strong> Traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345<br />
3.5 Some Applicati<strong>on</strong>s of a Result in C<strong>on</strong>trol Theory to Ec<strong>on</strong>omic Planning Models .................. 355<br />
3.6 Optimal Ec<strong>on</strong>omic Growth when CO2 C<strong>on</strong>straints are Critical ................................... 377<br />
3.7 Ec<strong>on</strong>omic Choice <strong>and</strong> <strong>Technology</strong> Diffusi<strong>on</strong> in New Product <strong>Markets</strong> ............................ 385<br />
I
HANS W. GOTTINGER<br />
CONTENTS<br />
II
SOURCES OF ARTICLES CORRESPONDING TO SECTIONS<br />
HANS W. GOTTINGER<br />
Sources of articles corresp<strong>on</strong>ding to Secti<strong>on</strong>s 1.1 – 3.7<br />
1 <strong>Technology</strong><br />
1.1 Journal of Ec<strong>on</strong>omics 49(2), 1989, 123-138<br />
1.2 Journal of Ec<strong>on</strong>omics 54(3), 1991, 267-281<br />
1.3 Technological Forecasting & Social Change 69, 2002, 607-624<br />
1.4 Japan <strong>and</strong> the World Ec<strong>on</strong>omy 18, 2006, 181-193<br />
1.5 Int. J. <strong>Technology</strong>, Policy <strong>and</strong> Management 4(3), 2004, 240-256<br />
1.6 Int. J. Business <strong>and</strong> Systems Research 10(1), 85-102<br />
1.7 Journal of Policy Studies (Japan) 15(9), 2003, 43-62<br />
1.8 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 6(5), 2017, 115-123<br />
2 <strong>Markets</strong><br />
2.1 Archives of Bus.Research 4(1), 2015, 1-16<br />
2.2 Adv. Soc.Sci. Research 4(7), 2017, 164-181<br />
2.3 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 4(6), 2015, 315-322<br />
2.4 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 4(2), 2015, 67-71<br />
2.5 Research Memo. Abstract, STRATEC Munich , 2012<br />
2.6 Int. J. Bus. <strong>and</strong> Ec<strong>on</strong>omics Research 2(2), 2013, 22-32<br />
2.7 Int. J. Revenue Management 8(1), 2008, 1-17<br />
3 Growth <strong>and</strong> <strong>Development</strong><br />
3.1 Unpublished Supplement to: <str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Mattheus F.A. Goosen , eds<br />
Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up, NovaScience, New York 2012<br />
3.2 Journal of Policy Studies (Japan) 21(11), 2005, 1-14<br />
3.3 J.Appl.Ec<strong>on</strong>omics <strong>and</strong> Business 2(5), 2014, 67-95<br />
3.4 (with C. Umali), J.Appl.Ec<strong>on</strong>omics <strong>and</strong> Business 3(1), 2015, 5-13<br />
3.5 Weltwirtschaftliches Archiv (Review of Ec<strong>on</strong>omics) 111. 1975, 728-751<br />
3.6 Energy Ec<strong>on</strong>omics, July 1992, 192-199<br />
3.7 Weltwirtschaftliches Archiv (Review of Ec<strong>on</strong>omics) 123, 1987, 93-120<br />
III
HANS W. GOTTINGER<br />
SOURCES OF ARTICLES CORRESPONDING TO SECTIONS<br />
IV
FOREWORD<br />
HANS W. GOTTINGER<br />
FOREWORD<br />
Yesterday (or was it Thursday) Apple was crowned as the first “trilli<strong>on</strong> dollar company”, willing a supposed<br />
race with Alphabet (Google), Amaz<strong>on</strong> <strong>and</strong> Microsoft. Many guru hours am<strong>on</strong>g financial journalists have<br />
been, <strong>and</strong> more will be spent <strong>on</strong> recounting the high (<strong>and</strong> a few low) points of Apple’s rise. This book has<br />
little to say <strong>on</strong> the dynamics of that particular competiti<strong>on</strong>, except perhaps to note that insights from the<br />
theory of games may play a role in the story.<br />
Speaking for myself, I find it more fascinating that Apple’s path to success, is littered with the corpses<br />
or dying remains of former corporate super-stars. Why, <strong>on</strong>e w<strong>on</strong>ders, did the inventors <strong>and</strong> innovators<br />
behind the key inventi<strong>on</strong>s <strong>and</strong> innovati<strong>on</strong>s, gain so little of the wealth they created? For that matter, why<br />
did formerly dominant tech firms, such as GE, IBM, Xerox, Kodak <strong>and</strong> HP, lose their way <strong>on</strong> the yellow<br />
brick road? Why did the current crop of winners get so powerful <strong>and</strong> why did the losers lose out? Moreover,<br />
can we expect Amaz<strong>on</strong> <strong>and</strong> Alphabet to fall by the wayside in their turn?<br />
These questi<strong>on</strong>s are of interest to inventors <strong>and</strong> venture capitalists, of course. They are also of interest<br />
to l<strong>on</strong>g-term instituti<strong>on</strong>al investors, such as pensi<strong>on</strong> funds, endowment funds <strong>and</strong> insurance companies.<br />
At another level, they are potentially important to our underst<strong>and</strong>ing of the ec<strong>on</strong>omic system as a whole.<br />
Ec<strong>on</strong>omists have l<strong>on</strong>g been at a loss to underst<strong>and</strong> the underlying drivers of booms <strong>and</strong> busts, al<strong>on</strong>g with<br />
their accompanying changes in wealth distributi<strong>on</strong> <strong>and</strong> c<strong>on</strong>sequences for political power.<br />
Ec<strong>on</strong>omic theory relies heavily <strong>on</strong> linear models. But the “boom-bust” pattern is evidently n<strong>on</strong>-linear.<br />
Let me indulge in a bit of theorizing of my own. I postulate that both booms <strong>and</strong> busts are likely to reflect<br />
two sides of the same phemomen<strong>on</strong>, the phenomen<strong>on</strong> of “lock-in/lock-out”. What this means is that a first<br />
choice am<strong>on</strong>g competing bidders or technologies or business models can become effectively irreversible<br />
<strong>on</strong>ce initial advantages (e.g. learning or experience) kick in. For example, the “experience curve” is a<br />
particularly important comp<strong>on</strong>ent of increasing productivity Returns to scale create (temporary) advantages<br />
for incumbents (booms) <strong>and</strong>, c<strong>on</strong>sequently, disadvantages to challengers. This phenomen<strong>on</strong> can (<strong>and</strong> does)<br />
often result in sub-optimal (or even bad) l<strong>on</strong>g-term outcomes (busts).<br />
The obvious example today is the fact that our whole industrial system (<strong>and</strong> urban development, in particular)<br />
is “locked in” to dependence <strong>on</strong> a road-based transport system requiring private automobiles (<strong>and</strong> trucks)<br />
using liquid hydrocarb<strong>on</strong> fuels. Electric <strong>and</strong> steam-powered vehicles were still competitive in 1900, but they<br />
had a short range, due to storage limitati<strong>on</strong>s. Steam power overcame that limit by introducing a c<strong>on</strong>denser<br />
in the 1920s (with the Doble), but it was much too late to compete. Electrificati<strong>on</strong> of cars <strong>on</strong>ly began again<br />
after the year 2000, thanks to the inventi<strong>on</strong> of lithium-i<strong>on</strong> batteries. Range <strong>and</strong> battery re-charge rate are<br />
still limitati<strong>on</strong>s, but performance in those domains is improving fast. What if the lithium-i<strong>on</strong> battery had<br />
been invented a century earlier?<br />
The “lock-in” of uranium-based nuclear power technology (<strong>and</strong> the “lock-out” of a potentially much superior<br />
thorium-based technology) is another example. It seems that uranium (specifically the isotope U-235) became<br />
the fuel of choice precisely because of the military uses of its by-products, neptunium <strong>and</strong> plut<strong>on</strong>ium. Now,<br />
in stati<strong>on</strong>ary power applicati<strong>on</strong>s plut<strong>on</strong>ium is an unwanted (except by terrorists) pollutant, not a valuable<br />
by-product. Yet uranium technology still gets all the investment.<br />
Positive returns-to-scale create a short-term advantage – hence a barrier to challenge by competitors –<br />
resulting from positive returns to scale. The assumpti<strong>on</strong> of c<strong>on</strong>stant or declining returns to scale is wr<strong>on</strong>g<br />
V
HANS W. GOTTINGER<br />
FOREWORD<br />
because “network” service industries (like the teleph<strong>on</strong>e system) dem<strong>on</strong>strably exhibit increasing returns<br />
to scale. (See pp 151-160 of this book). The simple explanati<strong>on</strong> for positive returns in network sectors is that<br />
the larger the network the greater the number of people c<strong>on</strong>nected to each node, resulting in increased value<br />
to each customer. This phenomen<strong>on</strong>, in more general terms, is also resp<strong>on</strong>sible for populati<strong>on</strong> clustering<br />
(cities), the success of department stores, browsers, the success of search engines (like Google) <strong>and</strong> a variety<br />
of other ec<strong>on</strong>omically important activities. All of these phenomena are impossible to explain in terms of<br />
ec<strong>on</strong>omic theories based <strong>on</strong> c<strong>on</strong>stant or decreasing returns to scale.<br />
Positive feedbacks in a deterministic system do not necessarily lead the system to a unique equilibrium<br />
state. The dynamic equati<strong>on</strong>s for the ec<strong>on</strong>omy in disequilibrium are n<strong>on</strong>-linear, <strong>and</strong> for n<strong>on</strong>-linear systems<br />
there can be multiple alternative equilibria. At a bifurcati<strong>on</strong> point the system then “chooses” <strong>on</strong>e path,<br />
often for short-term reas<strong>on</strong>s. There is no guarantee whatsoever that the path adopted at a bifurcati<strong>on</strong> point<br />
leads to the “best” l<strong>on</strong>g-term soluti<strong>on</strong>. Indeed, as, the path selected is often sub-optimal, <strong>and</strong> may lead<br />
the ec<strong>on</strong>omy “over a cliff” if there is no scope for remedial acti<strong>on</strong>. It is a priori unlikely that the optimum<br />
short-term choice is also the optimum l<strong>on</strong>g-term choice.<br />
This book will, hopefully, provide some clues to these mysteries.<br />
Robert Ayres, Novartis Professor of Ec<strong>on</strong>omics <strong>and</strong> <strong>Technology</strong>, INSEAD, Paris<br />
VI
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
<str<strong>on</strong>g>Essays</str<strong>on</strong>g> <strong>on</strong> Decisi<strong>on</strong>s, <strong>Technology</strong>, <strong>Markets</strong> <strong>and</strong> <strong>Development</strong><br />
PREFACE AND INTRODUCTION<br />
This collecti<strong>on</strong> of essays highlights the interface topics of innovati<strong>on</strong> <strong>and</strong> technology, markets <strong>and</strong> competiti<strong>on</strong>,<br />
growth <strong>and</strong> development in an evoluti<strong>on</strong>ary dynamic ec<strong>on</strong>omy of a dominantely capitalistic market ec<strong>on</strong>omy<br />
in a globalized c<strong>on</strong>text. They span a time frame from the 1980s to current developments.<br />
Previews<br />
In Part 1 the core of Secs. 1.1. to 1.8 emphasize the linkage between high speed technological competiti<strong>on</strong><br />
through technology races in firms, industries <strong>and</strong> nati<strong>on</strong>al ec<strong>on</strong>omies. We start with observati<strong>on</strong>s how<br />
stochastic innovati<strong>on</strong> processes emerge in a firm <strong>and</strong> are carried through by capital investments <strong>and</strong><br />
how the capital allocati<strong>on</strong> activity is governed by a stochastic process c<strong>on</strong>strained by optimal stopping<br />
of productivity enhancing investments (Secs.1.1, 1.2). Further <strong>on</strong>, the remaining Secs. 1.3 to 1.7 rely <strong>on</strong><br />
(stochastic) differential game analysis in symmetric <strong>and</strong> asymmetric rivalrous situati<strong>on</strong>s am<strong>on</strong>g firms that<br />
extends the modeling of R&D based competiti<strong>on</strong> bey<strong>on</strong>d c<strong>on</strong>venti<strong>on</strong>al models found in the ec<strong>on</strong>omics <strong>and</strong><br />
technology literature. Many high technology industries are characterized by positive network externalities.<br />
Firms essentially compete <strong>and</strong> cooperate <strong>on</strong> R&D <strong>and</strong> the producti<strong>on</strong> of goods <strong>and</strong> services that share a<br />
network. Some models of competiti<strong>on</strong> c<strong>on</strong>tain special features that apply equally well to network markets.<br />
One is the uncertainty in technological development or uncertainty in the realizati<strong>on</strong> of a firm’s R&D effort.<br />
The other is the dynamic nature of price competiti<strong>on</strong> between firms in the presence of network effects.<br />
Firms compete with each other over an extended period of time <strong>and</strong> must therefore strategically choose<br />
prices as the market shares of the firms evolve.<br />
Most existing models of network markets focus <strong>on</strong> <strong>on</strong>ly <strong>on</strong>e of the two features. Further, almost all the<br />
models have the comm<strong>on</strong>ality that <strong>on</strong>e firm captures the market instantaneously <strong>and</strong> sells to all c<strong>on</strong>sumers<br />
from then <strong>on</strong> (the “winner-takes-all” market) . However, the history of high technology industry abounds<br />
with instances where rival firms (<strong>and</strong> technologies) have had extended battles for providing the industry<br />
st<strong>and</strong>ard.<br />
In network markets evoluti<strong>on</strong> of market share is a very interesting phenomen<strong>on</strong> especially in the face of<br />
uncertainty about future product quality. Sec.1.5 attempts to provide a framework that could capture the<br />
richness of market share evoluti<strong>on</strong> in the presence of network externalities. In a specific situati<strong>on</strong> of a high<br />
-tech market if you are not the indisputable leader you have <strong>on</strong>ly little hope to achieve leadership unless<br />
you come up with a major innovati<strong>on</strong> replacing the leading firm. Incremental innovati<strong>on</strong> – making slight<br />
improvements in the leaders’ products – will not enable a small firm to overtake a leader that enjoys the<br />
benefits of network ec<strong>on</strong>omies. Therefore, it is not atypical for a fringe firm that invests heavily to displace<br />
the leader by leapfrogging the leader’s technology.<br />
We are encountering the problem of tradeoff between “network dominance” <strong>and</strong> “radical innovati<strong>on</strong>”<br />
that could tip the market the other way, with a significant caveat added that breakthrough R&D is highly<br />
uncertain. From a strategic perspective, in this envir<strong>on</strong>ment, for any two firms of asymmetric size, both<br />
compete dynamically over prices to win market share. In this dynamic process there are two ways to achieve<br />
(temporary) m<strong>on</strong>opolistic status. The ‘smaller’ firm can use dynamic pricing competiti<strong>on</strong> to delay the time<br />
in which the ‘larger’ firm wins a critical market share in the hope to hit the innovati<strong>on</strong> first <strong>and</strong> displace<br />
it. If the innovator is “patient”, the probability of innovati<strong>on</strong> <strong>and</strong> the discount factor are sufficiently high,<br />
there is an equilibrium in which duopoly persists (no firm achieves a critical market share) until <strong>on</strong>e of the<br />
two firms wins the race for innovati<strong>on</strong>.<br />
In Sec.1.7 we model the R&D process as <strong>on</strong>e that is composed of a series of stages, at each of which some<br />
set of product characteristics are determined. At each stage, the firm exercises discreti<strong>on</strong> as to the extent<br />
VII
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
to which the particular stage characteristics will be developed. Further, when the firm’s R&D resource<br />
endowment is fixed <strong>and</strong> stage-specific, the firm faces a tradeoff between the extent to which it will develop<br />
the stage characteristics <strong>and</strong> the time that it will spend in doing so. We will refer to the degree to which<br />
the stage characteristic is developed as the characteristic’s (targeted) level of strategic positi<strong>on</strong>ing. In this<br />
sense, the more “forward-looking” the product characteristics, the greater the profits the firm can expect<br />
to get from the product but the l<strong>on</strong>ger the expected time to completi<strong>on</strong> of R&D. We will speak of this as<br />
the speed–leadership tradeoff in each stage. When the completi<strong>on</strong> of R&D requires many such tradeoffs to<br />
be resolved sequentially, firms will, at the beginning of each stage, want to make their choices in light of<br />
the choices made by all firms in previous stages <strong>and</strong> their stage completi<strong>on</strong> records to that time. To carry<br />
technological innovati<strong>on</strong> through networks across firm/industry boundaries creates a new paradigm of<br />
network centered innovati<strong>on</strong> that integrate through intelligent software tools as outlined <strong>and</strong> reviewed <strong>on</strong><br />
the present state of affairs in Sec. 1.8.<br />
In highly competitive technological industries new challenges <strong>and</strong> opportunities are arising in the new<br />
product development arena. Driven by global markets, global competiti<strong>on</strong>, the global dispersi<strong>on</strong> of scientific/<br />
engineering talent, <strong>and</strong> the advent of new informati<strong>on</strong> <strong>and</strong> communicati<strong>on</strong> technologies (ICT) a new visi<strong>on</strong><br />
of product development is that of a highly disaggregated , distributive process with people <strong>and</strong> organizati<strong>on</strong>s<br />
spread throughout the world. At the same time products are becoming increasingly complex requiring<br />
numerous engineering decisi<strong>on</strong>s to bring them to market. Competitive pressures mean that “time to market”<br />
has become a key to new product success. However, at the same time it is important to keep the innovati<strong>on</strong><br />
<strong>and</strong> quality dimensi<strong>on</strong>s of the new product at their optimal level.<br />
A central questi<strong>on</strong> to address is how should firms invest in innovati<strong>on</strong> <strong>and</strong> what are the implicati<strong>on</strong>s of such<br />
investments for competitive advantage? Underst<strong>and</strong>ing why a firm benefits from investments in innovati<strong>on</strong><br />
<strong>and</strong> quality illuminates issues of competitive strategy <strong>and</strong> industrial organizati<strong>on</strong>. In the field of competitive<br />
strategy, much attenti<strong>on</strong> has been devoted to the c<strong>on</strong>cept of core capabilities. Underst<strong>and</strong>ing how firms make<br />
optimal investments in the face of competiti<strong>on</strong> reveals the nature of competiti<strong>on</strong> <strong>and</strong> provides theoretical<br />
<strong>and</strong> managerial implicati<strong>on</strong>s for developing core competence <strong>and</strong> dynamic capabilities.<br />
In major parts of competitive analysis involving R&D decisi<strong>on</strong>s the focus is <strong>on</strong> breakthrough innovati<strong>on</strong>s<br />
which could create entirely new markets, for example, in studies featuring patent racing between competing<br />
firms. In more comm<strong>on</strong> competitive situati<strong>on</strong>s we observe firms, however, competing by investing in<br />
incremental improvements of products. It is an important aspect when innovati<strong>on</strong> is c<strong>on</strong>sidered to be<br />
manifested in product quality, process improvements <strong>and</strong> in the overall quality culture of an organizati<strong>on</strong><br />
. For example, after product launch, incremental improvement of different aspects of product quality,<br />
improvements in various business processes <strong>and</strong> an incremental adopti<strong>on</strong> of a quality culture are quite realworld<br />
phenomena. Thus innovati<strong>on</strong> is c<strong>on</strong>ceptualized as being manifested in terms of quality. Moreover,<br />
the quality dimensi<strong>on</strong> extends to pursue c<strong>on</strong>tinuous total quality improvement for the entire product life<br />
cycle. Some firms operate in a simultaneous product launch situati<strong>on</strong> while others compete sequentially<br />
by adopting the role of leader or follower. The strategic implicati<strong>on</strong>s in these diverse circumstances can be<br />
treated within a unified framework of dynamic stochastic differential games.<br />
Part 2 , Competiti<strong>on</strong> <strong>and</strong> <strong>Markets</strong>, is devoted to look into the interacti<strong>on</strong> of innovati<strong>on</strong> culture, competitive<br />
dynamics <strong>and</strong> shaping of markets with c<strong>on</strong>tinuous emergence <strong>and</strong> proliferati<strong>on</strong> of open source technologies<br />
dominating in ICTs (Secs. 2.1, 2.2). In Sec. 2.2 we focus <strong>on</strong> a c<strong>on</strong>testable market with network externalities<br />
engaging an incumbent <strong>and</strong> an entrant. The incumbent, unlike the entrant, already has an installed base<br />
of c<strong>on</strong>sumers. We look at decisi<strong>on</strong> situati<strong>on</strong>s of firms regarding how proprietary they want to make their<br />
technology , either through patent protecti<strong>on</strong> or through development in open source systems (OSS).<br />
We explicitly model the direct <strong>and</strong> indirect effects of network externalities. For example, more software<br />
companies are willing to produce programs for an operating system (OS) if it has a larger c<strong>on</strong>sumer base.<br />
This increased competiti<strong>on</strong> could lead to an improvement of the quality of the OS. The model predicts that<br />
using open source technologies is likely to enhance the rate of R&D , <strong>and</strong> c<strong>on</strong>sequently the quality of the<br />
product.. An incumbent that would choose this strategy is likely to deter entrance of a newcomer because<br />
it can play out its advantage of a larger network.<br />
Competitive <strong>and</strong> cooperative forces, termed “coopetiti<strong>on</strong>”, are likely to synchr<strong>on</strong>ize in extending to supply<br />
VIII
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
chain networks (Secs. 2.3, 2.4) . This opens the door for strategic cooperati<strong>on</strong> as focussed <strong>on</strong> a timely case<br />
of a Chinese industry study (Secs. 2.5, 2.6). The degree of competitive positi<strong>on</strong>ing in view of forming<br />
strategic alliances (Sec. 2.7) is explored. This preview c<strong>on</strong>cludes by emphasizing the analytical , strategic<br />
<strong>and</strong> practical implicati<strong>on</strong>s <strong>and</strong> providing directi<strong>on</strong>s for future industry analysis.<br />
In recent years with the emergence of e-business <strong>and</strong> a supply chain view for product development processes,<br />
multiple firms with varying <strong>and</strong> at times c<strong>on</strong>flicting objectives enter into collaborative arrangements. In<br />
such situati<strong>on</strong>s, the competitive strategy based <strong>on</strong> quality <strong>and</strong> innovati<strong>on</strong> could potentially permeate in<br />
those collaborative setups. A recent example is the collaborative venture of S<strong>on</strong>y <strong>and</strong> Samsung to build<br />
a cutting edge plant for LCD flat screen TV though displaying <strong>on</strong>going fierce rivalry in new product<br />
launches in exactly the same product categories. When innovati<strong>on</strong> <strong>and</strong> quality levels form the core of a<br />
firm’s capabilities, each member in the supply chain would have an incentive to invest <strong>and</strong> improve their<br />
dynamic capabilities. This leads to tacit competiti<strong>on</strong> am<strong>on</strong>g collaborative product development partners<br />
by means of active investment in innovati<strong>on</strong> <strong>and</strong> quality.<br />
Although the forces of innovati<strong>on</strong> are central to competiti<strong>on</strong> in young, technically dynamic industries, they<br />
also affect mature industries where life cycles historically were relatively str<strong>on</strong>g, technologies mature, <strong>and</strong><br />
dem<strong>and</strong>s stable.<br />
A strategy for technology must c<strong>on</strong>fr<strong>on</strong>t primarily what the focus of technical development will be. The<br />
questi<strong>on</strong> is what technologies are critical to the firm’s competitive advantage. In this c<strong>on</strong>text, technology<br />
must include the know-how the firm needs to create, produce <strong>and</strong> market its products <strong>and</strong> deliver them<br />
to customers. As a major step in creating a technology strategy it has to define those capabilities where<br />
the firm seeks to achieve a distinctive advantage relative to competitors. For most firms, there are a large<br />
number of important areas of technological know-how but <strong>on</strong>ly a h<strong>and</strong>ful where the firm will seek to create<br />
truly superior capability.<br />
Having determined the focus of technical development <strong>and</strong> the source of capability, the firm must establish<br />
the timing <strong>and</strong> frequency for innovati<strong>on</strong> efforts. Part of the timing issue involves developing technical<br />
capabilities, <strong>and</strong> the rest involves introducing technology into the market. The frequency of implementati<strong>on</strong><br />
<strong>and</strong> associated risks will depend in part <strong>on</strong> the nature of the technology <strong>and</strong> the markets involved (e.g. disk<br />
drive vs. automotive technology), but in part <strong>on</strong> strategic choice. At the extreme, a firm may adopt a rapid<br />
incremental strategy, that is, frequent, small changes in technology that cumulatively lead to c<strong>on</strong>tinuous<br />
performance improvement. The polar opposite might be termed the great leap forward strategy. In this<br />
approach, a firm chooses to make infrequent but large-scale changes in technology that substantially<br />
advance the state of the art .<br />
Sec. 2.3 uncovers the importance of asymmetries by investigating the difference between firms in terms of<br />
key parameters. Moreover, for the examinati<strong>on</strong> of inter-firm competiti<strong>on</strong> through product development, the<br />
c<strong>on</strong>text of simultaneous entry <strong>and</strong> sequential entry are treated separately. This allows a deeper underst<strong>and</strong>ing<br />
of the implicati<strong>on</strong>s of informati<strong>on</strong> asymmetry <strong>and</strong> commitment which have been regarded as important<br />
determinants in the c<strong>on</strong>text of game-theoretic studies.<br />
The study of innovati<strong>on</strong> based competiti<strong>on</strong> has often c<strong>on</strong>sidered aspects related to patent races <strong>and</strong> incremental<br />
product-process innovati<strong>on</strong> to achieve distinctive advantage. However, recently innovati<strong>on</strong>-based competiti<strong>on</strong><br />
has become an aspect of buyer-supplier relati<strong>on</strong>ships. There are many instances in manufacturing where <strong>on</strong>e<br />
finds situati<strong>on</strong>s of lock-ins created by innovative suppliers. For example, in the computer industry Intel <strong>and</strong><br />
Microsoft as suppliers of microprocessor <strong>and</strong> operating systems, respectively, to desktop manufacturers like<br />
IBM , Hewlett-Packard <strong>and</strong> Dell illustrate such innovati<strong>on</strong>-based lock-ins. Indeed, there exists an evolving<br />
power structure (dubbed “channel power”) in a supply chain driven by innovati<strong>on</strong> competence of its members.<br />
A business c<strong>on</strong>text is envisaged in which, at any given point in time of the relati<strong>on</strong>ship, both the buyer <strong>and</strong><br />
the supplier could be pursuing innovati<strong>on</strong> simultaneously. We recognize that the primary motivati<strong>on</strong> for<br />
such investments in innovati<strong>on</strong> by members of the supply chain is to increase their differential or relative<br />
IX
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
channel power in the supply chain. Also in a situati<strong>on</strong> where the buyer is locked in by a supplier, the<br />
buyer may actively pursue the creati<strong>on</strong> of a substitute technology by investing in innovati<strong>on</strong>. The primary<br />
motivati<strong>on</strong> for the buyer in this case would be to eliminate the technology lock-in <strong>and</strong> become independent.<br />
Another c<strong>on</strong>tributi<strong>on</strong>, Sec.2.4, presents the competitive role of innovati<strong>on</strong> am<strong>on</strong>g collaborating firms.<br />
The models provide reas<strong>on</strong>ing for inter-firm incentives in forming collaborative arrangements for product<br />
development. The dynamics of relati<strong>on</strong>ship am<strong>on</strong>g supply chain partners is viewed in terms of their<br />
respective innovati<strong>on</strong> competence. It is emphasized that varying power arrangements in a supply chain leads<br />
to different implicati<strong>on</strong>s for investments in innovati<strong>on</strong> by buyer <strong>and</strong> supplier. It highlights the incentive for<br />
buyer <strong>and</strong> supplier to strategically maneuver their overall innovati<strong>on</strong> levels by appropriate investments.<br />
The purpose is to provide a strategic framework <strong>and</strong> insights regarding power <strong>and</strong> competiti<strong>on</strong> in a<br />
collaborative supply chain setup.<br />
Managerial dimensi<strong>on</strong>s<br />
The analytical results provide insights into managerial implicati<strong>on</strong>s regarding strategic issues c<strong>on</strong>cerned<br />
with product development <strong>and</strong> supply chain planning. The c<strong>on</strong>ceptual framework provides a foundati<strong>on</strong> for<br />
managers to view competitive advantage emanating from product development in a more integrated manner.<br />
The results of Sec.1.6 point towards a time variant innovati<strong>on</strong> investment strategy. It suggests that firms<br />
competing by means of new product development must choose instantaneous investment in innovati<strong>on</strong><br />
such that it increases with time. The quality manifestati<strong>on</strong>s exhibit an increasing trajectory. In a symmetric<br />
competiti<strong>on</strong>, the investment profile of the leader is sigmoidal while that of the follower is c<strong>on</strong>vex increasing.<br />
‘T’he impact of such an investment profile is aptly reflected in the quality improvement trajectory.<br />
The results provide implicati<strong>on</strong>s for a firm for innovati<strong>on</strong> investments by c<strong>on</strong>sidering their relative strengths<br />
<strong>and</strong> weaknesses. The explorati<strong>on</strong> of firm asymmetries provide some interesting implicati<strong>on</strong> that a firm<br />
should c<strong>on</strong>sider when competing via new product development. The results can be translated into executable<br />
decisi<strong>on</strong>s for investments. These results provide insights to evaluate different scenarios for pursuing strategies<br />
which would lead to best market outcome.<br />
Secs. 2.3 <strong>and</strong> 2.4 provide a way of thinking about business strategy <strong>and</strong> operati<strong>on</strong>al alignment in a<br />
collaborative network. The results provide directi<strong>on</strong>s for strategies to be adopted in a supply chain c<strong>on</strong>text.<br />
These are becoming increasingly important in the present business envir<strong>on</strong>ment, where many firms are<br />
joined together in a collaborative network. In industries driven by innovati<strong>on</strong> ,e,g. semi-c<strong>on</strong>ductors or<br />
biotechnology, having the c<strong>on</strong>trol <strong>on</strong> the overall innovati<strong>on</strong> levels that drive the technology l<strong>and</strong>scape could<br />
mean a str<strong>on</strong>g strategic advantage for <strong>on</strong>e partner over other c<strong>on</strong>tributing firms.<br />
Innovati<strong>on</strong> competence plays an important role in this approach <strong>on</strong> a supply chain relati<strong>on</strong>ship. In Sec. 2.4<br />
the specific scenario modeled is representative of practical situati<strong>on</strong>s. As an example, it can be observed<br />
that IBM is actively involved in creating quantum computers. One reas<strong>on</strong>ing for investments in quantum<br />
computers could be an extensi<strong>on</strong> of a technology l<strong>and</strong>scape. However, equally important is the realizati<strong>on</strong><br />
that a success of such an endeavour could lead to a lock-out of Intel from this newly formed market for<br />
quantum courputers. The simplistic model used in Sec. 2.4 allows for an explorati<strong>on</strong> of such a situati<strong>on</strong>.<br />
The stati<strong>on</strong>ary Markov perfect Nash equilibrium investment strategy of the buyer <strong>and</strong> supplier is found to<br />
be time invariant <strong>and</strong> is characterized by the parameters in the model. An interesting insight obtained as<br />
a result of analysis of the model is that the supplier can influence the motivati<strong>on</strong> of the buyer to invest in<br />
substitute technology. The underlying mechanism can be translated into pricing strategy that results in a<br />
l<strong>on</strong>g-term buyer-supplier relati<strong>on</strong>ship.<br />
The analysis of a model of buyer-supplier relati<strong>on</strong>ship allows to investigate a more generalized setup of<br />
buyer-supplier innovati<strong>on</strong>-based competiti<strong>on</strong>. The aspect of channel power is tightly integrated in the<br />
analysis to underst<strong>and</strong> the complex behavioral aspects using a simple framework. These results indicate<br />
that structural risk plays an important role in innovati<strong>on</strong> investment decisi<strong>on</strong>s. Moreover, the variance in<br />
wealth formati<strong>on</strong> is an important indicator of how much to invest. These results are quite intuitive <strong>and</strong><br />
X
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
enable managers to objectively resolve some of these strategic decisi<strong>on</strong>s.<br />
The propositi<strong>on</strong>s provide insights into the dynamics of relati<strong>on</strong>ship between a buyer <strong>and</strong> a supplier firm.<br />
These propositi<strong>on</strong>s present the role <strong>and</strong> resp<strong>on</strong>sibility of buyer in creating motivati<strong>on</strong> for the supplier to<br />
collaborate. The importance of critical assets is aptly amplified in the results. The possessi<strong>on</strong> of critical<br />
assets, often observed with the upstream partner gives the supplier a potential to achieve relative market<br />
closure through a positi<strong>on</strong> of dominance over competitors. It is likely that a firm in possessi<strong>on</strong> of such a<br />
critical asset also has the potential to achieve effective leverage over collaborating partner-, <strong>and</strong> suppliers.<br />
The resp<strong>on</strong>sibility of the buyer is to create a delicate balance between managing supplier <strong>and</strong> customers.<br />
Similar to the c<strong>on</strong>ceptual development the results suggest that for a buyer that wishes to be successful an<br />
underst<strong>and</strong>ing must be developed about how to own <strong>and</strong> c<strong>on</strong>trol critical assets that provide opportunities to<br />
create customer dependency <strong>and</strong> “lock-in” .A competence in procurement management forms the cornerst<strong>on</strong>e<br />
for success of l<strong>on</strong>g term business strategy.<br />
Extensi<strong>on</strong>s<br />
Several extensi<strong>on</strong>s can be c<strong>on</strong>sidered for supply chain modeling <strong>and</strong> they establish behavioral features for<br />
the Microfoundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth, Sec. 3.1.<br />
First; within the framework of Sec.1.6 the model can be enriched by evaluating the improvement in product<br />
quality with learning effect. The learning effect is very well documented in the literature <strong>and</strong> the extensi<strong>on</strong><br />
to incorporate learning effects is straightforward. Specifically, learning can be used to characterize the<br />
dynamics of evoluti<strong>on</strong> of product quality <strong>and</strong> in the characterizati<strong>on</strong> of costs associated with innovati<strong>on</strong><br />
investments.<br />
For analytical simplicity the revenue functi<strong>on</strong> is treated as a salvage value. As an extensi<strong>on</strong> the incorporati<strong>on</strong><br />
of the dynamics of evoluti<strong>on</strong> of revenue in the analysis of the model can be explored. In the sequential play<br />
game, the state dynamics of the follower can be c<strong>on</strong>sidered to be dependent <strong>on</strong> the state of the leader at<br />
previous time instant. This modificati<strong>on</strong> would allow for analysis of a model in which the leader uses openloop<br />
Nash equilibrium strategy for innovati<strong>on</strong> investments whereas the follower would adopt a Markovian<br />
Nash equilibrium strategy. Additi<strong>on</strong>al insights can be gained by this modificati<strong>on</strong> for leader-follower<br />
competitive dynamics in new product development.<br />
Firm asymmetries can be explored by c<strong>on</strong>sidering multiple parameters at the same time. This would enable<br />
a richer underst<strong>and</strong>ing of the strategies that a leader <strong>and</strong> a follower should adopt based <strong>on</strong> their strengths<br />
<strong>and</strong> weaknesses.<br />
The competitive dynamics of buyers <strong>and</strong> suppliers are explored. The essays present the competitive role<br />
of innovati<strong>on</strong> am<strong>on</strong>g collaborating firms. The research asserts that the buyer <strong>and</strong> the supplier could be<br />
in a competitive relati<strong>on</strong>ship due to efforts in innovati<strong>on</strong>. Existing literature in operati<strong>on</strong>s management<br />
doesn’t explicitly c<strong>on</strong>sider the implicati<strong>on</strong>s of innovati<strong>on</strong> based competitive strategies in a collaborative<br />
supply chain c<strong>on</strong>text. Models are presented that provide motivati<strong>on</strong> to explicitly c<strong>on</strong>sider such competitive<br />
situati<strong>on</strong>s. The noti<strong>on</strong> of lock-in <strong>and</strong> channel power is implicit in the model <strong>and</strong> presents an approach for<br />
theory-building activities in this area.<br />
We know that the model presented derives from a simplified versi<strong>on</strong> of the dynamics between a buyer <strong>and</strong><br />
a supplier. The c<strong>on</strong>straints placed <strong>on</strong> different parameters <strong>and</strong> the c<strong>on</strong>text of a m<strong>on</strong>opoly supplier <strong>and</strong> a<br />
buyer facing perfect competiti<strong>on</strong> is an approximati<strong>on</strong> for analytical c<strong>on</strong>venience. However, this enables a<br />
theoretical investigati<strong>on</strong> of some important aspects underlying buyer-supplier relati<strong>on</strong>ships. Specifically,<br />
the research provides a way of thinking about business strategy in collaborative networks. These are<br />
becoming increasingly important in the present business envir<strong>on</strong>ment, where many firms are part of a<br />
collaborative network. In industries driven by innovati<strong>on</strong> , the industrial internet, advanced manufacturing,<br />
semic<strong>on</strong>ductors <strong>and</strong> biotechnology, having c<strong>on</strong>trol over the overall innovati<strong>on</strong> levels could mean a str<strong>on</strong>g<br />
strategic advantage for <strong>on</strong>e partner over other c<strong>on</strong>tributing firms.<br />
At the same time, the firm already having the channel power can take some acti<strong>on</strong>s that would enable a<br />
XI
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
l<strong>on</strong>g-term relati<strong>on</strong>ship built <strong>on</strong> trust based governance.<br />
Several extensi<strong>on</strong>s could be c<strong>on</strong>sidered to better represent such “coopetitive” relati<strong>on</strong>ships between<br />
collaborating firm. Given the objective interests, it is necessary for collaborating partners to know how far<br />
the other side is prepared to c<strong>on</strong>cede before it is no l<strong>on</strong>ger profitable to be in a relati<strong>on</strong>ship. Informati<strong>on</strong> about<br />
the cost structures of the collaborating partners <strong>and</strong> their relative utility from the exchange relati<strong>on</strong>ship is<br />
critical to an underst<strong>and</strong>ing of power in exchange relati<strong>on</strong>ship. Hence, <strong>on</strong>e of the extensi<strong>on</strong> of the problem<br />
is to c<strong>on</strong>sider aspects of informati<strong>on</strong> asymmetry. The two key problems generated by private informati<strong>on</strong><br />
<strong>and</strong> imperfect observability are – adverse selecti<strong>on</strong> <strong>and</strong> moral hazard. The adverse selecti<strong>on</strong> is a c<strong>on</strong>diti<strong>on</strong><br />
of supplier opportunism that occurs prior to a signing of a c<strong>on</strong>tract. Moral hazard <strong>on</strong> the other h<strong>and</strong> refers<br />
to supplier opportunism that occurs <strong>on</strong>ce the buyer has signed a c<strong>on</strong>tract.. A rati<strong>on</strong>al buyer tries to possess<br />
informati<strong>on</strong> that counters the above motives <strong>on</strong> the part of suppliers. Aspects related to detailed price<br />
comparis<strong>on</strong> between the supplier; its competitors <strong>and</strong> the range of substitutes provides means to investigate<br />
buyer strategies. An investigati<strong>on</strong> of the implicati<strong>on</strong>s of threat by buyer of placing the c<strong>on</strong>tract elsewhere<br />
to get additi<strong>on</strong>al price c<strong>on</strong>cessi<strong>on</strong>s from its preferred supplier can be carried out.<br />
The assumpti<strong>on</strong>s made in the models <strong>and</strong> the c<strong>on</strong>straints put <strong>on</strong> variables can be relaxed to obtain a better<br />
representati<strong>on</strong>. A scenario with multiple suppliers/buyers enables enrichment of the findings . The result<br />
regarding guidelines for price b<strong>and</strong>width to create l<strong>on</strong>g-term relati<strong>on</strong>ships can be explored in further detail.<br />
The implicati<strong>on</strong>s <strong>on</strong> profits by adopting a strategy for l<strong>on</strong>g-term relati<strong>on</strong>ship can be compared <strong>and</strong> c<strong>on</strong>trasted<br />
with those of adopting the Markov-perfect Nash equilibrium strategies. The insights obtained from this<br />
approach also provide a background for future studies in supply chain c<strong>on</strong>tracts. The problem c<strong>on</strong>text <strong>and</strong><br />
results clearly dem<strong>on</strong>strate the competitive role of innovati<strong>on</strong> between the buyer <strong>and</strong> the supplier. It provides<br />
a rati<strong>on</strong>ale for investigati<strong>on</strong> of supply chain c<strong>on</strong>tracts by explicitly c<strong>on</strong>sidering the evoluti<strong>on</strong> of innovati<strong>on</strong><br />
competence <strong>and</strong> critical assets of collaborating partners.<br />
In Sec.2.4 the stochastic differential. game model c<strong>on</strong>siders the Brownian moti<strong>on</strong> for the evoluti<strong>on</strong> of<br />
wealth for supplier <strong>and</strong> buyer. These Brownian moti<strong>on</strong>s can be treated to be autocorrelated to investigate<br />
the associated implicati<strong>on</strong>s of investment strategy. The model can be extended to examine a sequential<br />
game <strong>and</strong> the associated investment strategy for buyer <strong>and</strong> supplier. Another extensi<strong>on</strong> to the model is<br />
c<strong>on</strong>siderati<strong>on</strong> of spill-over effects. The spill-over of knowledge created by investments in innovati<strong>on</strong> is<br />
comm<strong>on</strong> in literature pertaining to ec<strong>on</strong>omics <strong>and</strong> management science. Such a spill-over of knowledge<br />
may provide some interesting results regarding buyer-supplier power structure <strong>and</strong> investment strategy.<br />
In Part 3, Ec<strong>on</strong>omic Growth <strong>and</strong> <strong>Development</strong>, Secs. 3.1 – 3.7 we start with a broad survey in Sec. 3.1 <strong>on</strong><br />
the microfoundati<strong>on</strong>s of ec<strong>on</strong>omic growth where many model results <strong>on</strong> firm <strong>and</strong> industrial racing behavior<br />
in Parts 1 <strong>and</strong> 2 are sourced to built the structural basis for industrial growth <strong>and</strong> in the aggregate for<br />
macroec<strong>on</strong>omic growth. In Sec.3.2 the link micro-macro is the focus of a modified classical growth model<br />
(the Solow model) which at the center is built up<strong>on</strong> catch-up behavior in the aggregate. Secs. 3.3 <strong>and</strong> 3.4<br />
turn the page <strong>on</strong> development , Sec. 3.3 <strong>on</strong> catch-up behavior <strong>and</strong> development paths between nati<strong>on</strong>s in<br />
a global <strong>and</strong> historical c<strong>on</strong>text from the first industrial revoluti<strong>on</strong> <strong>on</strong> c<strong>on</strong>vergence <strong>and</strong> divergence of paths<br />
<strong>and</strong> changing positi<strong>on</strong>s <strong>on</strong> various performance criteria in ec<strong>on</strong>omic growth <strong>and</strong> prosperity. While Sec. 3.3<br />
is more c<strong>on</strong>cerned with “racing to the top” in a more closely related field Sec. 3.4 emphasizes development<br />
paths from the bottom to overcome or break out of development traps or poverty traps. Secs. 3.5 <strong>and</strong><br />
3.6 partially depart from the main theme, Sec. 3.5 portrays a a c<strong>on</strong>trol theoretic model of development<br />
planning, more a traditi<strong>on</strong>al macro model either to be superimposed <strong>on</strong> or interacting cooperatively with<br />
more decentralized market <strong>and</strong> trade oriented industrial structures. Sec. 3.6 emphasizes another aspect<br />
of opimal ec<strong>on</strong>omic growth under possible fixed/variable c<strong>on</strong>staints trough climate change restricti<strong>on</strong>s.<br />
In what follows we restrict comments <strong>on</strong> the main theme of Part 3 .<br />
On ec<strong>on</strong>omic growth <strong>and</strong> development we assume that a key driver of ec<strong>on</strong>omic growth am<strong>on</strong>g nati<strong>on</strong>s<br />
or regi<strong>on</strong>al ec<strong>on</strong>omic entities is an intrinsic, <strong>on</strong>going, perpetual <strong>and</strong> historically observable rivalry to<br />
propel a state‘s st<strong>and</strong>ing, prestige, power <strong>and</strong> ec<strong>on</strong>omic performance through getting ahead of or not to<br />
fall too much behind its rivals in a pecking order. As we observe in ec<strong>on</strong>omic history through the age of<br />
industrializati<strong>on</strong> from the mid 18th century, the cumulative process of industrial development in a dynamic<br />
time c<strong>on</strong>text can be correlated with some countries emerging as early leaders being challenged by others,<br />
XII
PREFACE AND INTRODUCTION<br />
HANS W. GOTTINGER<br />
later adopters that strived for new leadership positi<strong>on</strong>s. This is in c<strong>on</strong>trast to those that were not part of the<br />
techno-ec<strong>on</strong>omic race, falling behind in ec<strong>on</strong>omic prosperity, <strong>and</strong> inevitably not being able to catch-up<br />
over decades or even centuries.<br />
The design mechanism for strategies of ec<strong>on</strong>omic growth essentially embodies a science <strong>and</strong> technology<br />
competiti<strong>on</strong> between nati<strong>on</strong>s <strong>and</strong> its industries as the creator of value added products <strong>and</strong> processes through<br />
competiti<strong>on</strong> as reflected in the level of gross domestic product (GDP). This is in the Schumpeterian traditi<strong>on</strong>.<br />
It follows from exploring the varying implicati<strong>on</strong>s of the noti<strong>on</strong>, clearly advanced by Schumpeter (1942) that<br />
it is the expectati<strong>on</strong> of supernormal profits from a temporary m<strong>on</strong>opoly positi<strong>on</strong> following an innovati<strong>on</strong><br />
that is the chief driver of R&D investment.<br />
The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of advanced technology or the possibility of leapfrogging<br />
is mainly attributable to how social capability evolves, (i.e., which effects become more influential, growing<br />
resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of vested interests <strong>and</strong> established<br />
positi<strong>on</strong>s). In a time of overcoming a major financial crisis with a slowing pace of deregulati<strong>on</strong>, privatizati<strong>on</strong>,<br />
liberalizati<strong>on</strong> <strong>and</strong> lifting of trade barriers it remains an interesting questi<strong>on</strong> regarding an internati<strong>on</strong>al<br />
competitive order whether industry racing patterns are sufficiently c<strong>on</strong>trolled by open world-wide markets<br />
or whether complementary internati<strong>on</strong>al agreements (e.g. regulati<strong>on</strong>s, c<strong>on</strong>trols) are needed to eliminate or<br />
mitigate negative externalities without compromising the positive externalities that come with industry racing.<br />
We investigate multiple kinds of races, fr<strong>on</strong>tier races am<strong>on</strong>g leaders <strong>and</strong> would be’ leaders <strong>on</strong> a global,<br />
regi<strong>on</strong>al <strong>and</strong> local scale corresp<strong>on</strong>ding to different leagues like in major sports events, as there are in each<br />
category catchup races am<strong>on</strong>g laggards <strong>and</strong> imitators. Technological fr<strong>on</strong>tiers at the firm <strong>and</strong> industry race<br />
levels offer a powerful tool through which to view evolving technologies within an industry. By providing<br />
benchmarking roadmaps that show where an individual firm is relative to other firms in the industry, they<br />
highlight the importance of strategic interacti<strong>on</strong>s in the firm’s technology decisi<strong>on</strong>s.<br />
In an ec<strong>on</strong>omic history perspective getting ahead, catching up <strong>and</strong> falling behind processes are taking<br />
place toward leaders <strong>and</strong> followers, trailers <strong>and</strong> laggards within a group of industrialized, industrializing<br />
<strong>and</strong> developing countries in pursuit of higher levels of power, welfare <strong>and</strong> productivity. Moses Abramovitz<br />
(1986) explains the central idea of the catch-up hypothesis as the trailing countries adopting behaviour<br />
of a backlog of unexploited technology. Supposing that the level of labor productivity were governed<br />
entirely by the level of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the differentials in<br />
productivities am<strong>on</strong>g countries are caused by the “technological age” of the stock used by a country relative<br />
to its “chr<strong>on</strong>ological age”.<br />
The technological age of capital is an age of technology at the time of investment plus years elapsing from<br />
that time. Since a leading country may be supposed to be furnished with the capital stock embodying, in<br />
each vintage, technology which was at the very fr<strong>on</strong>tier at the time of investment, the technological age<br />
of the stock is, so to speak, the same as its chr<strong>on</strong>ological age. While a leader is restricted in increasing<br />
its productivity by the advance of new technology, trailing countries, al<strong>on</strong>g Abramovitz’ line, have the<br />
potential to make a larger leap as they are provided with the privilege of exploiting the backlog in additi<strong>on</strong><br />
to the newly developed technology. Hence, followers being behind with a larger gap in technology will<br />
have a str<strong>on</strong>ger potential for growth in productivity. The potential, however, will be reduced as the catchup<br />
process goes <strong>on</strong> because the unexploited stock of technology becomes smaller <strong>and</strong> smaller.<br />
This hypothesis explains the diffusi<strong>on</strong> process of best practice technology <strong>and</strong> gives the same sort of<br />
S-curve change in productivity rise of catching-up countries am<strong>on</strong>g a group of industrialized countries as<br />
that of followers to the leader in an industry. Although this view can explain the tendency to c<strong>on</strong>vergence<br />
of productivity levels of follower countries, it fails to answer the historical puzzles why a country, such as<br />
the United States, has preserved the st<strong>and</strong>ing of the technological leader (taking excepti<strong>on</strong>s for two world<br />
wars) for a l<strong>on</strong>g time since taking over leadership from Britain <strong>and</strong> Germany before the First World War<br />
<strong>and</strong> why the shifts have taken place in the ranks of follower countries in their relative levels of productivity,<br />
i.e., technological gaps between them <strong>and</strong> the leader.<br />
Abramovitz poses some extensi<strong>on</strong>s <strong>and</strong> qualificatios <strong>on</strong> this simple catchup hypothesis in an attempt to<br />
XIII
HANS W. GOTTINGER<br />
PREFACE AND INTRODUCTION<br />
explain these facts. Am<strong>on</strong>g other factors than technological backwardness he lays stress <strong>on</strong> a country’s<br />
social capability, i.e., years of educati<strong>on</strong> as a proxy of technical competence in its political, commercial,<br />
industrial, <strong>and</strong> financial instituti<strong>on</strong>s. The social capability of a country may become str<strong>on</strong>ger or weaker<br />
as technological gaps close <strong>and</strong> thus, he states, the actual catch-up process does not lend itself to simple<br />
formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what Mancur Ols<strong>on</strong> (1982) expresses to be public<br />
policies <strong>and</strong> instituti<strong>on</strong>s’ as his explanati<strong>on</strong> of the great differences in per capita income across countries,<br />
stating that any poorer countries that adopt relatively good ec<strong>on</strong>omic policies <strong>and</strong> instituti<strong>on</strong>s enjoy rapid<br />
catchup growth. This view should be taken seriously when we wish to underst<strong>and</strong> the technological catching<br />
up to American leadership in some key industries by Japan, in particular, during the post-war period <strong>and</strong><br />
explore the possibility of a shift in st<strong>and</strong>ing between these two countries. This c<strong>on</strong>siderati<strong>on</strong> will directly<br />
bear <strong>on</strong> the future trend of the state of the art which exerts a crucial influence <strong>on</strong> the development of the<br />
world ec<strong>on</strong>omy.<br />
On a macro scale this racing paradigm would suggest that the U.S. would mobilize all its technology/<br />
science/entrepreneurial resources to increase or keep its distance to other large emerging ec<strong>on</strong>omies (such<br />
as China or India). With science/ technology being an evoluti<strong>on</strong>ary cumulative enterprise, for a dominating<br />
country with a portfolio of major increasing returns industries, the odds of leapfrogging <strong>on</strong>eself are higher<br />
than being leapfrogged by close followers, thus this asymmetry could play a distinctive role. Abramovitz’<br />
advantage of backwardness may hold <strong>on</strong> up to a certain limit but with decreasing returns.<br />
In an interesting paper, under some seemingly reas<strong>on</strong>able assumpti<strong>on</strong>s <strong>on</strong> linear technology investment<br />
<strong>and</strong> dynamic equilibrium path of capital accumulati<strong>on</strong> in a neoclassical type model Lau <strong>and</strong> Wan (1993)<br />
obtain the following results which they argue are fully c<strong>on</strong>sistent with empirical growth ec<strong>on</strong>omics: These<br />
are: (a) Not all ec<strong>on</strong>omies c<strong>on</strong>verge in growth with each other. (b) Ec<strong>on</strong>omies with an initial technical<br />
capability will c<strong>on</strong>verge in growth with the advanced ec<strong>on</strong>omies. The difference in per capita output grows<br />
exp<strong>on</strong>entially, if the developing ec<strong>on</strong>omy engages <strong>on</strong>ly in imitati<strong>on</strong> (<strong>and</strong> not innovati<strong>on</strong>). (c) With an initial<br />
technical capability there is a ‘high growth’ period, preceded (followed) by a phase of ‘trend accelerati<strong>on</strong>’<br />
(‘trend decelerati<strong>on</strong>’). (d) With an initial technical capability the the technology gap widens forever. The<br />
previous analysis indicates that the possibilities opened through competitive industrial racing are far richer<br />
<strong>and</strong> more surprising than they would emerge from the macro-scale models.<br />
A dynamic competitive nati<strong>on</strong>al ec<strong>on</strong>omy with its underlying instituti<strong>on</strong>al factors is fed by a c<strong>on</strong>sistent<br />
industrial policy intrinsically supported by a strategic innovati<strong>on</strong> system. Although this view can explain the<br />
tendency to c<strong>on</strong>vergence of productivity levels of follower countries, it fails to answer the historical puzzle<br />
why a country, the United States, has preserved the st<strong>and</strong>ing of the technological leader for a l<strong>on</strong>g time this<br />
century, <strong>and</strong> why it has intermittently been threatened by Germany, Japan <strong>and</strong> lately by emerging China <strong>and</strong><br />
to a lesser degree by India. We show the empirical proliferati<strong>on</strong> of comparative growth <strong>and</strong> development<br />
over the main period of industrializati<strong>on</strong>. Why have the shifts taken place in the ranks of follower countries<br />
in their relative levels of productivity, i.e., technological gaps between them <strong>and</strong> the leader?<br />
XIV
1.<br />
INNOVATION<br />
AND<br />
TECHNOLOGY
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
3
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
4
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
5
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
6
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
7
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
8
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
9
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
10
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
11
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
12
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
13
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
14
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
HANS W. GOTTINGER<br />
15
HANS W. GOTTINGER<br />
1.1 STOCHASTICS OF INNOVATION PROCESSES<br />
16
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
17
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
18
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
19
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
20
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
21
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
22
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
23
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
24
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
25
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
26
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
27
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
28
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
29
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
30
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
HANS W. GOTTINGER<br />
31
HANS W. GOTTINGER<br />
1.2 STOCHASTIC MODELS FOR CAPITAL GROWTH<br />
32
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
33
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
34
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
35
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
36
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
37
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
38
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
39
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
40
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
41
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
42
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
43
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
44
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
45
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
46
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
47
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
48
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
49
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
50
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
HANS W. GOTTINGER<br />
51
HANS W. GOTTINGER<br />
1.3 MODELING STOCHASTIC INNOVATION RACES<br />
52
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
53
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
54
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
55
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
56
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
57
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
58
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
59
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
60
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
61
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
62
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
63
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
64
1.4 GLOBAL TECHNOLOGICAL RACES<br />
HANS W. GOTTINGER<br />
65
HANS W. GOTTINGER<br />
1.4 GLOBAL TECHNOLOGICAL RACES<br />
66
240 Int. J. <strong>Technology</strong>, Policy <strong>and</strong> Management, Vol. 4, No. 3, 2004<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network<br />
markets<br />
HANS W. GOTTINGER<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Institute of Management Sciences, Maastricht,<br />
Kwansei Gakuin University (KGU),<br />
Kobe-S<strong>and</strong>a Campus, Japan<br />
E-mail: bvz28020@ksc.kwansei.ac.jp<br />
Abstract: We look at a competitive situati<strong>on</strong> in network markets where there is<br />
uncertain technological development in product/process technologies. Firms<br />
‘price’ compete in those markets to gain market share before any of them<br />
succeeds in getting an innovati<strong>on</strong> to move ahead of its rival(s). If the firms are<br />
in a technology race <strong>and</strong> the probability of innovati<strong>on</strong> success is small, then a<br />
firm with a bigger network advantage is likely to attract more customers in the<br />
absence of innovati<strong>on</strong>. If, however, any of those firms expect innovati<strong>on</strong> with a<br />
high probability, <strong>and</strong> n<strong>on</strong>e of them have a big advantage, then they keep <strong>on</strong><br />
sharing the market until the innovati<strong>on</strong> occurs.<br />
Keywords: network ec<strong>on</strong>omics; market structure <strong>and</strong> pricing; firm strategy;<br />
market performance.<br />
Reference to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (2004)<br />
‘Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets’, Int. J. <strong>Technology</strong>,<br />
Policy <strong>and</strong> Management, Vol. 4, No. 3, pp.240–256.<br />
Biographical notes: Dr <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> is Professor of Ec<strong>on</strong>omics at the University<br />
of Osaka, <strong>and</strong> with the Institute of Management Science, Maastricht. His fields<br />
of interest include network ec<strong>on</strong>omics, strategy <strong>and</strong> competiti<strong>on</strong> issues, energy<br />
<strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics.<br />
1 Introducti<strong>on</strong><br />
Many high technology industries are characterised by positive network externalities.<br />
Firms essentially compete against each other for goods that share a network.<br />
Our model of competiti<strong>on</strong> c<strong>on</strong>tains two special features that apply equally well to<br />
network markets. One is the uncertainty in technological development or uncertainty in<br />
the realisati<strong>on</strong> of a firm’s R&D effort. The other is the dynamic nature of price<br />
competiti<strong>on</strong> between firms in the presence of network effects.<br />
Firms compete with each other over an extended period of time <strong>and</strong> must, therefore,<br />
strategically choose prices as the market shares of the firms evolve.<br />
Most existing models of network markets focus <strong>on</strong> <strong>on</strong>ly <strong>on</strong>e of the two features.<br />
Further, almost all the models have the feature that <strong>on</strong>e firm captures the market<br />
instantaneously <strong>and</strong> sells to all c<strong>on</strong>sumers from then <strong>on</strong> (the ‘winner-takes-all’ market).<br />
However, the history of the high technology industry abounds with instances where rival<br />
Copyright © 2004 Inderscience Enterprises Ltd.<br />
67
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 241<br />
firms (<strong>and</strong> technologies) have had extended battles for providing the industry st<strong>and</strong>ard<br />
(Uttenback, 1994).<br />
In network markets, evoluti<strong>on</strong> of market share is a very interesting phenomen<strong>on</strong><br />
especially in the face of uncertainty about the future product quality. This paper also<br />
attempts to provide a framework that could capture the richness of the market share<br />
evoluti<strong>on</strong> in the presence of network externalities.<br />
This paper is a follow-up to a stochastic racing model (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2002a): it also takes<br />
some cues <strong>on</strong> racing from the seminal c<strong>on</strong>tributi<strong>on</strong> by C. Futia (1980).<br />
In its strategic dimensi<strong>on</strong> it is closely related to Katz <strong>and</strong> Shapiro (1992) <strong>and</strong> Harris<br />
<strong>and</strong> Vickers (1987), particularly, in their ‘tug-of-war’ presentati<strong>on</strong> <strong>and</strong> multi-stage racing.<br />
Overall our approach is emerging from new ideas <strong>on</strong> anti-trust analysis in<br />
dynamically competitive industries (Evans <strong>and</strong> Schmalensee, 2001).<br />
It is illuminating to take a full quote out of their comprehensive analysis:<br />
“Firms that are not leaders in network industries generally have little hope of<br />
reaching that status unless they come up with a major innovati<strong>on</strong> – <strong>on</strong>e that can<br />
defeat the natural advantage that network effects bestow <strong>on</strong> the industry<br />
leaders. Incremental innovati<strong>on</strong> – making slight improvements in the leaders’<br />
products – will not enable a small firm to overtake a leader that enjoys the<br />
benefits of network ec<strong>on</strong>omies. Similarly, the possibility of being displaced by<br />
a major innovati<strong>on</strong> will shape leaders’ research agendas. If there is a chance<br />
that today’s products will be replaced by a major innovati<strong>on</strong>, a leader’s survival<br />
depends <strong>on</strong> bringing this innovati<strong>on</strong> to market <strong>and</strong> thereby replacing itself<br />
before others do. As a result, competiti<strong>on</strong> in network industries often involves<br />
intense R&D efforts aimed at capturing or retaining market leadership. It is not<br />
atypical for a fringe firm that invests heavily to displace the leader by<br />
leapfrogging the leader’s technology…” (Evans <strong>and</strong> Schmalensee, 2001,<br />
pp.10–12)<br />
We are exploring this problem of trade-off between ‘network dominance’ <strong>and</strong> ‘radical<br />
innovati<strong>on</strong>’ that could tip the market the other way, with the significant complexity added<br />
that breakthrough R&D is highly uncertain. From a strategic perspective, in this<br />
envir<strong>on</strong>ment, for any two firms of asymmetric size, both compete dynamically over<br />
prices to win market share. In this dynamic process there are two ways to achieve<br />
(temporary) m<strong>on</strong>opolistic status. The ‘smaller’ firm can use dynamic pricing competiti<strong>on</strong><br />
to delay the time in which the ‘larger’ firm wins a critical market share in the hope to hit<br />
the innovati<strong>on</strong> first <strong>and</strong> displace it. If the probability of innovati<strong>on</strong> <strong>and</strong> the discount factor<br />
are sufficiently high, there is an equilibrium in which duopoly persists (no firm achieves<br />
a critical market share) until <strong>on</strong>e of the two firms wins the race for innovati<strong>on</strong>.<br />
We take a historical example from markets for IT products. In the pers<strong>on</strong>al computer<br />
(PC) operating system market. Microsoft products (MS DOS <strong>and</strong> Windows) have been<br />
dominating the market since the mid 1980s. An operating system is the fundamental<br />
programme that c<strong>on</strong>trols the allocati<strong>on</strong> <strong>and</strong> use of computer resources. Thus, the utility<br />
that operating systems provide to c<strong>on</strong>sumers depends <strong>on</strong> the number of compatible<br />
applicati<strong>on</strong>s. As a general rule, an applicati<strong>on</strong> that relies <strong>on</strong> a specific operating system<br />
will not functi<strong>on</strong> <strong>on</strong> another operating system unless it is ported to that specific operating<br />
system. Therefore, because of its dominance, the majority of applicati<strong>on</strong>s have been<br />
written to run <strong>on</strong> Microsoft operating systems (MS DOS). The dominati<strong>on</strong> of MS DOS<br />
has become even str<strong>on</strong>ger since the arrival of Windows 95 in the PC operating system<br />
market. This, in turn, has provided a great indirect positive network externality to PC<br />
owners who adopted Windows 95 as an operating system. Many other firms, such as IBM<br />
68
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
242 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
<strong>and</strong> Be Inc., introduced their own operating systems <strong>and</strong> tried to compete with Windows.<br />
These products, however, lacked sufficient compatible applicati<strong>on</strong>s to efficiently compete<br />
with Microsoft products. The lack of compatible applicati<strong>on</strong>s prevented enough<br />
applicati<strong>on</strong> developers <strong>and</strong> c<strong>on</strong>sumers from regarding OS/2 Warp or BeOS as a viable<br />
alternative to the dominant incumbent, Windows. This obstacle prevented these potential<br />
entrants from obtaining a sizeable market share. Their failure to enter the market<br />
successfully, however, was not due to the inferior quality of their operating systems.<br />
In fact, OS/2 Warp is reported to be at least as good as Windows, <strong>and</strong> BeOS offers<br />
superior support for multimedia applicati<strong>on</strong>s. If c<strong>on</strong>sumers who use multimedia<br />
applicati<strong>on</strong>s frequently adopt Windows at the expense of BeOS, they have to give up the<br />
c<strong>on</strong>venience that is provided by BeOS. Thus, for multimedia specific users, adopting<br />
BeOS as their operating system at the expense of another operating system might provide<br />
the highest utility. Nevertheless, the lack of compatible applicati<strong>on</strong>s, which in turn<br />
implies the lack of positive network externality, has prevented c<strong>on</strong>sumers from adopting<br />
OS/2 Warp or BeOS. As a c<strong>on</strong>clusi<strong>on</strong>, the positive network externality for the dominant<br />
incumbent (Windows) has worked as an entry barrier against entrants (OS/2 Warp or<br />
BeOS), which do not have network externality. Such an entry barrier can <strong>on</strong>ly be<br />
overcome by a radical innovati<strong>on</strong>, virtually leapfrogging the dominant incumbent.<br />
2 Market histories<br />
We substantiate our claim <strong>on</strong> a natural tradeoff between network size <strong>and</strong> quality change<br />
(through significant technological change) by looking independently <strong>on</strong> four market<br />
histories for the IT industry.<br />
2.1 Internet explorer<br />
The first versi<strong>on</strong> of Internet Explorer was introduced in the web browser market in July<br />
1995. This versi<strong>on</strong> of Internet Explorer was reportedly inferior to Netscape’s Navigator.<br />
Microsoft made many improvements in its next versi<strong>on</strong>, Internet Explorer 3.0, in August<br />
1996. Internet Explorer 3.0 matched nearly all the features in Navigator. However,<br />
Microsoft was not able to draw significant market share at the expense of Navigator with<br />
this enhanced versi<strong>on</strong>. Internet Explorer 4.0, released in late 1997, finally managed to<br />
obtain roughly the same number of reviewers who preferred it to Navigator as the number<br />
of those who preferred Navigator. In 1996, most c<strong>on</strong>sumers were much more familiar<br />
with Navigator, the dominant web browser than with Internet Explorer. Thus although the<br />
quality of Internet Explorer was much improved since its release into the market, more<br />
than 80% of c<strong>on</strong>sumers were still using Navigator until late 1996. It was evident that<br />
Internet Explorer could not draw a significant percentage of c<strong>on</strong>sumers away from<br />
Navigator by simply improving the quality of its web browser. Senior executives within<br />
Microsoft c<strong>on</strong>cluded that Internet Explorer could not catch up with Navigator’s market<br />
share by itself. J. Allchin, senior executive within Microsoft, wrote to P. Martiz, in late<br />
December 1996:<br />
“I d<strong>on</strong>’t underst<strong>and</strong> how IE (Internet Explorer) is going to win. The current<br />
path is simply to copy everything that Netscape does, packaging <strong>and</strong> product<br />
wise. Let’s {suppose} IE is as good as Navigator/Communicator. Who wins?<br />
The <strong>on</strong>e with 80% market share. Maybe being free helps us, but <strong>on</strong>ce people<br />
are used to a product it is hard to change them. C<strong>on</strong>sider Office. We are most<br />
expensive today, but we’re still winning.”<br />
69
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 243<br />
From 1996 to 1998, Internet Explorer’s market share increased dramatically.<br />
The estimated market share of Navigator fell from 80% in January 1996 to 55% in<br />
November 1997, while that of Internet Explorer increased from 5% to 36% over the same<br />
period. These changes in market shares, however, were mainly due to the fact that<br />
Microsoft leveraged Windows to increase the usage of Internet Explorer, rather than<br />
c<strong>on</strong>stantly improving the quality of its browser.<br />
2.2 Server operating system<br />
The envir<strong>on</strong>ment of the server operating system market is different from that of the<br />
PC-operating system in the sense that a server is maintained by a group of computer<br />
experts (not by novices as in the case of a PC) <strong>and</strong> is intended for specific purposes,<br />
rather than for general uses. Moreover, a choice of server operating systems requires<br />
many technical features, such as functi<strong>on</strong>ality, system management, performance,<br />
reliability (stability) <strong>and</strong> security. Functi<strong>on</strong>ality <strong>and</strong> performance are important technical<br />
features for both PCs <strong>and</strong> servers, but other aspects are not very important for PCs. This<br />
difference comes from the fact that while a PC is designed for use by <strong>on</strong>e pers<strong>on</strong> at a<br />
time, a server is designed to provide data, services, <strong>and</strong> functi<strong>on</strong>ality through a digital<br />
network to multiple users. Most individual PC owners do not have an incentive to buy a<br />
server-operating system, since server operating systems generally cannot functi<strong>on</strong><br />
efficiently <strong>on</strong> PC hardware. In other words, the c<strong>on</strong>sumers in the two markets are<br />
basically separated. According to Data Quest, the Unix System is currently the most<br />
widely used operating system platform. It exp<strong>and</strong>ed its market share from 36% in 1996 to<br />
43% in 1998. The newcomer in the server operating system market, Windows NT, also<br />
achieved great success during that period; its market share rose from 10% in 1996 to 16%<br />
in 1998. Linux licence shipments grew at a rate of 212.5%, accounting for more than<br />
17% of all server operating system units shipped, Other than Unix servers <strong>and</strong> Windows<br />
NT, Open VMS <strong>and</strong> Novel.<br />
NetWare <strong>and</strong> OS/2 are the major competitors in the server-operating system market.<br />
A small product differentiati<strong>on</strong> in a server-operating system enables it to create <strong>and</strong> to<br />
maintain its own dem<strong>and</strong>. In sum, a dominant product cannot eliminate other products<br />
with its prevailing positive network externalities when substitutability for products are<br />
low.<br />
2.3 OS/2 warp<br />
In late 1994, IBM introduced its Intel-<strong>on</strong>ly compatible operating system, OS/2 Warp, as<br />
an alternative to Windows. OS/2 was originally designed to be optimised for Intel<br />
microprocessors. This resulted in a lack of portability. Because the Intel compatible PC<br />
dominated the PC market, the lack of portability did not prevent c<strong>on</strong>sumers from<br />
adopting OS/2 as an operating system. Furthermore, OS/2 was not reported to be inferior<br />
to Windows. The greatest disadvantage of adopting OS/2 as an operating sytem was the<br />
lack of compatible software; approximately 2,500 applicati<strong>on</strong>s were available. In 1995,<br />
IBM still attempted to compete with Windows 95 in the PC-operating system market<br />
with its release of OS/2 2.0. However, IBM figured that it would lose between 70% <strong>and</strong><br />
90% of its sales volume if it installed OS/2 to the exclusi<strong>on</strong> of Windows 95. In 1996,<br />
IBM withdrew from the PC-operating system market <strong>and</strong>, instead, has been repositi<strong>on</strong>ing<br />
OS/2 for the server-operating system market. The failure of OS/2 Warp to gain enough<br />
70
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
244 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
market share to survive is mainly due to its lack of applicati<strong>on</strong>s rather than inferiority to<br />
Windows. By the time OS/2 Warp was released, Windows was already the predominant<br />
operating system <strong>and</strong> ran an overwhelming number of applicati<strong>on</strong>s. Most c<strong>on</strong>sumers used<br />
Windows not because it was superior in quality to its competitors, but rather because it<br />
provided the greatest positive network externality am<strong>on</strong>g the available PC-operating<br />
systems.<br />
In other words, Microsoft successfully deterred entry into the PC-operating system<br />
market by using its predominating network externality.<br />
2.4 BeOS<br />
Be, Inc., founded in 1990 by a former executive of Apple’s product divisi<strong>on</strong>, released its<br />
cross-platform operating system, BeOS. BeOS functi<strong>on</strong>s virtually identically <strong>on</strong> both<br />
Intel-compatible PCs <strong>and</strong> Macintosh systems. In additi<strong>on</strong>, BeOS is optimised for<br />
multiprocessor media-based systems. These two features provide great c<strong>on</strong>venience for<br />
multimedia-specific users. Be has stated from the beginning that it intends to coexist<br />
within the Windows <strong>and</strong> Apple World as a specialised operating system. Even taking this<br />
specific purpose into account, the market share of BeOS is too small compared to the<br />
number of Windows users. The most outst<strong>and</strong>ing disadvantage is the lack of compatible<br />
applicati<strong>on</strong>s (approx. 1,000). Be, Inc. tried to overcome this problem by maintaining a<br />
website which provided a list of freeware <strong>and</strong> shareware applicati<strong>on</strong>s. Be, Inc. failed to<br />
gain sizeable market share, <strong>and</strong> was eventually sold to Palm. However, Palm c<strong>on</strong>tinues to<br />
produce BeOS <strong>and</strong> to maintain its BeWare website. BeOS still exists as a niche operating<br />
system, in spite of its superior support for multimedia applicati<strong>on</strong>s. The failure of BeOS<br />
to obtain a significant market share is evidence in support of the propositi<strong>on</strong> that the<br />
incumbent can deter entry by using its network externality.<br />
3 Some issues of technological racing<br />
Katz <strong>and</strong> Shapiro (1992) look at whether there is too much or too little technological<br />
innovati<strong>on</strong> in a market with network externalities, compared to the social optimum. In<br />
particular, they explore whether a new product, which embodies technological progress,<br />
is introduced too early or too late compared to the social optimum. In their model,<br />
technological progress is deterministic. Only the entrant gets the benefit of improving<br />
technology, but <strong>on</strong>ce it enters the market its product is fixed. Therefore, in any<br />
equilibrium of their model, either the new firm captures the market as so<strong>on</strong> as it enters or<br />
it does not enter at all. Their model is thus not able to capture the richness of pricing<br />
strategies <strong>and</strong> market evoluti<strong>on</strong> as a functi<strong>on</strong> of market shares. For instance, it is unable<br />
to explain the phenomen<strong>on</strong> of two competing technologies fighting for superiority in a<br />
market when both are making sales until <strong>on</strong>e firm finally leaves the market. Harris <strong>and</strong><br />
Vickers (1987) present two models of races in which there is both technological<br />
uncertainty <strong>and</strong> strategic interacti<strong>on</strong> between competitors as the race unfolds.<br />
Their aim is to see how the equilibrium effort levels of competitors vary with the<br />
intensity of rivalry between them. They show that under certain c<strong>on</strong>diti<strong>on</strong>s the leader in<br />
the race makes greater efforts than the follower, <strong>and</strong> efforts increase as the gap between<br />
competitors in their progress decreases. The model presented here can be seen al<strong>on</strong>g<br />
71
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 245<br />
similar lines to that of Harris <strong>and</strong> Vickers, but more realistically, with the network<br />
advantage as the state variable instead of progress, <strong>and</strong> prices providing strategic<br />
interacti<strong>on</strong> between firms instead of the amount of effort put into R&D. While there is<br />
<strong>on</strong>e ultimate goal for each firm in the multistage race of Harris <strong>and</strong> Vickers to reach the<br />
finishing line of the race first, we leave it open for firms to decide which network they<br />
want to build. It is this feature that can lend profit maximisati<strong>on</strong> by firms to a complex<br />
network evoluti<strong>on</strong> pattern. Also, the amount of effort put into R&D stochastically moves<br />
firms to a new stage in the race, whereas in our model prices – the variables of strategic<br />
interacti<strong>on</strong> – deterministically move firms to a new market share.<br />
While a lot of work has been d<strong>on</strong>e in the area of R&D, very little of it specifically<br />
addresses markets with network externalities. Katz <strong>and</strong> Shapiro look at the choice of<br />
product compatibility in a market with technological progress. They look at both the<br />
cases of deterministic <strong>and</strong> uncertain technological progress. However, technological<br />
progress in their model is exogenous <strong>and</strong> they look at a two-period game, which does not<br />
provide a very suitable framework for looking at the dynamic pricing strategy of firms<br />
<strong>and</strong> the resulting evoluti<strong>on</strong> of market shares. Choi (1994) looks at a two-period model of<br />
a m<strong>on</strong>opoly in a market with network externalities. He studies the incentive of the<br />
m<strong>on</strong>opolist to introduce an incompatible improved product in the presence of network<br />
externalities. Kristiansen (1998) studies the c<strong>on</strong>sequences of network externalities <strong>on</strong> the<br />
riskiness of R&D projects chosen by an entrant <strong>and</strong> an incumbent. He shows that the<br />
incumbent chooses a risky project that very often lets a new firm with an incompatible<br />
technology enter. In additi<strong>on</strong>, the entrant has an incentive to choose more certain projects<br />
than are socially optimal <strong>and</strong> these strengthen the possibility of adopti<strong>on</strong> of an<br />
incompatible technology.<br />
This paper develops <strong>and</strong> analyses a model of technological uncertainty where firms<br />
compete strategically in prices in the presence of network externalities. Technological<br />
uncertainty is modelled as an improvement in the quality of the firm’s product.<br />
Whether the firm gets the innovati<strong>on</strong>, however, is uncertain. In the simplest model,<br />
<strong>on</strong>ce a firm gets the innovati<strong>on</strong>, the race ends <strong>and</strong> there is no further innovati<strong>on</strong>.<br />
Once the innovati<strong>on</strong> occurs, when there is no uncertainty, the model looks like that of<br />
Katz <strong>and</strong> Shapiro – the firm with the quality plus network advantage sells to all the<br />
c<strong>on</strong>sumers subsequently. We attempt to analyse how firms share the market by<br />
strategically choosing their prices before they succeed in getting an innovati<strong>on</strong>.<br />
The advantage of a firm is measured by the sum of its network size <strong>and</strong> the quality of<br />
its product. If the firms are in a technological race but the probability of innovati<strong>on</strong> is<br />
small, then again, in the absence of innovati<strong>on</strong>, the firm with the bigger advantage sells to<br />
all the c<strong>on</strong>sumers. If, however, those firms expect innovati<strong>on</strong> with a high probability then<br />
so l<strong>on</strong>g as neither firm has a very big advantage both firms take turns in selling to the<br />
c<strong>on</strong>sumers in the absence of innovati<strong>on</strong>. Thus, if the firms are <strong>on</strong> an equal base, neither<br />
firm gets pushed out unless <strong>on</strong>e of the firms gets an insurmountable advantage in the<br />
process. That is because with firms that are sufficiently close in terms of their advantage,<br />
neither firm finds it worthwhile to push the other firm out of the market since the ‘losing’<br />
firm prices more aggressively than the ‘leader’ as it has more to gain by staying in the<br />
market.<br />
72
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
246 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
4 Asymmetric R&D races<br />
We provide some anecdotal observati<strong>on</strong>s across various industries that show a str<strong>on</strong>g<br />
rivalry between a leader in the field <strong>and</strong> the nearest rival(s) <strong>on</strong> the technological fr<strong>on</strong>t <strong>and</strong><br />
its efforts to gain market share.<br />
We identify those interactive patterns as technological races. We can even extend<br />
those observati<strong>on</strong>s to megamergers <strong>and</strong> alliances as products of network ec<strong>on</strong>omies that<br />
did not protect those networks from relentless technological attacks by upstarts, new<br />
entrants or Schumpeterian entrepreneurs hailing ‘creative destructi<strong>on</strong>’. This asymmetric<br />
view reflects the match between the ‘Old Guard vs the Vanguard’ (Hamel <strong>and</strong> Switzer,<br />
2004) in the sense that an incumbent’s positi<strong>on</strong> endows her with a natural network<br />
advantage which could <strong>on</strong>ly be overcome by a breakthrough innovati<strong>on</strong> of a new entrant<br />
depending <strong>on</strong> some probability of innovati<strong>on</strong> success. We have manifestati<strong>on</strong>s of intense<br />
rivalry <strong>and</strong> ‘head-to-head’ competiti<strong>on</strong> across industries, be it advanced microprocessors<br />
between Intel <strong>and</strong> American Micro Devices (AMD), (Markoff <strong>and</strong> Lohr 2003), specialty<br />
pharmaceuticals against chr<strong>on</strong>ic diseases between Merck, Glaxo <strong>and</strong> Pfizer, <strong>and</strong> between<br />
biotech companies Amgen <strong>and</strong> Biogen. In the exp<strong>and</strong>ing market of software related web<br />
services it is Microsoft against IBM, Sun Microsystems <strong>and</strong> Oracle (Lohr, 2003), in<br />
servers it is IBM vs Sun <strong>and</strong> H-P (Clark, 2003) in c<strong>on</strong>sumer electr<strong>on</strong>ics <strong>and</strong> design it is<br />
S<strong>on</strong>y against Matsushita or Sanyo (Belsen, 2003).<br />
5 Modelling a race in a network<br />
We start out with a simple model in which two firms, Y <strong>and</strong> Z, are manufacturing a<br />
product using incompatible technologies: the product is infinitely durable. The firms have<br />
products of identical qualities that we normalise to zero. For simplicity the marginal cost<br />
of producti<strong>on</strong> is assumed to be zero for both firms.<br />
We define: The network base of firm i at time t, b it , is the total number of c<strong>on</strong>sumers<br />
it has sold to until time t, i ∈ {Y, Z}.<br />
The network advantage of firm i at time t, n it ≡ b it – b jt , i, j∈ {Y, Z}.<br />
A new c<strong>on</strong>sumer arrives in the market periodically. Each c<strong>on</strong>sumer has an inelastic<br />
unit dem<strong>and</strong> for the product <strong>and</strong> must decide the firm from which to buy the product as<br />
so<strong>on</strong> as he/she enters the market. C<strong>on</strong>sumers care about how many other c<strong>on</strong>sumers buy<br />
the same product because of network externalities. That is, they are likely to choose<br />
products of the larger network over those of the smaller. They also care about the quality<br />
of the product <strong>and</strong> its price. For simplicity, all ‘representative’ c<strong>on</strong>sumers have identical<br />
preferences. We may state their utility functi<strong>on</strong>s as u (n t , q t, p t ) where q t is the quality of<br />
the product as a perfect substitute for the network advantage <strong>and</strong> p t is its price.<br />
Throughout the phase of adding uncertainty to the innovati<strong>on</strong> process the c<strong>on</strong>sumers are<br />
expected to be risk-neutral. We assume that there is a threshold difference, D, in network<br />
sizes of the two firms such that if neither firm’s network is D (critically bigger than the<br />
other’s) then c<strong>on</strong>sumers get the same network benefits from the two firms. More<br />
realistically, <strong>on</strong>e could think of D not just as a number but also as an ‘order of<br />
magnitude’. Therefore, if neither firm has a network advantage of D then c<strong>on</strong>sumers buy<br />
from the firm which charges a lower price if the quality is the same. C<strong>on</strong>sumers are<br />
indifferent about the two firms if they charge the same price. Once a firm establishes a<br />
73
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 247<br />
network advantage of D over the other firm, c<strong>on</strong>sumers get a much bigger benefit from it.<br />
Thus, if both firms charge the same price then c<strong>on</strong>sumers prefer the firm that has the<br />
network advantage of D (or more). It follows that <strong>on</strong>ce a firm gets a network advantage<br />
of D it becomes a (temporary) m<strong>on</strong>opolist of the market.<br />
If both firms did not have identical qualities to start out with, then we could think of<br />
the network advantage as comprising the network difference plus the quality difference,<br />
that is, treat network <strong>and</strong> quality as perfect substitutes in the c<strong>on</strong>sumer’s utility.<br />
The proposed rule for c<strong>on</strong>sumer decisi<strong>on</strong>s implies that c<strong>on</strong>sumers are myopic, that is,<br />
the current base of c<strong>on</strong>sumers with the firms is what matters to them. While in practice<br />
we would expect c<strong>on</strong>sumers to form expectati<strong>on</strong>s about the future networks of firms, <strong>and</strong><br />
thus introduce the possibility of multiple fulfilled-expectati<strong>on</strong>s equilibria, our assumpti<strong>on</strong><br />
of myopic c<strong>on</strong>sumers would help select the most efficient equilibrium outcome am<strong>on</strong>g<br />
them. Both assumpti<strong>on</strong>s – of myopic c<strong>on</strong>sumers <strong>and</strong> c<strong>on</strong>sumers with fully rati<strong>on</strong>al<br />
expectati<strong>on</strong>s – are extreme <strong>on</strong>es, <strong>and</strong> what is more realistic is something in between the<br />
two. Having myopic c<strong>on</strong>sumers has the added advantage of making the analysis more<br />
tractable.<br />
Let n Yt represent the state of the system (we could equivalently choose n Zt to be the<br />
state variable). Thus state 1 denotes the situati<strong>on</strong> where firm Y has <strong>on</strong>e (order of<br />
magnitude) more c<strong>on</strong>sumer than firm Z, state – 1 denotes the situati<strong>on</strong> where firm Z has<br />
<strong>on</strong>e (order of magnitude) more c<strong>on</strong>sumer than firm Y. We define 3 to be the absorbing<br />
state for firm Y <strong>and</strong> – 3 the absorbing state for firm Z, that is we let D equal 3. The net<br />
discounted payoff of a firm in its absorbing state is M, for the other firm likewise. The<br />
absorbing state can be thought of as the threshold advantage in quality or network, which<br />
is big enough to deter the other firm from ever trying to catch up. Thus the firm that has<br />
this advantage becomes a m<strong>on</strong>opolist in the market. It is the sole seller in the market<br />
making m<strong>on</strong>opoly profits of M whereas the other firm leaves the market <strong>and</strong> thus gets<br />
zero profits.<br />
The product is subject to technological innovati<strong>on</strong>. We assume that an innovati<strong>on</strong><br />
pushes a firm to its absorbing state. Thus, we assume that the innovati<strong>on</strong> is equivalent to<br />
a big enough network advantage that gives a firm an insurmountable lead over the other<br />
firm, in many given cases of asymmetrical network advantage the innovati<strong>on</strong>s need to be<br />
‘radical’. The innovati<strong>on</strong>, however, is uncertain, either a firm gets the required quality<br />
improvement or there is no improvement at all. The probability that a firm gets the<br />
innovati<strong>on</strong> in any time period is α ∈ [0,1/2] that is the same for both firms. We could<br />
think of α > 0 as representing the technological envir<strong>on</strong>ment in which the firms operate.<br />
We assume that <strong>on</strong>ce <strong>on</strong>e firm succeeds, there are no further innovati<strong>on</strong>s in the market.<br />
We also assume for notati<strong>on</strong>al simplicity that at most <strong>on</strong>e firm gets the innovati<strong>on</strong>. Thus,<br />
the probability that neither firm gets the innovati<strong>on</strong> in any <strong>on</strong>e time is 1 – 2α. In this<br />
situati<strong>on</strong>, we look at the innovati<strong>on</strong> process as ‘exogenous’ (though we could make it<br />
‘endogenous’ by letting the probability of innovati<strong>on</strong> success depend <strong>on</strong> the network size<br />
as a proxy for the degree of accumulated knowledge (learning) in determining innovati<strong>on</strong><br />
success).<br />
We could think of the technological innovati<strong>on</strong> as a new feature that the firms try to<br />
introduce in their respective products. Once <strong>on</strong>e firm gets its innovati<strong>on</strong>, it can get a<br />
patent even though the firms produce incompatible products. We assume that the new<br />
improved product that a firm might introduce is compatible with its older product.<br />
74
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
248 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Compatibility, in this model, translates into network sharing – compatible products share<br />
a network.<br />
Firms compete in prices <strong>and</strong> thus must decide what price to charge at each of the<br />
states in S = {–2, –1,0, 1,2}, in order to maximise their respective profits given by<br />
π Y (s 0 ,p) = ∑ t δ t (1–2α) t–1 [(1–2α)p Yt x Yt + αM] + δ τ+1 I(s τ+1 = 3)M, t = 0,1,2, …. , τ<br />
π Z (s 0 ,p) = ∑ t δ t (1–2α) t–1 [(1–2α)p Zt x Zt + αM] + δ τ+1 I(s τ+1 = 3)M, t = 0,1,2, …. , τ<br />
x<br />
it<br />
s<br />
t+<br />
1<br />
⎧1<br />
⎪<br />
= ⎨0<br />
⎪<br />
⎩<br />
0 or 1<br />
= s<br />
t<br />
+ x<br />
if p<br />
if p<br />
if p<br />
Yt<br />
it<br />
it<br />
it<br />
,<br />
< p<br />
> p<br />
= p<br />
jt<br />
jt<br />
jt<br />
where p it is the price charged by firm i in period t, x it ∈ {0, 1} is the number of period t<br />
c<strong>on</strong>sumers that the firm i sells to: τ + 1 is the time period in which a firm reaches its<br />
absorbing state even without getting the innovati<strong>on</strong>: s 0 ∈ S is the initial state; <strong>and</strong><br />
p = (p 1 , p 2 ), where<br />
p i = { p it }, t = 0,1,2,...,τ.<br />
We thus have a game where the firms are identical except in the sizes of their networks.<br />
There is uncertain technological innovati<strong>on</strong> but the size of the innovati<strong>on</strong> <strong>and</strong> the<br />
probability of innovati<strong>on</strong> are again identical for both firms. Therefore, by comparing the<br />
soluti<strong>on</strong> of a game with no uncertainty with the soluti<strong>on</strong> of our game, we should be able<br />
to gain valuable insights into the effects uncertainty has <strong>on</strong> the competiti<strong>on</strong> between two<br />
firms in the presence of network externalities.<br />
We look for Markov equilibria (Fudenberg <strong>and</strong> Tirole 1993), which are<br />
n<strong>on</strong>-collusive <strong>and</strong> symmetric. In such an equilibrium we c<strong>on</strong>sider a firm that is indifferent<br />
about selling or not selling. Thus if firm Y is the firm not selling at state s, then<br />
p Y (s) + δ ∏ Y (s+1) = δ ∏ Y (s–1)<br />
where ∏(s) represents the c<strong>on</strong>tinuati<strong>on</strong> payoff starting from state s. If Z is the firm not<br />
selling at state s then<br />
p Z (s) + δ ∏ Z (s–1) = δ ∏ Z (s+1).<br />
In such an equilibrium the n<strong>on</strong>-selling firm ensures that even if against all expectati<strong>on</strong>s<br />
the c<strong>on</strong>sumer buys its product then it will not make negative profits in equilibrium.<br />
Definiti<strong>on</strong> 1 An equilibrium is n<strong>on</strong>-collusive if neither firm charges a positive price<br />
at a n<strong>on</strong>-absorbing state. Alternatively, if either firm charges a positive price at a<br />
n<strong>on</strong>-absorbing state, equilibrium is collusive.<br />
Definiti<strong>on</strong> 2 An equilibrium is symmetric if p Y (s) = p Z (–s) for all permissible s <strong>and</strong><br />
x Y (s) = x Z (–s) for all permissible s ≠ 0.<br />
If instead of having an infinite horiz<strong>on</strong> game, we look at a game with a l<strong>on</strong>g but finite<br />
horiz<strong>on</strong>, the n<strong>on</strong>-collusive equilibria are the <strong>on</strong>es we expect would most likely be<br />
sustainable.<br />
We first c<strong>on</strong>sider the case in absence of uncertainty, that is, α = 0.<br />
75
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 249<br />
Propositi<strong>on</strong> 1 For α = 0, the symmetric n<strong>on</strong>-collusive Markov equilibrium prices are<br />
unique: the firm with a positive network advantage makes all the sales in equilibrium.<br />
Proof: Suppose firm Z sells at s = 2. If firm Y were to sell to the c<strong>on</strong>sumer at s = 2,<br />
then in the next period it would get a profit of M. Therefore, if in a n<strong>on</strong>-collusive<br />
equilibrium firm Y does not sell at s = 2 as we have assumed, it must be getting a<br />
c<strong>on</strong>tinuati<strong>on</strong> payoff at least as big as M next period. There are two possibilities for s = 1:<br />
either firm Y sells or firm Z does.<br />
Suppose firm Y sells at s = 1. Then <strong>on</strong>ce the state reaches 1, it must alternate between<br />
1 <strong>and</strong> 2 forever more from then <strong>on</strong>. The c<strong>on</strong>tinuati<strong>on</strong> payoff to firm Y at state 1 then is<br />
just the discounted sum of the price it charges to the c<strong>on</strong>sumer at s = 1 every other period.<br />
For this to be at least M, the price at s = 1 must be positive <strong>and</strong> equilibrium is then<br />
collusive. Suppose firm Z sells at s = 1 also. Then <strong>on</strong>ce s = 1 is reached, it must remain<br />
within the states –1, 0, 1 forever more. But then again for firm Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at<br />
s = 1 to be at least M, it must get a positive payoff at s = 0 or –1, a c<strong>on</strong>tradicti<strong>on</strong>. Thus, Y<br />
must sell at s = 2. Suppose firm Z sells at s = 1 <strong>and</strong> Y sells at s = 2. Then <strong>on</strong>ce state 1 is<br />
reached, the state must remain within the states –1, 0, 1. In a n<strong>on</strong>-collusive equilibrium<br />
where neither firm charges a positive price, Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 0 cannot be<br />
positive. Z’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 2 is zero. Similarly, the smallest price Z is willing<br />
to charge at s = 1 cannot be less than zero. However, firm Y’s c<strong>on</strong>tinuati<strong>on</strong> payoff at s =<br />
2 is then positive <strong>and</strong> at most zero at s = 0, it is therefore willing to charge a negative<br />
price at s = 1 in order to lure the c<strong>on</strong>sumer into buying its product. But that means the<br />
c<strong>on</strong>sumer should buy from firm Y at s = 1 since its price is lower than Z’s, a<br />
c<strong>on</strong>tradicti<strong>on</strong>. Thus firm Y must sell at s = 1 also.<br />
We have so far shown that if a symmetric, n<strong>on</strong>-collusive Markov equilibrium exists,<br />
then it must have the firm with the network advantage selling. We now show that such an<br />
equilibrium does indeed exist <strong>and</strong> that it is (almost) unique.<br />
If in an equilibrium the firm with the network advantage always wins then the<br />
c<strong>on</strong>tinuati<strong>on</strong> payoffs of both firms at s = 0 must be zero in a symmetric equilibrium since<br />
both firms are willing to give up exactly the same surplus in order to win the c<strong>on</strong>sumer.<br />
This surplus is equal to the c<strong>on</strong>tinuati<strong>on</strong> payoff from winning the current, which has to be<br />
at least zero.<br />
At s = 1, the lowest firm Z is willing to charge is zero since whether it sells to the<br />
current c<strong>on</strong>sumer or not, it is going to get a payoff of zero from then <strong>on</strong>. But that means it<br />
must charge a price of zero at s = 2 also. The c<strong>on</strong>tinuati<strong>on</strong> payoff at s = 0 from winning is<br />
then δ 2 M.<br />
Thus, both firms charge –δ 3 M at s = 0 <strong>and</strong> get zero discounted payoffs. It can be<br />
checked that these are equilibrium prices.<br />
Thus, the unique equilibrium prices are as follows:<br />
p(2) = p(–2) = 0<br />
p(1) = p(–1) = 0<br />
p(0) = – δ 3 M.<br />
Both firms charge the same prices. Firm Y sells to the c<strong>on</strong>sumer at s = 1, 2,... firm Z sells<br />
to the c<strong>on</strong>sumer at s = –1, –2, either firm could sell at s = 0.<br />
Thus, we see that in the absence of uncertainty, the firm that is already ahead in terms<br />
of network size is the ‘winning’ firm. The other firm that has a network disadvantage is<br />
76
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
250 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
unable to overtake the firm that is ahead. This result is similar to that obtained by Katz<br />
<strong>and</strong> Shapiro (1992). Once a firm acquires a bigger network, the other firm is unable to<br />
overcome its disadvantage. Since in our model, quality <strong>and</strong> network are perfect<br />
substitutes in the c<strong>on</strong>sumer’s utility, we could replace network by quality to get the same<br />
result. The firm with the higher quality sells to all c<strong>on</strong>sumers, the lower quality firm is<br />
unable to push out the better quality firm from the market, given that both firms start out<br />
with the same network size.<br />
We now introduce uncertainty into the game <strong>and</strong> see how firms behave. As in the<br />
previous case, the <strong>on</strong>ly difference between the firms is in their networks. The uncertainty<br />
faced by firms is symmetric: both firms have the same probability of getting an<br />
innovati<strong>on</strong> <strong>and</strong> the effect of an innovati<strong>on</strong> is symmetric for both firms.<br />
Propositi<strong>on</strong> 2 There exist δ * <strong>and</strong> α * such that for δ * < δ < 1 <strong>and</strong> 0 < α < α * the<br />
n<strong>on</strong>-collusive symmetric equilibrium prices are unique. The firm with the bigger network<br />
sells to the c<strong>on</strong>sumer, either firm could sell when s = 0.<br />
Proof:<br />
The proof proceeds al<strong>on</strong>g a sequence of Lemmas.<br />
Since we are looking for symmetric equilibria, at s = 0 the lowest price both firms are<br />
willing to charge must be the same. Thus, in equilibrium both firms must be indifferent<br />
between selling or not at s = 0. In other words, at s = 0 both firms have the expected<br />
payoffs. Again, since we are looking at symmetric equilibria we need to <strong>on</strong>ly focus <strong>on</strong><br />
behaviour at s = 0, 1, 2. The different possibilities at s = 1 <strong>and</strong> s = 2 determine the<br />
different c<strong>and</strong>idates for equilibrium outcomes. There are four possible c<strong>and</strong>idates for<br />
equilibrium network advantage evoluti<strong>on</strong>, given that no innovati<strong>on</strong> occurs in the<br />
meantime.<br />
(i) x Y (1) = 1, x Y (2) = 1,<br />
(ii) x Z (1) = 1, x Y (2) = 1,<br />
(iii) x Z (1) = 1, x Z (2) = 1,<br />
(iv) x Y (1) = 1, x Z (2) = 1<br />
where x i (s) = 1 means that at state s firm i sells to <strong>on</strong>e c<strong>on</strong>sumer (<strong>and</strong> thus j ≠ i sells to no<br />
<strong>on</strong>e since there is <strong>on</strong>ly <strong>on</strong>e c<strong>on</strong>sumer in the market at any time period) in the absence of<br />
innovati<strong>on</strong> having taken place.<br />
Define p i * (s) to be the lowest price firm i is willing to charge <strong>and</strong> p(s) to be the<br />
equilibrium price at s. The lowest a firm is willing to charge at any state is just the<br />
difference between the profits if it lost the current c<strong>on</strong>sumer <strong>and</strong> if it w<strong>on</strong> it. In other<br />
words, the maximum a firm is willing to give up in order to attract a c<strong>on</strong>sumer is exactly<br />
the amount of extra profit the firm can expect to make by selling to that c<strong>on</strong>sumer. Define<br />
π i W (s), π i L (s) to be the expected profits of firm i next period <strong>on</strong>wards from respectively<br />
winning <strong>and</strong> losing the current c<strong>on</strong>sumer at state s, let π i (s) be the ex ante expected<br />
payoff of i at state s.<br />
We prove the propositi<strong>on</strong> using the following Lemmas:<br />
Lemma 1<br />
Proof:<br />
Case (iii) cannot occur in equilibrium.<br />
Suppose (iii) described the equilibrium network evoluti<strong>on</strong>. Then,<br />
p Y * (1) = p(1) = δ(1 – 2α)p(0) (1)<br />
77
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 251<br />
This is just the difference in payoffs to firm Y from losing versus winning the current<br />
c<strong>on</strong>sumer.<br />
Similarly,<br />
π Y W (1) = 0 + δαM + δ (1 – 2α) π Y (1)<br />
π Y L (1) = (1 – 2α)p(0) + δαM + δ (1 – 2α) π Y (1)<br />
p Y * (1) = π Y L (1) – π Y W (1).<br />
p Z * (0) = p(0) = δ(1 – 2α)p(1) (2)<br />
Equati<strong>on</strong>s 1 <strong>and</strong> 2 together imply that<br />
p(1) = δ 2 (1 – 2α) 2 p(1).<br />
But that, in turn, implies that either p(1) = 0 or δ = 1 <strong>and</strong> α = 0. Since we are looking at<br />
the case where α ≠ 0 , p(1) must be zero, which means from (2), p(0) = 0. Thus<br />
π<br />
Y<br />
αM<br />
(0) = π<br />
Z<br />
(0)<br />
≡ π(0)<br />
1 − δ(<br />
1 − 2α)<br />
.<br />
We can calculate p Z * (2) <strong>and</strong> p Y * (2) (= p(2) since Y is the losing firm at s = 2)<br />
p Z * (2) = – δαM – δ 2 (1 – 2α)π(0)<br />
p Y * (2) = δαM + δ 2 (1 – 2α)π(0) – δM.<br />
But for δ < 1, p Y * (2) < p Z * (2) which means firm Y must be the <strong>on</strong>e selling to the<br />
c<strong>on</strong>sumer at s = 2 , a c<strong>on</strong>tradicti<strong>on</strong>.<br />
Lemma 2 Case (iv) describes the network evoluti<strong>on</strong> in equilibrium <strong>on</strong>ly when<br />
δ 2 (1 – 2α) 2 > 1/2 .<br />
Proof: π Z W (1) = 0 + δαM + δ (1 – 2α) π Z (1)<br />
π Z L (1) = (1 – 2α)p(2) + δαM + δ (1 – 2α) π Z (1)<br />
p Z * (1) = p(1) = π Z L (1) – π Z W (1)<br />
⇒ p(1) = δ(1 – 2α)p(2) (3)<br />
We can calculate p(2) = p Y * (2) using the same method:<br />
δαM δ(1−2 α)<br />
p(2) =<br />
1−δ( 1−2α )<br />
p(1)<br />
−δM<br />
−<br />
1 −δ 2<br />
( 1 − 2 α ) 2<br />
)<br />
*<br />
z<br />
2<br />
2<br />
δαM δ (1−<br />
2α)<br />
(2) =<br />
−<br />
2<br />
1−<br />
δ(<br />
1−<br />
2α)<br />
1−<br />
δ ( 1−<br />
2α)<br />
p<br />
2<br />
p(2<br />
(4)<br />
Substituting the expressi<strong>on</strong> for p(2) in (3), we get,<br />
2<br />
2<br />
2<br />
2<br />
δ (1−<br />
2α)<br />
αM<br />
δ (1−<br />
2α)<br />
2<br />
p(1) = +<br />
p(1) − δ (1−<br />
2α)M<br />
(5)<br />
2<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
Similarly, we calculate p(0) using the relati<strong>on</strong> of (3) to get p(0) = 0 since<br />
78
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
252 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
2 2<br />
δ (1−2 α) δ(1−2 α)<br />
p(0) = p(2) −<br />
1 − (1−2 ) 1 − (1−2 )<br />
2<br />
2 2 2<br />
δ α δ α<br />
p(1) .<br />
This in turn implies that p Y * (1) = 0.<br />
For case (iv) to occur in equilibrium, p(1) must be n<strong>on</strong>-negative. That leads to p Y * (1)<br />
≤ p(1) ; it also implies that p(2) is n<strong>on</strong>-negative, therefore p Z * (2) ≤ 0 ≤ p(2) from Equati<strong>on</strong><br />
(4). From (5) we see that p(1) is n<strong>on</strong>-negative <strong>on</strong>ly if δ 2 (1 – 2α) 2 ≥ 1/2.<br />
Lemma 3 As δ → 1 <strong>and</strong> α → 0, the pattern of network evoluti<strong>on</strong> described in (iv)<br />
can <strong>on</strong>ly occur in a collusive equilibrium.<br />
Proof: As δ → 1 , α → 0 , δ 2 (1 – 2α) 2 ≥ 1/2. Thus from Lemma 2 case (iv) can occur<br />
in an equilibrium. But when δ 2 (1 – 2α) 2 ≥ 1/2 , we see from (5) that p(1) > 0, that is, the<br />
equilibrium is collusive.<br />
Lemma 4 Case (i) describes the pre-innovati<strong>on</strong> network evoluti<strong>on</strong> in a n<strong>on</strong>-collusive<br />
equilibrium <strong>on</strong>ly when 3α [1 + δ (1 – 2α) ] ≤ 1.<br />
Proof:<br />
We calculate the lowest prices firms are willing to charge as described earlier <strong>on</strong><br />
p Y * (2) = δαM + δ 2 (1 – 2α) αM + δ 3 (1 – 2α) 2 M – δM + δ (1 – 2α) p(1) +<br />
δ 2 (1 – 2α) 2 p(2)<br />
p Z * (2) = p(2) = − δ α M − δ 2 (1 – 2α) α M<br />
p Y * (1) = δ 2 (1 – 2α) αM + δ 3 (1 – 2α) 2 αM – δ 2 (1 – 2α) M – δ (1 – 2() p(2)<br />
p Z *(1) = p(1) = δ 2 (1 – 2α) αM – δ 3 (1 – 2α) 2 αM<br />
p(0) = – δ (1 – 2α) p(1) – δ 2 (1 – 2α) 2 p(2) – δ 3 (1 – 2α) 2 M<br />
= δ 3 (1 – 2α) 2 M [2α + 2δα (1 – 2α) – 1].<br />
We can use the above equati<strong>on</strong>s to solve for p(0), p (1), <strong>and</strong> p(2), simultaneously. It can<br />
be checked that p Y *(2) ≤ p(2) always holds, but that p Y *(1) ≤ p(1) holds <strong>on</strong>ly if<br />
3α [1 + δ (1 – 2α) ] ≤ 1 (6)<br />
It can be checked that when inequality (6) is satisfied, then p(0), p(1), <strong>and</strong> p(2) are all<br />
negative. Note that (6) holds when δ → 1 <strong>and</strong> α → 0. Thus, when inequality (6) is<br />
satisfied the network evoluti<strong>on</strong> described in case (i) occurs as a result of the n<strong>on</strong>-collusive<br />
equilibrium described by the prices given above. Both firms charge p(s), s ∈ {−2, −1, 0,<br />
1, 2}, where p(0), p(1) <strong>and</strong> p(2) are as given above <strong>and</strong> p(−2) = p(2), p(−1) = p(1). At s =<br />
1, 2 firm Y sells to the c<strong>on</strong>sumer in the market, at s = − 1, − 2 firm Z sells to the<br />
c<strong>on</strong>sumer, <strong>and</strong> at s = 0 either firm could sell.<br />
We see that the equilibrium when there is no uncertainty is just the limit of the above<br />
equilibrium as α → 0.<br />
Lemma 5 Case (ii) describes the pre-innovati<strong>on</strong> network evoluti<strong>on</strong> in a n<strong>on</strong>collusive<br />
equilibrium <strong>on</strong>ly when 1 − 3δ 2 (1 – 2α) 2 ≤ 0 or when<br />
δ 4 (1 – 2α) 4 + 4 δ 2 (1 – 2α) 2 – 1 ≤ 0 <strong>and</strong> α − 3δ 2 (1 – 2α) 2 − (1 – 2α) (1 – δ) ≥ 0.<br />
Proof: Suppose (ii) describes the pattern of equilibrium network evoluti<strong>on</strong> in the<br />
absence of innovati<strong>on</strong>. Then, p Z * (2) = p(2) <strong>and</strong> p Y * (1) = p(1):<br />
79
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 253<br />
M (1 2 )<br />
* (2) = δα δ − α<br />
−<br />
p(1)<br />
2<br />
1 − δ(1<br />
− 2α)<br />
1 − δ (1 − 2α)<br />
(7)<br />
p<br />
Z<br />
2<br />
2<br />
2<br />
M (1 2 )<br />
*(2) = δα δ − α<br />
+<br />
p(0) − δM<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
(8)<br />
pY<br />
2<br />
p<br />
Y<br />
2<br />
M (1 2 )<br />
*(1) = δα δ − α<br />
2<br />
+<br />
p(0) − δαM<br />
− δ(1−<br />
2α)p(2)<br />
δ (1−<br />
2α)M<br />
2<br />
2<br />
1−<br />
δ(1−<br />
2α)<br />
1−<br />
δ (1−<br />
2α)<br />
(9)<br />
2<br />
2<br />
M (1 2 )<br />
* (1) = δα δ − α<br />
−<br />
p(1) δαM<br />
2<br />
1 − δ(1<br />
− 2α)<br />
1 − δ (1 − 2α)<br />
(10)<br />
p<br />
Z<br />
+<br />
2<br />
p(0) = δ (1 – 2α) p(1) (11)<br />
Using Equati<strong>on</strong>s (7) <strong>and</strong> (11) to substitute in the values of p(2)<strong>and</strong> p(0) in Equati<strong>on</strong> (9),<br />
we get<br />
M<br />
p (1) =−<br />
(12)<br />
⎡<br />
2 2⎤<br />
2 2<br />
⎢ ⎛ ⎞<br />
1 12<br />
⎥ ⎛ ⎞ ⎛ ⎞<br />
⎢ −δ 12 1<br />
⎜<br />
− α⎟ ⎥δ ⎜<br />
− α⎟ ⎜<br />
−δ⎟<br />
⎢ ⎝ ⎠ ⎥ ⎝ ⎠ ⎝ ⎠<br />
⎣<br />
⎦<br />
⎡<br />
⎤<br />
⎡<br />
2⎤<br />
⎛ ⎞<br />
1 12<br />
⎢13 2⎛ ⎞<br />
⎢ −δ−α ⎥ 12<br />
⎥<br />
⎢ ⎜ ⎟ ⎥ ⎢ −δ ⎜<br />
−α ⎟<br />
⎣ ⎝ ⎠⎦⎢ ⎝ ⎠<br />
⎣<br />
⎥<br />
⎦<br />
For (ii) to occur in equilibrium, it must be true that p Z * (1) ≤ p(1) <strong>and</strong> p Y * (2) ≤ p(2). It can<br />
be checked that p Y * (2) ≤ p(2) iff 1 − 3δ 2 (1 – 2α) 2 ≤ 0, or δ 4 (1 – 2α) 4 + 4 δ 2 (1– 2α) 2 –<br />
1 ≤ 0. It can also be checked that p Z * (1) ≤ p(1) iff 1 − 3δ 2 (1 – 2α) 2 ≤ 0 or α − 3δ 2 α<br />
(1 – 2α) 2 − (1 – 2α) ( 1 – δ) ≥ 0, thus proving the lemma.<br />
Lemma 6<br />
For δ → 1 <strong>and</strong> α → 0 the equilibrium described in Lemma 5 is collusive.<br />
Proof: For δ → 1 <strong>and</strong> α → 0, 1 − 3δ 2 (1 – 2α) 2 < 0, thus case (ii) can occur in equilibrium<br />
from Lemma 5. But it can be checked from (12) that for 1 − 3δ 2 (1 – 2α) 2 ≤ 0 <strong>and</strong> δ < 1,<br />
p(1) is positive.<br />
Thus from Lemmas 1 to 6 we can infer that as δ → 1 <strong>and</strong> α → 0, the (almost) unique<br />
n<strong>on</strong>-collusive equilibrium shows the following properties. In the absence of innovati<strong>on</strong><br />
both firms charge the prices as below: As δ → 1 <strong>and</strong> α → 0, the equilibrium prices in the<br />
absence of innovati<strong>on</strong> are:<br />
p(2) = p(−2) = − δαM − δ 2 (1 − 2α) αM<br />
p(1) = p(−1) = − δ 2 (1 – 2α) αM − δ 3 (1 – 2α) 2 αM<br />
p(0) = δ 3 (1 – 2α) 2 M [2α + 2δα (1 – 2α) – 1].<br />
As α → 0, the equilibrium prices in the absence of innovati<strong>on</strong> are:<br />
p(2) = p(−2) → 0<br />
p(1) = p(−1) → 0<br />
p(0) → −δ 3 M.<br />
Firm Y sells to the c<strong>on</strong>sumer in the market at s = 1, 2 <strong>and</strong> firm Z sells when s = − 1, − 2.<br />
When s = 0, either firm could sell but both earn the same expected profits when s = 0,<br />
that is, x Y (1) = x Y (2) = x Z (−1) = x Z (−2) = 1 <strong>and</strong> x Y (0) = 1 or x Z (0) = 1.<br />
80
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
254 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Thus, in the limit as the probability of innovati<strong>on</strong> approaches zero, the equilibrium<br />
approaches the n<strong>on</strong>-collusive equilibrium without uncertainty.<br />
Propositi<strong>on</strong> 3 There exist δ * <strong>and</strong> α * such that for δ * < δ < 1 <strong>and</strong> α * < α < 1/2, the<br />
n<strong>on</strong>-collusive symmetric equilibrium prices are unique. The firm with network advantage<br />
of 2 <strong>and</strong> the firm with network advantage of 1 sell.<br />
Proof: For Lemmas 1, 2, 4 <strong>and</strong> 5, we infer that the <strong>on</strong>ly possible outcome in<br />
equilibrium has the firm with the advantage of 2 <strong>and</strong> the firm with the disadvantage of 1<br />
sell prior to an innovati<strong>on</strong> occurring. From Lemma 5 we see that for δ close to 1 <strong>and</strong> α<br />
close to 1/2, the equilibrium prices are negative. The equilibrium is, therefore, n<strong>on</strong>collusive.<br />
As δ → 1 <strong>and</strong> α → 1/2, the equilibrium prices are as follows:<br />
p(2) = p(−2) → − M/2<br />
p(1) = p(−1) → 0<br />
p(0) → 0.<br />
Propositi<strong>on</strong>s 2 <strong>and</strong> 3 illustrate the effect of uncertainty <strong>on</strong> equilibrium behaviour in a<br />
market with network externalities. In Propositi<strong>on</strong> 1 we saw that when the <strong>on</strong>ly difference<br />
between two firms is in the sizes of their networks of c<strong>on</strong>sumers, the firm with the bigger<br />
network sells to all c<strong>on</strong>sumers that subsequently enter the market.<br />
However, <strong>on</strong>ce we introduce uncertainty about the quality of a firm’s product this no<br />
l<strong>on</strong>ger is true. Even if the uncertainty faced by the firms is fully symmetric <strong>and</strong> the <strong>on</strong>ly<br />
difference between the firms is in the size of their respective networks, the firm with the<br />
bigger network does not always sell in equilibrium. For a sufficiently high success rate in<br />
innovati<strong>on</strong>, it is the firm with the smaller network that sells provided the network<br />
distance is not too large.<br />
The reas<strong>on</strong> for this strikingly different outcome compared to the situati<strong>on</strong> without<br />
uncertainty is as follows: When there is no uncertainty, the <strong>on</strong>ly way either firm can<br />
realise positive profits is by reaching its absorbing state. Outside of their respective<br />
absorbing states both firms make the same profits that is zero. Thus the way for either<br />
firm to make positive profits is to sell to all the c<strong>on</strong>sumers <strong>and</strong> reach its absorbing state.<br />
However, since the firm with the larger network needs a smaller additi<strong>on</strong> to its network to<br />
reach its absorbing state, it needs a shorter time span to reach its absorbing state, hence its<br />
discounted profits from selling to all c<strong>on</strong>sumers is larger. Not selling to any c<strong>on</strong>sumers,<br />
<strong>on</strong> the other h<strong>and</strong>, gives both firms zero profits. Thus, since the firm with the bigger<br />
network has more to gain from selling to all subsequent c<strong>on</strong>sumers while not selling<br />
gives it the same profits as the firm that is behind, it is willing to price more aggressively<br />
in order to lure the c<strong>on</strong>sumers. Thus, in equilibrium the firm that is ahead always sells.<br />
The introducti<strong>on</strong> of uncertainty, even when it is uniform across the firms, alters the<br />
incentives of the firms to fight for the c<strong>on</strong>sumers. When the difference in network (size)<br />
between the two firms is sufficiently small <strong>and</strong> the probability of success is high the<br />
presence of uncertainty reduces the incentive of the firm that is ahead to fight for the<br />
current c<strong>on</strong>sumer relative to that of the firm with the network disadvantage. That is<br />
because with a high rate of innovati<strong>on</strong> the leading firm has a high probability of getting<br />
its m<strong>on</strong>opoly profits whether it sells to the current c<strong>on</strong>sumer or not. The firm with the<br />
network disadvantage, <strong>on</strong> the other h<strong>and</strong>, has a bigger stake in winning the current<br />
c<strong>on</strong>sumer since by not letting the other firm get close to its absorbing state, it keeps its<br />
own chances of getting the innovati<strong>on</strong> <strong>and</strong> thus reaping the high m<strong>on</strong>opoly profits alive.<br />
81
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
Stochastic racing <strong>and</strong> competiti<strong>on</strong> in network markets 255<br />
However, when the leading firm has a big network advantage, that is, is close to its<br />
absorbing state, its incentive to sell is much higher since it can then guarantee itself the<br />
m<strong>on</strong>opoly profits by selling to the current customer. Thus as δ → 1 <strong>and</strong> α → 1/2, we see<br />
the firm with a network advantage of 2 selling <strong>and</strong> also the firm with a network<br />
disadvantage of 1.<br />
Until now all the discussi<strong>on</strong> has been in terms of network advantage or disadvantage.<br />
We could equally well have had the discussi<strong>on</strong> in terms of quality advantage or<br />
disadvantage since network <strong>and</strong> quality are perfect substitutes in a c<strong>on</strong>sumer’s utility.<br />
Thus without uncertainty, in the presence of network externalities, if the <strong>on</strong>ly difference<br />
between two firms was in the quality of their products then the firm with the better<br />
quality product would sell to all the c<strong>on</strong>sumers. With uncertain quality improvement<br />
however, if the probability of quality improvement were high enough then the firm with<br />
the lower quality would sell to the c<strong>on</strong>sumer. Both firms would remain in the market until<br />
<strong>on</strong>e of the firms got an innovati<strong>on</strong> that gave it an insurmountable lead over the other.<br />
6 Extensi<strong>on</strong>s <strong>and</strong> c<strong>on</strong>clusi<strong>on</strong>s<br />
We have looked at a simple model of a market with network externalities where there is<br />
technological innovati<strong>on</strong>. It would be interesting what kind of modificati<strong>on</strong>s to our basic<br />
model might change some of the results. The model is based <strong>on</strong> the assumpti<strong>on</strong> of<br />
symmetry in terms of the innovati<strong>on</strong> that the firms expect.<br />
If we introduce asymmetric innovati<strong>on</strong> this would permit us to endogenise innovati<strong>on</strong><br />
to the extent that we could look at the decisi<strong>on</strong>s of firms to improve quality. If we look at<br />
a more dynamic model of having sequential innovati<strong>on</strong> <strong>and</strong> multiple races<br />
(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> [2002b]) we might be able to see whether there is a clustering of innovati<strong>on</strong>s,<br />
<strong>and</strong> by examing the size of innovati<strong>on</strong>s, whether firms would rather tend toward <strong>on</strong>e big<br />
risky innovati<strong>on</strong> or rather for a series of small <strong>on</strong>es which are less risky. Under welfare<br />
theoretic c<strong>on</strong>siderati<strong>on</strong>s there may be related questi<strong>on</strong>s <strong>on</strong> whether there is too much or<br />
too little innovati<strong>on</strong> in view of a social optimum (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2003).<br />
Our simple model shows how competing firms behave in a market with network<br />
externalities before the resoluti<strong>on</strong> of technological uncertainty. If our firms are in waiting<br />
(for the breakthrough innovati<strong>on</strong> to come true) but the probability of innovati<strong>on</strong> is small,<br />
then the firm that has a network advantage makes all the sales. However, if the<br />
probability of innovati<strong>on</strong> is large, then it is not always the firm with the bigger network<br />
that sells. While a firm with a network advantage of 2 would make the sale, it is the firm<br />
with a network disadvantage of 1 that makes the sale if it is lucky to catch this<br />
innovati<strong>on</strong>.<br />
Alternatively, it could also serve as an explanati<strong>on</strong> for why a firm, that is in the end,<br />
driven out of the market may just hang <strong>on</strong> there for a while, with a network disadvantage,<br />
before it finally gives up <strong>and</strong> exits (or is exited from) the market.<br />
Furthermore, as some sort of by-product of this analysis, we could pursue the<br />
interacti<strong>on</strong> between network effects <strong>and</strong> innovati<strong>on</strong> for market structures. If network<br />
effects are ‘competitive barriers’ that could <strong>on</strong>ly be overcome by radical innovati<strong>on</strong> then,<br />
by implicati<strong>on</strong>, the synergy of network effects <strong>and</strong> radical innovati<strong>on</strong> through the<br />
incumbent would c<strong>on</strong>stitute a formidable powerhouse to c<strong>on</strong>solidate m<strong>on</strong>opolistic<br />
positi<strong>on</strong>s in dynamic product markets where ‘winner-takes-all markets’ are created.<br />
82
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
HANS W. GOTTINGER<br />
256 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Acknowledgments<br />
I want to say thank to A. Aykac, Sophia Antipolis, J. McGee, Warwick, I. Png,<br />
Singapore, for their helpful comments <strong>and</strong> suggesti<strong>on</strong>s.<br />
References<br />
Belsen, K. (2003) ‘S<strong>on</strong>y again turns to designs to lift electr<strong>on</strong>ics’, New York Times<br />
(Online Editi<strong>on</strong>), Feb. 02.<br />
Choi, J.P. (1994) ‘Irreversible choice of uncertain technologies with network externalities’, R<strong>and</strong><br />
Journal of Ec<strong>on</strong>omics, Vol. 25, pp.382–401.<br />
Clark, D. (2003) ‘Sun unveils blade servers in bid to counter IBM, H-P’, Wall Street Journal<br />
(Online Editi<strong>on</strong>), Feb.22.<br />
Evans, D.S. <strong>and</strong> Schmalensee, R. (2001) ‘Some ec<strong>on</strong>omic aspects of antitrust analysis in<br />
dynamically competitive industries’, Nati<strong>on</strong>al Bureau of Ec<strong>on</strong>omic Research, Cambridge,<br />
MA. Working Paper 8268, http://www.nber.org/papers/w8268.<br />
Fudenberg, D. <strong>and</strong> Tirole, J. (1993) Game Theory, MIT Press, Cambridge, MA.<br />
Futia, C. (1980) ‘Schumpeterian competiti<strong>on</strong>’, Quarterly Journal of Ec<strong>on</strong>omics, Vol. 94,<br />
pp.675–695.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W (2002b) Sequential Innovati<strong>on</strong>s <strong>and</strong> R&D Racing, http://www.csef.it/ publicati<strong>on</strong>s<br />
2001/.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W (2003) Ec<strong>on</strong>omies of Network Industries, Routledge, L<strong>on</strong>d<strong>on</strong>.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2002a) ‘Modeling stochastic innovati<strong>on</strong> races’, Technological Forecasting <strong>and</strong><br />
Social Change, Vol. 69, pp.607–624.<br />
Hamel, G. <strong>and</strong> Switzer, L. (2004) ‘The old guard vs. the vanguard’, The Wall Street Journal<br />
(Online Editi<strong>on</strong>), Feb. 23.<br />
Harris, C. <strong>and</strong> Vickers, J. (1987) ‘Racing with uncertainty’, Review of Ec<strong>on</strong>omic Studies, Vol. 54,<br />
pp.1–21.<br />
Katz, M. <strong>and</strong> Shapiro, C. (1992) ‘Product introducti<strong>on</strong> with network externalities’, The Journal of<br />
Industrial Ec<strong>on</strong>omics, Vol. 40, pp.55–83.<br />
Kristiansen, E.G. (1998) ‘R&D in the presence of network externalities: timing <strong>and</strong> compatibility’,<br />
R<strong>and</strong> Journal of Ec<strong>on</strong>omics, Vol. 29, pp.531–547.<br />
Lohr, S. (2003) ‘Competitors shape strategy to gain edge in web services’, New York Times (Online<br />
Editi<strong>on</strong>), Feb. 3.<br />
Markoff, J. <strong>and</strong> Lohr, S. (2003) ‘Advanced micro issues challenge to rivals’, New York Times<br />
(Online Editi<strong>on</strong>), April 19.<br />
Uttenback, J.M. (1994) ‘Mastering the dynamics of innovati<strong>on</strong>’, Harvard Business School Press,<br />
Bost<strong>on</strong>.<br />
83
HANS W. GOTTINGER<br />
1.5 STOCHASTIC RACING AND COMPETITION IN NETWORK MARKETS<br />
84
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
Int. J. Business <strong>and</strong> Systems Research, Vol. X, No. Y, XXXX 1<br />
High speed technology competiti<strong>on</strong><br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
IMS, University of Maastricht,<br />
Maastricht, The Netherl<strong>and</strong>s<br />
IIR, Hitotsubashi University,<br />
Tokyo, Japan<br />
E-mail: hg528@bingo-ev.de<br />
Abstract: We c<strong>on</strong>sider a technological race in which competitive firms engage<br />
in a class of differential games <strong>and</strong> some firms have priority of moves over<br />
others. The firm that has the right to move first is called the technological<br />
leader <strong>and</strong> the other competing firm is called the follower. An example of<br />
this type of sequential move game is the Stackelberg model of duopoly.<br />
The open-loop Nash equilibrium c<strong>on</strong>diti<strong>on</strong>s in a sequential move game is<br />
derived <strong>and</strong> the results for technological race problems are derived as is a<br />
comparis<strong>on</strong> of the strategies of leader <strong>and</strong> the follower.<br />
Keywords: industrial organisati<strong>on</strong>; dynamic competiti<strong>on</strong>; differential games;<br />
technological leadership.<br />
Reference to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (XXXX)<br />
‘High speed technology competiti<strong>on</strong>’, Int. J. Business <strong>and</strong> Systems Research,<br />
Vol. X, No. Y, pp.XXX–XXX.<br />
Biographical notes: <str<strong>on</strong>g>Hans</str<strong>on</strong>g>-Werner <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> has been the Director of the<br />
Institute of Management Science, Maastricht, NL <strong>and</strong> the Professor of<br />
Ec<strong>on</strong>omics in the Institute of Innovati<strong>on</strong> Research., Tokyo, Japan. He has<br />
internati<strong>on</strong>ally taught <strong>and</strong> widely published in the areas of industrial, energy<br />
<strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics <strong>and</strong> management. In these areas he has also been<br />
an advisor for internati<strong>on</strong>al organisati<strong>on</strong>s. He is the author for the recently<br />
published book titled Ec<strong>on</strong>omies of Network Industries (Routledge, 2003) <strong>and</strong><br />
Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong> (Routledge, 2006).<br />
1 Introducti<strong>on</strong><br />
In highly competitive technological industries new challenges <strong>and</strong> opportunities are<br />
arising in the new product development arena. Driven by global markets, global<br />
competiti<strong>on</strong>, the global dispersi<strong>on</strong> of scientific/engineering talent <strong>and</strong> the advent of new<br />
Informati<strong>on</strong> <strong>and</strong> Communicati<strong>on</strong> Technologies (ICT) a new visi<strong>on</strong> of product<br />
development is that of a highly disaggregated, distributive process with people <strong>and</strong><br />
organisati<strong>on</strong>s spread throughout the world. At the same time, products are becoming<br />
increasingly complex requiring numerous engineering decisi<strong>on</strong>s to bring them to market.<br />
Competitive pressures mean that ‘time to market’ has become a key to new product<br />
success. However, at the same time, it is important to keep the innovati<strong>on</strong> <strong>and</strong> quality<br />
dimensi<strong>on</strong>s of the new product at their optimal level.<br />
Copyright © XXXX Inderscience Enterprises Ltd.<br />
85
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
2 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
A central questi<strong>on</strong> to address is, how should firms invest in innovati<strong>on</strong> <strong>and</strong> what are<br />
the implicati<strong>on</strong>s of such investments for competitive advantage? Underst<strong>and</strong>ing why a<br />
firm benefits from investments in innovati<strong>on</strong> <strong>and</strong> quality illuminates issues of<br />
competitive strategy <strong>and</strong> industrial organisati<strong>on</strong>. In the field of competitive strategy,<br />
much attenti<strong>on</strong> has been devoted to the c<strong>on</strong>cept of core capabilities (Teece et al., 1997).<br />
Underst<strong>and</strong>ing how firms make optimal investments in the face of competiti<strong>on</strong> reveals<br />
the nature of competiti<strong>on</strong> <strong>and</strong> provides theoretical <strong>and</strong> managerial implicati<strong>on</strong>s for<br />
developing core competence <strong>and</strong> dynamic capabilities.<br />
In major parts of competitive analysis involving R&D decisi<strong>on</strong>s the focus is <strong>on</strong><br />
breakthrough innovati<strong>on</strong>s which could create entirely new markets, for example, in<br />
studies featuring patent racing between competing firms. In more comm<strong>on</strong> competitive<br />
situati<strong>on</strong>s we observe firms, however, competing by investing in incremental<br />
improvements of products. It is an important aspect when innovati<strong>on</strong> is c<strong>on</strong>sidered to be<br />
manifested in product quality, process improvements <strong>and</strong> in the overall quality culture of<br />
an organisati<strong>on</strong> (Kanter, 2006). For example, after product launch, incremental<br />
improvement of different aspects of product quality, improvements in various business<br />
processes <strong>and</strong> an incremental adopti<strong>on</strong> of a quality culture are quite real-world<br />
phenomena. Some firms operate in a simultaneous product launch situati<strong>on</strong> while others<br />
compete sequentially by adopting the role of leader or follower. The strategic<br />
implicati<strong>on</strong>s in these diverse circumstances can be treated within a unified framework of<br />
dynamic stochastic differential games.<br />
In recent years, with the emergence of e-business <strong>and</strong> a supply chain view for<br />
product development processes, multiple firms with varying <strong>and</strong> at times c<strong>on</strong>flicting<br />
objectives enter into collaborative arrangements. In such situati<strong>on</strong>s, the competitive<br />
strategy based <strong>on</strong> quality <strong>and</strong> innovati<strong>on</strong> could potentially permeate in those<br />
collaborative setups. A recent example is the collaborative venture of S<strong>on</strong>y <strong>and</strong> Samsung<br />
to build a cutting edge plant for LCD flat screen TV though displaying <strong>on</strong>going fierce<br />
rivalry in new product launches in exactly the same product categories. When innovati<strong>on</strong><br />
<strong>and</strong> quality levels form the core of a firm’s capabilities, each member in the supply chain<br />
would have an incentive to invest <strong>and</strong> improve their dynamic capabilities. This leads to<br />
tacit competiti<strong>on</strong> am<strong>on</strong>g collaborative product development partners by means of active<br />
investment in innovati<strong>on</strong> <strong>and</strong> quality.<br />
Although the forces of innovati<strong>on</strong> are central to competiti<strong>on</strong> to emerging technically<br />
dynamic industries, they also affect mature industries where life cycles, historically, were<br />
relatively str<strong>on</strong>g, technologies were mature <strong>and</strong> dem<strong>and</strong>s stable.<br />
A strategy for technology must c<strong>on</strong>fr<strong>on</strong>t primarily what the focus of technical<br />
development will be. The questi<strong>on</strong> is what technologies are critical to the firm’s<br />
competitive advantage. In this c<strong>on</strong>text, technology must include the ‘know-how’ the firm<br />
needs to create, produce, market its products <strong>and</strong> deliver them to customers. As a major<br />
step in creating a technology strategy it has to define those capabilities where the firm<br />
seeks to achieve a distinctive advantage relative to competitors. For most firms, there are<br />
a large number of important areas of technological ‘know-how’ but <strong>on</strong>ly a h<strong>and</strong>ful where<br />
the firm will seek to create truly superior capability.<br />
Having determined the focus of technical development <strong>and</strong> the source of capability,<br />
the firm must establish the timing <strong>and</strong> frequency for innovati<strong>on</strong> efforts. Part of the timing<br />
issue involves developing technical capabilities, <strong>and</strong> the rest involves introducing<br />
technology into the market. The frequency of implementati<strong>on</strong> <strong>and</strong> associated risks will<br />
depend in part <strong>on</strong> the nature of the technology <strong>and</strong> the markets involved (e.g. disk drive<br />
86
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 3<br />
versus automotive technology), but in part <strong>on</strong> strategic choice. At the extreme, a firm<br />
may adopt a rapid incremental strategy, that is, frequent, small changes in technology<br />
that cumulatively lead to c<strong>on</strong>tinuous performance improvement. The polar opposite<br />
might be termed the great leap forward strategy. In this approach, a firm chooses to make<br />
infrequent but large-scale changes in technology that substantially advance the state of<br />
the art (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006, Chap. 6).<br />
As an example of the importance of innovati<strong>on</strong> strategy in product development, we<br />
notice that IBM created <strong>and</strong> c<strong>on</strong>tinues to dominate the mainframe segment, but it missed<br />
by many years, the emergence of the minicomputer architecture <strong>and</strong> market.<br />
The minicomputer was developed <strong>and</strong> its market applicati<strong>on</strong>s exploited by firms such as<br />
Digital Equipment Corporati<strong>on</strong> (DEC) <strong>and</strong> Data General (DG) back in the 1970s.<br />
Summarising, in a mainstream scenario, high speed technological competiti<strong>on</strong> is<br />
likely to be driven by<br />
1 companies competing in fast changing, diverse networks <strong>and</strong><br />
2 companies c<strong>on</strong>fr<strong>on</strong>ted with ever shortening product/technology life cycles<br />
because of multiple interacti<strong>on</strong>s between current <strong>and</strong> emerging technologies <strong>and</strong><br />
product diversity or what may be circumscribed as ‘combinatorial innovati<strong>on</strong>’<br />
(Varian et al., 2004).<br />
A clear example provided by Arthur (2000, p.3):<br />
“If you look at genomics it is certainly heavily based <strong>on</strong> biology, but it is also<br />
highly dependent <strong>on</strong> digital computing!<br />
Genomics, then, is as much a manifestati<strong>on</strong> of digital computing as it is a<br />
manifestati<strong>on</strong> of molecular biology. What happens is that some of the new<br />
technologies become base technologies, <strong>and</strong> these give rise to manifestati<strong>on</strong>s,<br />
some of which themselves become base technologies. Manfestati<strong>on</strong>s then<br />
appear at an even faster rate.”<br />
To represent the features of leader-follower type in a sequence of technological racing<br />
we c<strong>on</strong>sider a class of differential games in which some firms have priority of moves<br />
over others.<br />
The firm that has the right to move first is called the leader <strong>and</strong> the other competing<br />
firm is called the follower. A well-known example of this type of sequential move game<br />
is the Stackelberg model of duopoly. In this type of interacti<strong>on</strong> the open-loop Nash<br />
equilibrium c<strong>on</strong>diti<strong>on</strong>s in a sequential move game can be derived. It would lead to a<br />
comparis<strong>on</strong> of the strategies of leader <strong>and</strong> the follower.<br />
Secti<strong>on</strong> 2 presents the essence of leader-follower type interacti<strong>on</strong>s through the<br />
format of a differential game which is accommodated to a competitive market<br />
situati<strong>on</strong> in Secti<strong>on</strong> 3. The properties of their interacti<strong>on</strong>s are derived in Secti<strong>on</strong> 4.<br />
Secti<strong>on</strong> 5 h<strong>and</strong>les various market asymmetries in the leader follower type situati<strong>on</strong>.<br />
Finally, Secti<strong>on</strong> 6 summarises <strong>and</strong> draws up c<strong>on</strong>clusi<strong>on</strong>s <strong>and</strong> directi<strong>on</strong>s for future<br />
research.<br />
2 A differential game formulati<strong>on</strong><br />
For our modelling effort we follow the notati<strong>on</strong>s <strong>and</strong> symbols as explained below.<br />
N: number of firms, T: finite time horiz<strong>on</strong> for the strategies, i, j: superscripts to<br />
denote competing firms in a duopoly, t: an instant of time in the dynamic game setup,<br />
87
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
4 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
u(t): investments in innovati<strong>on</strong> effort (expenditure per unit time), R(t): net revenue rate<br />
for the firm at t, R o<br />
: product category net revenue rate for the existing product,<br />
R 1<br />
: product category net revenue rate for the new product, x o<br />
: quality level of the existing<br />
product.<br />
Several examples illustrate the importance of innovati<strong>on</strong> <strong>and</strong> quality as<br />
competitive weap<strong>on</strong>s. In the computer mainframe arena IBM replaced Remingt<strong>on</strong><br />
R<strong>and</strong> <strong>and</strong> Sperry Univac as market leader in the 1950s, <strong>and</strong> its subsequent growth<br />
outperformed its competitors so rapidly that in 1963 its data process revenues were<br />
four times larger than the combined revenues of its eight main rivals in the US market<br />
(Hoffman, 1976).<br />
In the model, we assume that enhancements of quality are achieved by climbing a<br />
performance ladder that may squarely embrace technology but could exp<strong>and</strong> to other<br />
criteria uniquely identified with quality. Let the quality of the product at time t be x(t).<br />
In the c<strong>on</strong>text of total quality management, as quality levels increase it becomes even<br />
more difficult to climb the performance ladder. The hypothesis behind the formulati<strong>on</strong> is<br />
that a firm needs to make higher innovati<strong>on</strong> investments targeted towards the<br />
improvement of quality. To capture this dynamics a negative feedback effect of the<br />
present state quality <strong>on</strong> the rate of change of product quality (x " (t)) is c<strong>on</strong>sidered.<br />
The state dynamics is<br />
"<br />
#<br />
x ! t" $ K# u! t "$ % %x! t"<br />
(1)<br />
where K is proporti<strong>on</strong>al to the level of capital investment in development technology <strong>and</strong><br />
L is the proporti<strong>on</strong>ality c<strong>on</strong>stant for the influence of present quality <strong>on</strong> the speed of<br />
further quality improvements, # is the innovati<strong>on</strong> resource productivity parameter.<br />
Based <strong>on</strong> (1) the quality of the product at time t is<br />
t<br />
#<br />
x! t" $ xo<br />
*<br />
&<br />
( K# u! s "$ % %x! s" '<br />
) & s<br />
(2)<br />
%<br />
The product quality provides a means for evaluating the product’s attractiveness in the<br />
market in the presence of other competing products. The firm’s market share is a<br />
functi<strong>on</strong> of both its own product quality <strong>and</strong> the product quality of rivals (the initial<br />
product quality x o<br />
may reflect the reputati<strong>on</strong>al capital of the firm, either low or high, <strong>and</strong><br />
may induce her to different innovati<strong>on</strong> investment efforts).<br />
The difference lies in the planning horiz<strong>on</strong>. In this competitive setup, the<br />
planning horiz<strong>on</strong> extends bey<strong>on</strong>d the date of launch <strong>and</strong> the competing firms c<strong>on</strong>tinue<br />
investing in product-process innovati<strong>on</strong> until the end of the growth phase of a product.<br />
Further,<br />
+ x%<br />
R%<br />
( % , t - ,<br />
i *<br />
p<br />
. x% / x%<br />
R! t"<br />
$ 0<br />
i<br />
. x !,<br />
"<br />
R<br />
'<br />
( ,<br />
i *<br />
p<br />
, t - ,<br />
.<br />
1 x !, " / x !,<br />
"<br />
The cumulative development cost of the new product at time t is given as:<br />
T*! t"<br />
$ * ,<br />
+ u! s",&<br />
s (4)<br />
%<br />
(3)<br />
88
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 5<br />
the firm’s cumulative profit at time t is determined as follows,<br />
, 2 ! t" $ TR! t" % T*! t"<br />
(5)<br />
where TR(t) <strong>and</strong> TC (T) are total revenues <strong>and</strong> costs at time t, respectively. The total<br />
revenue functi<strong>on</strong> is given by:<br />
TR! t"<br />
$ * ,<br />
R! s"&<br />
s (6)<br />
%<br />
where R(3) is given in (3). The firm’s decisi<strong>on</strong> set is 4 $ {u(t)}. Notice that the firm<br />
precommits <strong>on</strong> the date of product launch T p<br />
. The cumulative profit functi<strong>on</strong>, T 2(5), is<br />
defined as the total profit by end of the window of opportunity with decisi<strong>on</strong> 564.<br />
The firm’s decisi<strong>on</strong> problem can be stated as,<br />
7 9 8 7 9 8 9 7 9<br />
8<br />
./x<br />
5 64<br />
, 2 ! 5 " $ TR 5 % T* 5 $ , 2 5<br />
(7)<br />
The combinati<strong>on</strong> of Equati<strong>on</strong>s (1)–(6) generates an explicit representati<strong>on</strong> of firm i’s<br />
cumulative profit by the end of time horiz<strong>on</strong>. This substituti<strong>on</strong> yields:<br />
& x%<br />
, 2 9 75<br />
9<br />
8 $ ./x5<br />
64 : R<br />
%<br />
,<br />
i *<br />
( x% / x%<br />
,<br />
#<br />
xo<br />
/ * K<br />
%<br />
'# u! s "$ % %x! s"&<br />
s<br />
/<br />
'<br />
%<br />
,<br />
#<br />
*<br />
xo<br />
/ * K# u! s "$ % %x! s"& s / x !,<br />
"<br />
%<br />
%<br />
*<br />
,<br />
%<br />
+ u! s",&<br />
s'<br />
;)<br />
p<br />
7 p 8<br />
R , ,<br />
(8)<br />
*<br />
where x<br />
%<br />
<strong>and</strong> x j<br />
(T), respectively are the quality of existing <strong>and</strong> new products of the<br />
competitor. C<strong>on</strong>sidering a duopoly, the optimisati<strong>on</strong> problem written above can be<br />
reformulated as a differential game problem with state variable for the firm i<br />
(the competing firm is represented by superscript j) given as x i (t); the c<strong>on</strong>trol variable<br />
u i (t). In the terminology used in optimal c<strong>on</strong>trol <strong>and</strong> differential games, the salvage term<br />
for firm i, < i (T, x(T)) is defined as follows:<br />
i<br />
i<br />
x%<br />
< !, ( x !, ""141R%<br />
,<br />
i *<br />
x / x<br />
/ R<br />
'<br />
i<br />
p<br />
% %<br />
i<br />
,<br />
i i<br />
#<br />
i<br />
= x% / * K &<br />
'<br />
! "'<br />
%<br />
'<br />
! "& >7 % 8<br />
% ( u s ) % x s s , ,<br />
p<br />
i<br />
,<br />
i i<br />
#<br />
i *<br />
& '<br />
%<br />
/ *%<br />
' ( ) %<br />
'<br />
/<br />
x K u ! s " % x ! s"& s x !,<br />
"<br />
(9)<br />
where<br />
i<br />
,<br />
* *<br />
! "<br />
& *<br />
#<br />
*<br />
x , $ x &<br />
%<br />
/<br />
%<br />
2<br />
! "'<br />
%<br />
2<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s<br />
;<br />
& s (10)<br />
( )<br />
89
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
6 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
3 A sequential differential game between two competing firms<br />
We extend the c<strong>on</strong>ceptualisati<strong>on</strong> of a hyper-competitive scenario by c<strong>on</strong>sidering a<br />
sequential differential game as being representative of a leader-follower or<br />
incumbent-entrant competitive situati<strong>on</strong>. The game is characterised by informati<strong>on</strong><br />
asymmetry where the follower is aware of the innovati<strong>on</strong> <strong>and</strong> quality levels of the<br />
leader’s products. The motivati<strong>on</strong> for c<strong>on</strong>sidering this scenario is its close<br />
corresp<strong>on</strong>dence with many real life competitive cases. Knowing the investment strategy<br />
of the leader, the rival firms can formulate their own strategies. Therefore, the firm acting<br />
as a leader chooses a decisi<strong>on</strong> path that maximises the objective for all c<strong>on</strong>ceivable<br />
resp<strong>on</strong>ses that can be taken by the follower(s). In the case of sequential games a<br />
hierarchical play differential game approach is used to model the competitive situati<strong>on</strong><br />
<strong>and</strong> to obtain the open-loop Stackelberg Nash equilibrium. The issue of subgame<br />
perfectness <strong>and</strong> commitment is extremely important in these soluti<strong>on</strong>s. Adding to the<br />
previous notati<strong>on</strong> we let T p<br />
, date of product launch by leader, T p<br />
/ ?, date of product<br />
launch by follower. The leader is represented by the superscript i <strong>and</strong> the follower is<br />
represented by the superscript j. Since it is a sequential game the time of product launch<br />
for the two players is such that the leader launches the product at time T p<br />
. Later, the<br />
follower launches the product after time ? at T p<br />
/ ?. A c<strong>on</strong>tinuous improvement in the<br />
product is c<strong>on</strong>sidered <strong>and</strong> therefore the entire time interval t 6 @0, TA for the leader <strong>and</strong><br />
t 6 @T p<br />
B TA for the follower needs to be optimised. The state dynamics of the leader is<br />
" i i<br />
#<br />
i<br />
$ & '<br />
'<br />
%<br />
'<br />
x ! t" K ( u ! t ") % x ! t"<br />
(11)<br />
The rate of quality improvement x "i (t) increases with investments u i (t). The factor L 1<br />
x i (t)<br />
suggests a natural decay in quality in the absence of any investments. The differential<br />
game formulati<strong>on</strong> for leader is:<br />
i i i<br />
7 p 8 = 7 8 7 8><br />
* ,<br />
%<br />
./x , C u ! t "(, $ % u ! s"& s / < ,( x !, "<br />
(12)<br />
where < i (T, x i (T)) is defined as<br />
s. t.<br />
i i i<br />
i<br />
R% x ,<br />
% p<br />
/ R'<br />
x !,<br />
"! " '<br />
i * i * i i<br />
%<br />
/<br />
%<br />
/<br />
%<br />
/<br />
7 p 8<br />
? R x !, " , %,<br />
%?<br />
i i<br />
< 7, ( x !,<br />
" 8 D /<br />
x x x !, " x x !, " x !,<br />
"<br />
3<br />
i<br />
#<br />
i i i i<br />
x t $ K &<br />
'<br />
u t ' % %' x t x $ x<br />
%<br />
, x ,<br />
(13)<br />
! " ( ! ") ! "( !%" ( 4ixe&( ! "47ee<br />
(14)<br />
Next, the follower’s problem formulati<strong>on</strong> is discussed. Owing to the sequential nature of<br />
the game, the informati<strong>on</strong> about the leader’s quality <strong>and</strong> about the leader’s investments is<br />
known to the follower. Specifically, it is assumed that the knowledge gained by the<br />
leader’s investments ‘spills over’ to the follower’s quality improvement dynamics. It is<br />
plausible to assume that in the c<strong>on</strong>text of a ‘leader-follower’ competiti<strong>on</strong>, the level of<br />
quality improvements of the follower indeed depends not <strong>on</strong>ly <strong>on</strong> its own innovati<strong>on</strong><br />
efforts but also <strong>on</strong> the knowledge pool available because of the leader’s investments.<br />
The fact that the leader often cannot wholly c<strong>on</strong>ceal his efforts or can he credibly<br />
announce the commitment he has made, make the situati<strong>on</strong> quite complicated. To address<br />
these issues fully in this c<strong>on</strong>text it would require a subtle <strong>and</strong> rich analysis of games with<br />
incomplete informati<strong>on</strong>s (Robs<strong>on</strong>, 1990).<br />
90
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 7<br />
A formal treatment of ‘spillover effects’ would enrich this model. The spillover<br />
effect is modelled by c<strong>on</strong>sidering a linear additive term M 2<br />
u i (t). The follower’s state<br />
dynamics is<br />
3 * *<br />
#<br />
* i<br />
$ & '<br />
2<br />
%<br />
2<br />
/<br />
2<br />
x ! t" K ( u ! t ") % x ! t" - u ! t"<br />
(15)<br />
It is assumed that the spillover is a functi<strong>on</strong> of the investments made by the leader u i (t).<br />
In this term M 2<br />
is a very small number which quantifies the amount of spillover from the<br />
leader to the follower. The term M 2<br />
u i (t) would restrict the analysis to the case where the<br />
follower emulates the innovati<strong>on</strong> <strong>and</strong> product quality of the leader, as opposed to setting<br />
his own technology <strong>and</strong> quality st<strong>and</strong>ards. Now with all c<strong>on</strong>siderati<strong>on</strong>s wrapped up, a<br />
leader-follower dynamics can be stated.<br />
The follower’s differential game formulati<strong>on</strong> is given as:<br />
,<br />
7 p 8 * = > 7 8<br />
, p<br />
, %,<br />
p * * *<br />
$ %*<br />
= u 7s / ,<br />
p 8> & s / < 7, ( x !,<br />
" 8<br />
* * * * *<br />
./x , 2 u ! t "(, $ % u ! s" & s / < , ( x !,<br />
"<br />
where < j (T, x (T)) is defined as,<br />
s.t.<br />
*<br />
x<br />
< !, ( x !, "" 4R %<br />
,<br />
x<br />
*<br />
%<br />
i *<br />
%<br />
/ x%<br />
p<br />
%<br />
*<br />
= x !, " 7, %,<br />
p<br />
%?<br />
8><br />
x<br />
/ R<br />
/<br />
x !, " x x !, " x !,<br />
"<br />
*<br />
% i<br />
%<br />
? R<br />
* '<br />
/<br />
%<br />
i<br />
/<br />
*<br />
* *<br />
#<br />
* i<br />
$ & '<br />
2<br />
%<br />
2<br />
/<br />
2<br />
x ! t" K ( u ! t ") % x ! t" - u ! t"<br />
(18)<br />
where M 2<br />
u i<br />
(t) represent the spillover of knowledge, assumed to be a functi<strong>on</strong> of<br />
investments by the leader, M 2<br />
is a small c<strong>on</strong>stant such that 0 - M 2<br />
-- 1.<br />
where<br />
Furthermore, x<br />
* !%" $ x<br />
*<br />
%<br />
( T fixed, x j (T) free<br />
i<br />
t<br />
i i<br />
! "<br />
& i<br />
#<br />
i<br />
x , $ x &<br />
%<br />
/<br />
%<br />
'<br />
! "'<br />
%<br />
'<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s<br />
;<br />
& s<br />
( )<br />
*<br />
t<br />
* *<br />
! "<br />
& *<br />
#<br />
* i<br />
x , $ x &<br />
%<br />
/<br />
2<br />
! "'<br />
%<br />
2<br />
! " /<br />
2<br />
! "<br />
'<br />
* :<br />
K ( u s ) % x s - u s<br />
;<br />
& s<br />
, p ( )<br />
(16)<br />
(17)<br />
4 Analysis of the model <strong>and</strong> discussi<strong>on</strong><br />
A c<strong>on</strong>venti<strong>on</strong>al tool for solving the problem is the applicati<strong>on</strong> of P<strong>on</strong>tryagin’s maximum<br />
principle for open-loop Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s. Interested readers can refer<br />
Dockner et al. (2000) for details regarding P<strong>on</strong>tryagin’s approach to solving sequential<br />
91
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
8 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
differential game problems. The equilibrium results are expressed in the form of<br />
properties as given below.<br />
4.1 Equilibrium results<br />
Property 1: The maximised costate variables for the follower are functi<strong>on</strong> of time <strong>and</strong> are<br />
given by,<br />
E<br />
9 $ E e<br />
(19)<br />
'<br />
E<br />
% 2t<br />
%'<br />
E e<br />
9 '<br />
$ (20)<br />
2 %2<br />
% t<br />
where<br />
E 9 '<br />
<strong>and</strong> E 9<br />
2<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in follower firm’s own state <strong>and</strong> the state of the leader i, E 1<br />
(0) $ E 01<br />
, E 2<br />
(0) $ E 02<br />
are known c<strong>on</strong>stants.<br />
Property 2: The Stackelberg equilibrium investment by the follower in product<br />
development is given as:<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ ( K e ) (21)<br />
Next the leader’s problem is investigated. The leader knows the follower’s best resp<strong>on</strong>se<br />
to each c<strong>on</strong>trol path u i (3).<br />
Property 3: The maximised costate variables for the leader are given by,<br />
F<br />
F<br />
F e<br />
9 %'<br />
t<br />
'<br />
$<br />
%'<br />
(22)<br />
9 % 2t<br />
2<br />
$ F<br />
%2e<br />
(23)<br />
F F e<br />
%<br />
9 $ % 2t<br />
(24)<br />
9 %9<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the follower.<br />
F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Property 4: The Stackelberg equilibrium investment by the leader in product<br />
development is given as:<br />
i<br />
u ! t"<br />
t<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
Property 5: The equilibrium state trajectory of performance improvement of the follower<br />
is given as:<br />
(25)<br />
&<br />
x t % x e -<br />
i<br />
'8!'%#<br />
"<br />
%% 2t<br />
i %'<br />
t<br />
* e<br />
&<br />
2<br />
'# F '<br />
%'<br />
! " :<br />
*<br />
% t K e<br />
$<br />
2 %<br />
/ 7% '/ 8 2 : % ;<br />
2t<br />
%<br />
2<br />
: ('%<br />
-<br />
2<br />
F<br />
%2e<br />
)<br />
(<br />
% 2t * % 2t<br />
7 ' e 8 K2 7K2 # E%'<br />
e 8<br />
/ % /<br />
* * 2<br />
8!' % "<br />
# #<br />
'<br />
;)<br />
(26)<br />
92
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 9<br />
Property 6: The equilibrium state trajectory of performance improvement of the leader is<br />
given as:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
%% t<br />
i e<br />
&<br />
i<br />
%'<br />
i<br />
& K'<br />
# F<br />
%'<br />
e '<br />
'<br />
x ! t " $ :%' x% / 7% '/ e 8 K'<br />
: % 2t<br />
;<br />
;<br />
%' : ('%<br />
-<br />
2<br />
F<br />
%2e<br />
) ;<br />
( )<br />
For the analysis, the leader <strong>and</strong> the follower are assumed to be symmetric <strong>and</strong> therefore<br />
the corresp<strong>on</strong>ding values for the follower j are also assumed to be the same.<br />
The first observati<strong>on</strong> is that the costate variable of both the leader <strong>and</strong> the follower<br />
increases more rapidly with time<br />
The marginal utility of a unit increase in quality, exhibits a c<strong>on</strong>vex increasing trajectory.<br />
With comparable values of parameters the plot of the costate trajectory of the follower<br />
has an upward exp<strong>on</strong>ential drift with time. The plot begins at time t $ T r<br />
, since the<br />
follower initiates product development activities <strong>on</strong>ly after the leader has already<br />
launched the product. Regarding the costate trajectory of the leader we see that F 1<br />
represents marginal utility from a unit increase in quality of the leader’s product.<br />
The costate variable for both the leader <strong>and</strong> the follower firm increase for the<br />
entire planning horiz<strong>on</strong>. This follows since the leader <strong>and</strong> follower are competing <strong>on</strong> the<br />
basis of their quality levels. Firms increase their market shares based <strong>on</strong> their<br />
relative quality to that of the competitor’s. After launch the product faces its introducti<strong>on</strong><br />
<strong>and</strong> growth stage.<br />
In these phases, maintaining higher quality becomes even more important since the<br />
product sales are directly influenced by product quality. This leads to a c<strong>on</strong>vex<br />
increasing trajectory for the costate variable. Thus, as a sec<strong>on</strong>d observati<strong>on</strong>.<br />
The investment in innovati<strong>on</strong> made by the follower increases more rapidly with time<br />
The investment strategies of the follower are purely a functi<strong>on</strong> of its own costate<br />
variable. The follower’s costate variable increases in a c<strong>on</strong>vex fashi<strong>on</strong>. This in turn<br />
results in a c<strong>on</strong>vex increase in follower’s c<strong>on</strong>trol trajectory. With identical parameter<br />
values we can derive that the follower compensates for a delayed entry into the market<br />
by increasing its investment intensity. Such an increase in investment results in increased<br />
quality <strong>and</strong> therefore high total revenue for the follower. Thus we observe.<br />
The investment in innovati<strong>on</strong> made by the leader initially increases rapidly with time but<br />
later increases at a decreasing rate<br />
The leader’s investment trajectory is sigmoidal or S-shaped. Because of the nature of the<br />
game, the leader enters the market before the follower. While formulating his<br />
equilibrium investment strategy, the leader takes into account the evoluti<strong>on</strong> of his own<br />
costate <strong>and</strong> also the costate of the follower.<br />
Being early in the market, the leader takes into account all possible courses of acti<strong>on</strong><br />
that a follower may choose.<br />
We note that the leader’s rate of investment increases in the initial phase <strong>and</strong> then<br />
starts to decrease. That is, the leader capitalises <strong>on</strong> the advantage of early market entry by<br />
increasing the intensity of investments <strong>and</strong> gaining a high market share. Subsequently,<br />
after the follower’s entry, the leader reduces the intensity of investments <strong>and</strong> thereby<br />
reduces the amount of spillover that is potentially possible.<br />
(27)<br />
93
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
10 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
When the parameter values of the leader <strong>and</strong> follower are identical, the follower invests<br />
higher than the leader<br />
In the course of time, initially the follower maps its investment strategy to that of the<br />
leader. However, subsequently the leader starts reducing the rate of investments, while<br />
the follower c<strong>on</strong>tinues with the high investment rate based strategy.<br />
As a further observati<strong>on</strong>, the rate of increase of the follower’s product quality increases<br />
with time<br />
The quality of follower’s product increases for the entire time-horiz<strong>on</strong>. Moreover,<br />
this increase is a c<strong>on</strong>vex functi<strong>on</strong>. As can be inferred from the problem formulati<strong>on</strong>, the<br />
quality level of the follower is a functi<strong>on</strong> of investments made by the leader <strong>and</strong> the<br />
follower. Specifically, the quality trajectory is influenced by both the c<strong>on</strong>vex profile of<br />
investments made by the follower <strong>and</strong> also .<br />
by the spillover effect of the investments<br />
made by the leader. This results in a c<strong>on</strong>vex increase in quality improvements. The state<br />
trajectory is therefore upward sloping in c<strong>on</strong>vex form.<br />
The follower starts accruing revenue <strong>on</strong>ly after T p<br />
/ ?, where T p<br />
is the date of launch<br />
of the leader. In the game c<strong>on</strong>text of quality-based competiti<strong>on</strong>, the follower<br />
compensates for the delayed entry by increasing the quality levels at a fast rate.<br />
The rate of increase of the leader’s product quality decreases with time<br />
The leader’s investment trajectory follows an S-shaped trajectory. In the problem<br />
formulati<strong>on</strong>, the leader's quality improvement is <strong>on</strong>ly a functi<strong>on</strong> of its own investments.<br />
Moreover, it is assumed that the leader doesn’t obtain the advantages of spillover of<br />
knowledge gained by the follower’s investments. Furthermore, the leader needs to<br />
deliberately avoid maintaining very high investments to ensure that the follower<br />
doesn’t achieve huge gains from spillovers. Under such a setup it must be noted<br />
that the leader chooses an S-shaped trajectory for investments. In such an investment<br />
profile the investment increases rapidly in the initial phase but then eventually<br />
tapers off. Corresp<strong>on</strong>ding to this investment profile, the rate of quality improvement<br />
is high in the initial phase but then eventually the rate of improvement starts<br />
decreasing. Thus, the quality trajectory of the leader is c<strong>on</strong>cave. This leads to another<br />
observati<strong>on</strong>.<br />
When the parameter values of the leader <strong>and</strong> follower are identical, the quality of the<br />
follower’s product is higher than that of the leader’s product<br />
With identical parameter values the follower <strong>and</strong> leader are perfectly symmetric in their<br />
capabilities. Moreover, the leader has the advantage of earlier market entry, whereas the<br />
follower obtains the gains of informati<strong>on</strong> asymmetry <strong>and</strong> the associated knowledge<br />
spillovers from the leader. It is reas<strong>on</strong>able to assume that the follower employs his<br />
inherent competence <strong>and</strong> the benefit of spillover of knowledge from the leader’s<br />
investments to increase quality at a much faster rate. Under such circumstances the<br />
overall gain <strong>and</strong> loss in the competitive game is dictated by the relative revenues<br />
achieved <strong>and</strong> the costs incurred by the two players. With higher investments than the<br />
leader the follower incurs higher costs. At the same time these investments also lead to<br />
higher quality levels <strong>and</strong> therefore revenues for the follower. In the event of the value<br />
of R 1<br />
being very high the follower wins the game while the results favour the leader if<br />
R 1<br />
is low.<br />
94
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 11<br />
5 Firm asymmetries<br />
We now turn to firm asymmetries <strong>and</strong> their implicati<strong>on</strong>s. It is likely that firms take up the<br />
role of leader <strong>and</strong> follower based <strong>on</strong> inherent strengths <strong>and</strong> weaknesses. Therefore, an<br />
explicit c<strong>on</strong>siderati<strong>on</strong> of firm asymmetries is very important <strong>and</strong> could potentially lend<br />
more insights in a sequential game setup. Asymmetries may be addressed in at least four<br />
ways: Firstly, it is possible that organisati<strong>on</strong>al approaches <strong>and</strong> techniques (such as total<br />
quality management) can be used to make the product development process more<br />
cost-efficient <strong>and</strong> effective. This would influence the value of the resource productivity<br />
parameter #. Sec<strong>on</strong>dly, it is possible to have an advantage in product development by<br />
making capital investment in technology development. This is equivalent to c<strong>on</strong>sidering<br />
asymmetries in the value of K. Thirdly, it is possible that the obsolescence parameter<br />
of firms denoted by L is different. Finally, different values of M suggest different<br />
levels of spillover of knowledge from the leader’s investments to the follower. The<br />
firm with higher value of L has a higher obsolescence or decay in quality. We<br />
restrict assessment of asymmetries by c<strong>on</strong>sidering different values for the parameters:<br />
(# i <strong>and</strong> # j ), (K 1<br />
<strong>and</strong> K 2<br />
), (L 1<br />
<strong>and</strong> L 2<br />
) <strong>and</strong> (M l<br />
<strong>and</strong> M 2<br />
).<br />
5.1 Innovati<strong>on</strong> resource productivity<br />
Parameters # i <strong>and</strong> # j denote the innovati<strong>on</strong> resource productivity parameter of the two<br />
firms. An asymmetry could result if the competing firms have differing capabilities in<br />
making a productive use of investments. The skill set of employees, training <strong>and</strong><br />
development activities, the quality culture are some of the reas<strong>on</strong>s for such an<br />
asymmetry. For example, (# i $ 0.2) G (# j $ 0.1) suggest that firm i has a higher<br />
innovati<strong>on</strong> productivity over j. An investigati<strong>on</strong> of different values of # suggest that the<br />
firm with higher innovati<strong>on</strong> productivity also invests a higher amount in product<br />
development. From the results we observe that the follower has a relatively higher level<br />
of investments in innovati<strong>on</strong> as compared to the leader.<br />
Therefore, an increase in the value of # of the follower will <strong>on</strong>ly result in making<br />
these investments still higher. Instead, an increase in leader’s # would provide some<br />
interesting insights in that it will show that the leader sails ahead with the follower<br />
having no chance in catching up. It clearly shows that the leader has a higher investment<br />
than that of the follower for almost the entire planning horiz<strong>on</strong>. As can also be noted, the<br />
quality of the leader’s product is always higher than that of the follower. Hence, an<br />
asymmetry in terms of resource productivity parameters provides a totally different result<br />
than what was observed when the competing firms were symmetric. It suggests that the<br />
leader with an advantage in terms of early market entry as well as a higher level of<br />
resource productivity indeed c<strong>on</strong>tinues having higher revenues. Under this situati<strong>on</strong> the<br />
leader doesn’t worry much about the spillover to the follower since its own resource<br />
productivity enables attaining higher revenues by increasing quality levels relative to that<br />
of the follower. With a high value of R 1<br />
it can be c<strong>on</strong>jectured that the result favours<br />
the leader.<br />
5.2 Capital investment parameters<br />
Firm asymmetries can be analysed by evaluating different values for K <strong>and</strong> the<br />
implicati<strong>on</strong>s <strong>on</strong> c<strong>on</strong>trol <strong>and</strong> state trajectories. K l<br />
<strong>and</strong> K 2<br />
are proporti<strong>on</strong>al to the level of<br />
95
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
12 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
capital investment in development technology by the two competing firms. If<br />
(K 1<br />
$ 20) G (K 2<br />
$ 10), it suggests that the leader i has higher levels of capital<br />
investment in development technology as compared to the follower j. The implicati<strong>on</strong>s of<br />
a higher value of K for the leader can easily be derived. The parameter K exerts a<br />
positive effect <strong>on</strong> innovati<strong>on</strong> investments. Similar to the c<strong>on</strong>jecture regarding different<br />
values of #, the intuiti<strong>on</strong> behind this effect is that with a higher level of capital<br />
investments in development technology, a firm targets the investments towards<br />
increasing the product quality. Thus with a higher value of K the investment by the<br />
leader is relatively higher than that of the follower for most part of the planning horiz<strong>on</strong>.<br />
However, in the latter part of the planning horiz<strong>on</strong>, the follower’s investment in fact<br />
shoots up while that of the leader tapers off. Such an investment profile would have an<br />
impact <strong>on</strong> the state trajectory. Furthermore, the product quality of the leader is higher<br />
than that of the follower for most part of the planning horiz<strong>on</strong>. Eventually, at the end<br />
of the planning horiz<strong>on</strong>, the high growth in the lower’s investment results in higher<br />
quality than that of the leader. In the model, the relative difference in resources due to<br />
strategic investments is captured in asymmetries in the value of K. Given adequately high<br />
values of R 1<br />
, the leader i would have relatively higher profits than the follower j.<br />
5.3 The obsolescence parameter<br />
L 1<br />
<strong>and</strong> L 2<br />
characterise the decay or in other words the obsolescence effect of quality.<br />
Different values of L 1<br />
<strong>and</strong> L 2<br />
help evaluate the difference am<strong>on</strong>g the two firms regarding<br />
this effect. (L 1<br />
$ 1) G (L 2<br />
$ 0.7) suggests that the obsolescence effect for leader i is higher<br />
than that for the follower j. An investigati<strong>on</strong> into the dynamics of innovati<strong>on</strong> investments<br />
reveals that with higher value of the parameter L, a firm would make higher level of<br />
investments. Interestingly, with a difference in L 1<br />
<strong>and</strong> L 2<br />
the shape of investment<br />
trajectory of the leader is now c<strong>on</strong>vex. That is when the leader faces a high obsolescence<br />
effect the investments are increased at a faster rate to ensure an increasing state<br />
trajectory. However, as opposed to parameter a <strong>and</strong> K such an increased c<strong>on</strong>trol<br />
trajectory doesn’t translate into higher quality. In fact, in this situati<strong>on</strong>, the leader has<br />
lower product quality inspite of higher investments. A high decay effect puts the leader<br />
into a disadvantageous positi<strong>on</strong>. In such a situati<strong>on</strong>, under equilibrium c<strong>on</strong>trol the leader<br />
always loses the competitive game.<br />
5.4 Spillover factor M 2<br />
M 2<br />
characterises the amount of spillover from the leader to the follower. In this secti<strong>on</strong><br />
the impact of reducti<strong>on</strong> of the spillover factor M 2<br />
is c<strong>on</strong>sidered. Evidently in this case the<br />
leader would be less worried about the amount of advantage the follower obtains from<br />
the leader’s own investments. If the spillover effect is low, the leader pursues<br />
investments more aggressively, c<strong>on</strong>sistent with the incentive enhancement of spillover<br />
effects in the c<strong>on</strong>text of network externalities (Saaskilahti, 2006). Under such<br />
circumstances, the c<strong>on</strong>trol trajectory of the leader is also c<strong>on</strong>vexly increasing <strong>and</strong> is <strong>on</strong>ly<br />
marginally lower than that of the follower.<br />
Regarding the impact of such investments <strong>on</strong> the state trajectory, the quality of the<br />
follower is higher than that of the leader <strong>and</strong> both obtain a c<strong>on</strong>vex increasing state<br />
trajectory. But unlike #, K <strong>and</strong> L, a different value of M 2<br />
doesn’t corresp<strong>on</strong>d to any<br />
advantage related to resources, core competence or dynamic capabilities. This is purely<br />
96
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
High speed technology competiti<strong>on</strong> 13<br />
an exogenously defined variable based <strong>on</strong> technology <strong>and</strong> industry types. Under changed<br />
circumstances with respect to M 2<br />
the leader doesn’t really gain much in the competitive<br />
game. The results in fact point to the follower winning the game due to much higher<br />
levels of quality <strong>and</strong> <strong>on</strong>ly marginally higher levels of investments.<br />
6 C<strong>on</strong>clusi<strong>on</strong>s <strong>and</strong> extensi<strong>on</strong>s<br />
There is apparently a paradox in trying to assess, both empirically <strong>and</strong> theoretically, the<br />
impact of competitive pressure <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> growth. On <strong>on</strong>e h<strong>and</strong>, according to the<br />
traditi<strong>on</strong> originating in Schumpeter (1942) the prospective reward provided by m<strong>on</strong>opoly<br />
rent to a successful innovator is required to stimulate sufficient R&D investment <strong>and</strong><br />
technological progress.<br />
On the other h<strong>and</strong>, the incentives to innovate are weaker for an incumbent<br />
m<strong>on</strong>opolist than for a firm in a competitive industry (Arrow, 1962). When competiti<strong>on</strong> is<br />
intense in the product market, innovati<strong>on</strong> may even be seen as the <strong>on</strong>ly way for a firm to<br />
survive. In neo-Schumpeterian models of endogenous growth (Aghi<strong>on</strong> <strong>and</strong> Howitt, 1992;<br />
Grossman <strong>and</strong> Helpman, 1991), innovati<strong>on</strong> allows a firm in an industry to take the lead<br />
<strong>and</strong> gain profit. But the m<strong>on</strong>opoly rent enjoyed by the winner is <strong>on</strong>ly temporary, <strong>and</strong> a<br />
new innovator, capitalising <strong>on</strong> accumulated knowledge, is always able to ‘leapfrog’ the<br />
leader unless the leader is endowed with advantages of firm asymmetries. In the recent<br />
research literature, Aghi<strong>on</strong> et al. (1997) <strong>and</strong> Aghi<strong>on</strong> et al. (2001), supposing a duopoly in<br />
each sector, both at the research <strong>and</strong> producti<strong>on</strong> levels, have introduced what they call<br />
‘step-by-step innovati<strong>on</strong>’, according to which technological progress allows a firm to<br />
take the lead, but with the lagging firm remaining active <strong>and</strong> eventually capable of<br />
catching up. This model has been extended by Encaoua <strong>and</strong> Ulph (2000), allowing for<br />
the possibility that the lagging firm leapfrogs the leader, without driving it out of the<br />
market. Here we adopt a simplified approach to c<strong>on</strong>sider spillover effects. Griliches<br />
(1979) lays out the c<strong>on</strong>ceptual framework <strong>and</strong> provides an early discussi<strong>on</strong> of the<br />
importance of spillover effects of R&D. Later in Griliches (1992) reviewed the recent<br />
empirical evidence <strong>on</strong> spillovers, <strong>and</strong> tentatively c<strong>on</strong>cluded that spillover effects may be<br />
substantial. The model can be extended by evaluating the improvement in product quality<br />
with learning effect. This effect is very well documented in the literature <strong>and</strong> thus the<br />
extensi<strong>on</strong> to incorporate learning effects is straightforward. Specifically, learning can be<br />
used to characterise the dynamics of evoluti<strong>on</strong> of product quality <strong>and</strong> the characterisati<strong>on</strong><br />
of costs associated with innovati<strong>on</strong> investments.<br />
For analytical simplicity the revenue functi<strong>on</strong> is treated as a salvage value. As an<br />
extensi<strong>on</strong> the incorporati<strong>on</strong> of the dynamics of revenue in the analysis of the model can<br />
be explored. In the sequential play game, the state dynamics of the follower can be<br />
c<strong>on</strong>sidered to be dependent <strong>on</strong> the state of the leader at previous time instant.<br />
This modificati<strong>on</strong> would allow for analysis of a model in which the leader uses<br />
open-loop Nash equilibrium strategy for innovati<strong>on</strong> investments whereas the follower<br />
would adopt a Markovian Nash equilibrium strategy.<br />
Additi<strong>on</strong>al insights can be gained by this modificati<strong>on</strong> for leader-follower<br />
competitive dynamics in new product development. Firm asymmetries can be explored<br />
by c<strong>on</strong>sidering multiple parameters at the same time. This would enable a richer<br />
underst<strong>and</strong>ing of the strategies that a leader <strong>and</strong> a follower should adopt based <strong>on</strong> their<br />
strengths <strong>and</strong> weaknesses.<br />
97
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
14 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
The c<strong>on</strong>venti<strong>on</strong>al approach used in innovati<strong>on</strong> <strong>and</strong> R&D research viewed the process<br />
as <strong>on</strong>e with c<strong>on</strong>stant returns, competitive output <strong>and</strong> factor markets <strong>and</strong> no externalities.<br />
However, such a framework doesn’t offer a full explanati<strong>on</strong> of productivity growth.<br />
For a better underst<strong>and</strong>ing it therefore becomes very important to c<strong>on</strong>sider increasing<br />
returns to scale, R&D spillovers <strong>and</strong> other externalities <strong>and</strong> disequilibria.<br />
References<br />
Aghi<strong>on</strong>, P. <strong>and</strong> Howitt, P. (1992) ‘A model of growth through creative destructi<strong>on</strong>’, Ec<strong>on</strong>ometrica,<br />
Vol. 60, pp.323–352.<br />
Aghi<strong>on</strong>, P., Harris, C. <strong>and</strong> Vickers, J. (1997) ‘Competiti<strong>on</strong> <strong>and</strong> growth with step-by-step<br />
innovati<strong>on</strong>: an example’, Review of Ec<strong>on</strong>omic Studies, Vol. 41, pp.771–782.<br />
Aghi<strong>on</strong>, P., Harris, C., Howitt, P. <strong>and</strong> Vickers, J. (2001) ‘Competiti<strong>on</strong>, imitati<strong>on</strong> <strong>and</strong> growth with<br />
step-by-step innovati<strong>on</strong>’, Review of Ec<strong>on</strong>omic Studies, Vol. 68, pp.467–492.<br />
Arrow, K. (1962) ‘The ec<strong>on</strong>omic implicati<strong>on</strong>s of learning by doing’, Review of Ec<strong>on</strong>omic Studies,<br />
Vol. 29, pp.155–173.<br />
Arthur, W.B. (2000) ‘Myths <strong>and</strong> realities of the high-tech ec<strong>on</strong>omy’, Talk given at Credit Suisse<br />
First Bost<strong>on</strong> (CSFB) Thought Leader Forum, 10 September.<br />
Dockner, E., Jorgensen, S., VanL<strong>on</strong>g, N. <strong>and</strong> Sorger, G. (2000) Differential Games in Ec<strong>on</strong>omics<br />
<strong>and</strong> Management Science, Cambridge: Cambridge University Press.<br />
Encaoua, D. <strong>and</strong> Ulph, D. (2000) ‘Catching-up or leapfrogging’, Cahiers de la MSE, Vol. 5,<br />
pp.10–22.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2006) Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Griliches, Z. (1979) ‘Issues in assessing the c<strong>on</strong>tributi<strong>on</strong> of Research <strong>and</strong> <strong>Development</strong> to<br />
productivity growth’, The Bell Journal of Ec<strong>on</strong>omics, Vol. 10, No. 1, pp.92–116.<br />
Griliches, Z. (1992) ‘The search of R&D spillovers’, The Sc<strong>and</strong>inavian Journal of Ec<strong>on</strong>omics,<br />
Vol. 94, pp.S29–S47.<br />
Grossman, G. <strong>and</strong> Helpman, E. (1991) Innovati<strong>on</strong> <strong>and</strong> Growth in the Global Ec<strong>on</strong>omy, Cambridge,<br />
MA: MIT Press.<br />
Hoffman, W.D. (1976) ‘Market structure <strong>and</strong> strategies of R&D behavior in the data processing<br />
market – theoretical thoughts <strong>and</strong> empirical findings’, Research Policy, Vol. 5, pp.334–353.<br />
Kanter, R.M. (2006) ‘Innovati<strong>on</strong>: the classical traps’, Harvard Business Review, November,<br />
pp.72–83.<br />
Robs<strong>on</strong>, A.J. (1990) ‘Duopoly with endogeneous strategic timing: stackelberg regained’,<br />
Internati<strong>on</strong>al Ec<strong>on</strong>omic Review, Vol. 31, pp.25–35.<br />
Saaskilahti, P. (2006) ‘Strategic R&D <strong>and</strong> network compatibility’, Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong><br />
New <strong>Technology</strong>, Vol. 15, No. 8, pp.711–733.<br />
Schumpeter, J.A. (1942) Capitalism, Socialism <strong>and</strong> Democracy, New York: Harper.<br />
Teece, D.J., Pisano, G. <strong>and</strong> Shuen, A. (1997) ‘Dynamic capabilities <strong>and</strong> strategic management’,<br />
Strategic Management Journal, Vol. 18, No. 7, pp.509–533.<br />
Varian, H.R., Farrell, J. <strong>and</strong> Shapiro, C. (2004) The Ec<strong>on</strong>omics of Informati<strong>on</strong> <strong>Technology</strong>,<br />
Cambridge: Cambridge University Press.<br />
98
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
Appendix<br />
High speed technology competiti<strong>on</strong> 15<br />
Mathematical Preliminaries, Definiti<strong>on</strong>s <strong>and</strong> Theorems<br />
This secti<strong>on</strong> covers some of the salient analytical aspects specific to a sequential play<br />
game.<br />
For a finite horiz<strong>on</strong> T let L <strong>and</strong> F denote the leader <strong>and</strong> follower respectively. Let<br />
x denote the vector of state variables, u L the vector of c<strong>on</strong>trol variables of the leader <strong>and</strong><br />
u F the vector of c<strong>on</strong>trol variables of the follower. Assume x 6 R n , u L 6 R mL <strong>and</strong> u F 6 R mF .<br />
Definiti<strong>on</strong> 1: The initial value of the follower’s costate variable E i<br />
is said to be<br />
n<strong>on</strong>-c<strong>on</strong>trollable if E i<br />
(0) is independent of the leader’s c<strong>on</strong>trol path u L (t). Otherwise, it is<br />
said to be c<strong>on</strong>trollable.<br />
The definiti<strong>on</strong> suggests that if the costate variable is c<strong>on</strong>trollable, the follower’s c<strong>on</strong>trol<br />
variable u F (t) at time t depends also <strong>on</strong> future values of u L (3); that is, <strong>on</strong> values u L (s) with<br />
s G t.<br />
Theorems <strong>and</strong> Proofs<br />
Theorem 1: The maximised costate variables for the follower are a functi<strong>on</strong> of time <strong>and</strong><br />
is given by<br />
where<br />
E<br />
9 $ E e<br />
(A1)<br />
'<br />
E<br />
% 2t<br />
%'<br />
E e<br />
9 '<br />
$ (A2)<br />
2 %2<br />
% t<br />
E 9 '<br />
<strong>and</strong> E 9<br />
2<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the follower firm's own state <strong>and</strong> the state of the leader i; E 1<br />
(0) $ E 01<br />
, E 2<br />
(0) $<br />
E 02<br />
are c<strong>on</strong>stants.<br />
Proof: Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s are derived by c<strong>on</strong>structing the Hamilt<strong>on</strong>ians.<br />
The analytical soluti<strong>on</strong> is derived for the follower firm j by writing the Hamilt<strong>on</strong>ian as:<br />
*<br />
* * #<br />
! " E &<br />
* * i<br />
/ $ % & '<br />
' 2<br />
! "<br />
2<br />
! "<br />
2<br />
! "'<br />
( u t ) / K u t % % x t / - u t<br />
( )<br />
i<br />
#<br />
/ E &<br />
i<br />
i<br />
2<br />
K' u ! t " % %'<br />
x ! t"<br />
'<br />
( )<br />
(A3)<br />
The necessary c<strong>on</strong>diti<strong>on</strong>s for optimality are:<br />
*<br />
/ $ %<br />
(A4)<br />
*<br />
u<br />
From (A5)<br />
E $ % / : E $ % /<br />
(A5)<br />
9 * 9 *<br />
' *<br />
x 2<br />
i<br />
x<br />
x !%" $ x<br />
(A6)<br />
* *<br />
%<br />
E E E E<br />
9 9 % 2t<br />
'<br />
$ %<br />
2 ':<br />
'<br />
$<br />
%'<br />
e<br />
(A7)<br />
E E E E<br />
9 9 %'<br />
t<br />
2<br />
$ %' 2:<br />
2<br />
$<br />
%2e<br />
(A8)<br />
99
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
16 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
where E 1<br />
(0) $ E 01<br />
is the known positive c<strong>on</strong>stant denoting the initial value of the<br />
costate E 1<br />
<strong>and</strong> E 2<br />
(0) $ E 02<br />
is a known negative c<strong>on</strong>stant denoting the initial value of the<br />
costate E 2<br />
.<br />
Theorem 2: The Stackelberg equilibrium investment by the follower in product<br />
development is given as<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ (<br />
K e )<br />
(A9)<br />
Proof: Differentiating (A3) with respect to u j (t), we obtain:<br />
*<br />
* *<br />
# %'<br />
u*<br />
$ %'/ E<br />
'#<br />
&<br />
2 ( ! ")<br />
'<br />
/ K u t<br />
(A10)<br />
By Equati<strong>on</strong> (A10) to zero <strong>and</strong> some algebraic manipulati<strong>on</strong>s, the following expressi<strong>on</strong> is<br />
obtained for optimal effort in product development:<br />
* 9 * 9<br />
2 '<br />
*<br />
' 8!' %#<br />
"<br />
u ! t"<br />
$ & ( K # E ' )<br />
(A11)<br />
Substituting the expressi<strong>on</strong> for E i<br />
in (A11):<br />
*<br />
* 9 * %<br />
'8!' "<br />
2t<br />
%#<br />
! " &<br />
2<br />
# E '<br />
%'<br />
u t $ ( K e )<br />
(A12)<br />
Theorem 3: The maximised costate variables for the leader are give n by,<br />
F<br />
9 $ F e<br />
(A13)<br />
'<br />
%'<br />
t<br />
%'<br />
F<br />
F e<br />
9 % 2t<br />
2<br />
$<br />
%2<br />
(A14)<br />
F F %<br />
9 $ % % 2t<br />
9 %9e<br />
(A15)<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the<br />
follower; F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Proof: The Stackelberg equilibrium c<strong>on</strong>diti<strong>on</strong>s are derived by c<strong>on</strong>structing the<br />
Hamilt<strong>on</strong>ian. An analytical soluti<strong>on</strong> is derived for the leader firm I by writing the<br />
Hamilt<strong>on</strong>ian as:<br />
/ & u t ' & K u t % x t '<br />
i<br />
i i i # i<br />
$ % ( ! ") / F<br />
'<br />
! " %<br />
'<br />
! "<br />
( )<br />
/ F & % / ' / F E<br />
( )<br />
*<br />
* #<br />
* i<br />
2<br />
K2u ! t " %<br />
2x ! t" -<br />
2u ! t "<br />
9 '%<br />
2<br />
@<br />
A<br />
(A16)<br />
The necessary c<strong>on</strong>diti<strong>on</strong>s for optimality are<br />
i<br />
/ $ %<br />
(A17)<br />
i<br />
u<br />
F $ %/ : F $ %/ : F $ %/ E<br />
(A18)<br />
9 i 9 i 9 i<br />
' i 2 *<br />
x<br />
x 9<br />
'<br />
i<br />
x !%" $ x<br />
(A19)<br />
i<br />
%<br />
100
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
HANS W. GOTTINGER<br />
From (A18)<br />
High speed technology competiti<strong>on</strong> 17<br />
F F F F<br />
9 9 %'<br />
t<br />
'<br />
$ %' ':<br />
'<br />
$<br />
%'<br />
e<br />
(A20)<br />
F F F F<br />
9 9<br />
% 2t<br />
2<br />
$ %<br />
2 2:<br />
2<br />
$<br />
%2e<br />
(A21)<br />
9 % 2t<br />
9<br />
$ %<br />
9% 2:<br />
9<br />
9<br />
$ %<br />
%9e<br />
(A22)<br />
F F F F %<br />
where F 9 ( '<br />
F 9 2<br />
<strong>and</strong> F 9<br />
9<br />
are the costate variables reflecting the marginal price for a unit<br />
increase in the leader’s own state, the state of the follower <strong>and</strong> the costate of the<br />
follower; F 01<br />
, F 02<br />
<strong>and</strong> F 03<br />
are c<strong>on</strong>stants.<br />
Theorem 4: The Stackelberg equilibrium investment by the leader in product<br />
development is given as:<br />
i<br />
u ! t"<br />
i<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
Proof: Differentiating (A3) with respect to u i (t), we obtain:<br />
(A23)<br />
i<br />
i i<br />
# %'<br />
& '<br />
' '<br />
F<br />
2 2<br />
/ i $ %'/ F # K u ! t"<br />
/ -<br />
u<br />
( )<br />
(A24)<br />
By Equati<strong>on</strong> (A24) to zero <strong>and</strong> some algebraic manipulati<strong>on</strong>s, the following expressi<strong>on</strong> is<br />
obtained for optimal effort in product development:<br />
i<br />
u ! t"<br />
i<br />
i 9<br />
'8!'%#<br />
"<br />
9<br />
& K'<br />
# F '<br />
'<br />
$ : 9 ;<br />
'%<br />
-<br />
2F<br />
2<br />
( )<br />
(A25)<br />
Substituting the expressi<strong>on</strong> for<br />
i<br />
u ! t"<br />
i<br />
i %<br />
'8!' "<br />
' t<br />
%#<br />
9<br />
& K'<br />
# F<br />
%'<br />
e '<br />
$ : % 2t<br />
;<br />
'%<br />
-<br />
2<br />
F<br />
%2e<br />
( )<br />
F 9 '<br />
<strong>and</strong> F 9 2<br />
in (A25) yields<br />
(A26)<br />
Theorem 5: The equilibrium state trajectory of performance improvement of the follower<br />
is given as<br />
&<br />
* *<br />
x ! t " $ % x / %'/ e - : ;<br />
i<br />
'8!'%<br />
# "<br />
%% 2t<br />
i %'<br />
t<br />
e<br />
&<br />
2<br />
'# F '<br />
:<br />
% t K<br />
%'<br />
e<br />
2 % 7 8 2<br />
% 2t<br />
% :<br />
2 (: '%-2 F%2e<br />
;<br />
(<br />
)<br />
* * 2<br />
# 8!' % # "<br />
% 2t * % 2t<br />
'<br />
/ %'/ e K K # E e ; )<br />
7 8 2 7 2 %' 8<br />
(A27)<br />
Proof: The expressi<strong>on</strong> for the optimal state trajectory can be obtained by c<strong>on</strong>sidering the<br />
state dynamics given in Equati<strong>on</strong> (11). Substituting u i (t) $ u i (t) 9 <strong>and</strong> u j (t) $ u i (t) 9 in<br />
(11) the following first-order differential equati<strong>on</strong> can be obtained<br />
* i i 9<br />
* * 9<br />
# 8!' %# "<br />
*<br />
& K'<br />
# F '<br />
'<br />
x ! t" $ K &<br />
2 ( K<br />
2<br />
# E '<br />
' ) % %<br />
2x ! t " 7, p<br />
/ H % t8<br />
/ -<br />
2 : 9 ;<br />
('%<br />
-<br />
2F<br />
2 )<br />
i<br />
'8!'%#<br />
"<br />
(A28)<br />
101
HANS W. GOTTINGER<br />
1.6 HIGH SPEED TECHNOLOGY COMPETITION<br />
18 H-W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
The first order differential equati<strong>on</strong> can be solved with the initial c<strong>on</strong>diti<strong>on</strong> x<br />
* !%" $ x<br />
*<br />
%<br />
;<br />
The resulting expressi<strong>on</strong> for x j (t) 9 is:<br />
&<br />
x t % x e -<br />
i<br />
'8!'%#<br />
"<br />
%% 2t<br />
i %'<br />
t<br />
* e<br />
&<br />
2<br />
'# F '<br />
%'<br />
! " :<br />
*<br />
% t K e<br />
$<br />
2 %<br />
/ 7% '/ 8 2 : % ;<br />
2t<br />
%<br />
2<br />
: ('%<br />
-<br />
2<br />
F<br />
%2e<br />
)<br />
(<br />
% 2t * % 2t<br />
7 ' e 8 K2 7K2 # E<br />
%'<br />
e 8<br />
/ % /<br />
* * 2<br />
# 8!' %#<br />
"<br />
'<br />
;)<br />
(A29)<br />
Theorem 6: The equilibrium state trajectory of performance improvement of the leader is<br />
given as:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
%% t<br />
i e<br />
&<br />
i<br />
%'<br />
t<br />
& K'<br />
# F<br />
%'<br />
e '<br />
'<br />
x ! t " $ :%' x% / 7% '/ e 8 K'<br />
: % 2t<br />
;<br />
;<br />
%' : ('%<br />
-<br />
2<br />
F<br />
%2e<br />
) ;<br />
( )<br />
(A30)<br />
Proof: The expressi<strong>on</strong> for the optimal state trajectory can be obtained by c<strong>on</strong>sidering the<br />
state dynamics given in Equati<strong>on</strong> (14). Substituting u i (t) $ u i (t) 9 in (14) the following<br />
first-order differential equati<strong>on</strong> can be obtained:<br />
i i<br />
i 9<br />
# 8!' %#<br />
"<br />
*<br />
& K'<br />
# F '<br />
'<br />
i<br />
! " $<br />
' : '<br />
! "<br />
9 ; %<br />
'%<br />
-<br />
2F<br />
2<br />
x t K % x t<br />
( )<br />
(A31)<br />
i<br />
i<br />
The first order differential equati<strong>on</strong> can be solved with the initial c<strong>on</strong>diti<strong>on</strong> x !%" $ x%<br />
;<br />
The resulting expressi<strong>on</strong> for x i (t) 9 is:<br />
i i<br />
'<br />
i %<br />
8!' "<br />
' t<br />
# %#<br />
% t<br />
i<br />
%'<br />
t<br />
& K'<br />
# F<br />
%'<br />
e '<br />
:<br />
' % 7 8<br />
;<br />
'<br />
% 2t<br />
'<br />
: '%<br />
2<br />
F<br />
%2<br />
;<br />
%<br />
i e<br />
& '<br />
x ! t " $ % x / %'/ e K : ;<br />
% ( - e<br />
(<br />
)<br />
)<br />
(A32)<br />
102
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
103
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
104
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
105
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
106
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
107
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
108
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
109
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
110
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
111
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
112
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
113
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
114
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
115
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
116
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
117
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
118
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
119
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
120
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACING<br />
HANS W. GOTTINGER<br />
121
HANS W. GOTTINGER<br />
1.7 SEQUENTIAL TECHNOLOGY CHOICE AND R&D RACINGN<br />
122
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'()*+!$$),$"-!<br />
.//0+112223456785709:;64.68?1@16@:7=!<br />
A>6+!$#3$$'BC1@36@:7=3"#$%#'#)3$)!<br />
DEEF+!"-"C,%)B-!(G=68/*&!DEEF+!"-"C,%)'H!(I8;687*! !<br />
!<br />
65$*!"#$%"$#*(-*65'"7/8*12("(3'2*&"0*!"0,/#%'&)*<br />
9'3$"/'("/*<br />
!"#$%&'(#'(%)*++,#-'(%<br />
E/=J/78>?654L!EMNOMKP!P>849;/683!)L!"#$%L!003!$$),$"-3!A>6+!$#3$$'BC1@36@:7=3"#$%#'#)3$)!<br />
6'4',7'1+!O9,2JS! 5>??9865J/6>8! :75J?7!<br />
5>??>8L!2.65.!78J:;7A!947=!0J=/6560J/6>8L!5>;;J:>=J/6>8!J8A!<br />
68/7=J5/6>83! W7:! "3#! /75.8>;>56J;! 87/2>=]68;>?7!74478/6J;!/>!?>A7=8!<br />
4>56J;!68/7=J5/6>8!J4!27;;!J4!X>=!:J;!:94687443!W.6;7!W7:!<br />
"3#!59==78/;S!A>?68J/74!/.7!D8/7=87/L!87/2>=]!78=]682J=A4!J8>/.7=!J;L!5>??>8;S!=7X7==7A!/>!J4!<br />
/.7!E7?J8/65!W7:!J8A!4>?7/6?74!=7X7==7A!/>!J4!W7:!-3#3!M.7!<br />
J;! >X! /.7! E7?J8/65! W7:! (7?:=J5689=]4*!64!/>!?J=]!90!27:!5>8/78/!<br />
68!J!2JS!/.J/!?J]74!6/!98A7=4/J8AJ:;7!:S!?J5.6874L!J;;>268! :7.J[7! ?>=7! 68/7;;6=?J/4!2>9;A!J;;>2!<br />
?J5.6874!/>!0=>5744!J8A!4.J=7!AJ/J!>8!/.76=!>28L!26/.>9/!/.7!<br />
877A! X>=! .9?J8! ?7A6J/6>83! O;>80?78/4! 68!<br />
D8/7=87/! /75.8>;>;>=! F7/2>=]4! J8A!<br />
F7J=! d67;A! P>??9865J/6>8! (FdP*! 9468! d=7c9785S!<br />
DA78/6/S! (NdDY*! /J! :778! 7[>;[688[7=X! /.747! /2>! /75.8>;>=!<br />
F7/2>=]4L! 64! ;7JA68! 872! 0>446:6;6/674! J8A! [646>843! M.7!<br />
0>446:6;6/S!>X!J!X=J?72>=]!/.J/!2>9;A!J;;>2!A6=75/!?J5.687,<br />
/>,! ?J5.687! 5>??9865J/6>8! >[7=! /.7! D8/7=87/! .J4! ;7A!<br />
=747J=5.7=4! />! 78[646>8! /.7! :787X6/4! >X! :=688;687!J8A!J;;>268!0J=/6560J/7!68!/.7!27:!<br />
J4! J! [J4/! 87/2>=]! >X! J9/>8>?>94L! 47;X,>=A957A!J!0J=JA6
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $$'!<br />
!<br />
D8/7=87/,;6]7!87/2>=]3!Y7[6574!J/!.>?7!8>!?J//7=!/.76=!46`7L!<br />
J=7!=!0J446[7!NdDY!/J8A6/6>87=! J=7! J//J5.7A! />! J! /=J84576[7=! />!<br />
5>??9865J/7! 26/.! J! ;>5J;! 47=[7=3! M.64! ;>5J;! 47=[7=! 90AJ/74!<br />
/.7! 4784>=! >9/09/! AJ/J! 68/>! /.7! D8/7=87/! /.=>95J;!<br />
47=[7=3! M.7! AJ/J! X=>?! D8/7=87/! 64! J557447A! X=>?! J! A74]/>0L!<br />
?>:6;7!0.>87!>=!J8S!A7[657!/.J/!64!5>8875/7A!/>!/.7!D8/7=87/3!<br />
I87! >X! /.7! ?J68! JA[J8/JX! /.64! ]68A! >X! 87/2>=]68;[68?7! J00;6J8574! J8A! A7[6574! 64! /.J/!<br />
27! 5J8! .J[7! 5>8/=>;! >[7=! >9=! A7[6574! /.=>99/! J! :J;!<br />
=7J5.!2.7=7!5>??9865J/6>8!64!0>446:;73!M.7!0>/78/6J;!J=6474!<br />
>8! 87J=;S! 7[7=S! 45J;7! J4! 27! 5>9;A! 6?J8>?>94! 47;X,A=6[68874*L!<br />
4?J=/! .>?74! J8A! 4?J=/! 56/6743! D>M! 5J8! :7! 6?0;7?78/7A! 68!<br />
:9468744!>=!68A94/=6J;!78[6=>8?78/4!:J47A!>8!/.7!=7c96=7?78/!<br />
78J:;7A! /75.865J;;S! :S! ?J5.687,/>,?J5.687! (Q"Q*!<br />
5>??9865J/6>8!(O;0JSA68!["]*3!<br />
M.7=7! J=7! ?J8S! A7[6574! 5>8875/7A! />! /.7! D8/7=87/3! d>=!<br />
7\J?0;7L! />! :! 27! .JA! />! 84!?J89J;;SL!:9/!8>2!27!.J[7!0=68/7=4!/.J/!5J8!:7!<br />
5>8875/7A! 26=7;744;S! /.=>9?7! 87/2>=]3! M.7!<br />
0=68/68:!5J8!:7!A>87!X=>?!J8S2.7=7!J8A!>8!J8S!A7[657!<br />
495.!J4!J!5>?09/7=!>=!J!4?J=/0.>873!D8!/.7!4J?7!2JSL!J;;!/.7!<br />
J00;6J8574!J8A!A7[6574!J/!.>?7!5J8!:7!5>8/=>;;7A!=7?>/7;S3!D/!<br />
64! J5.67[7A! :S! J==J8=]!J8A!5>8875/68!J!A75646>8,<br />
?J]68=]! :75J947! /.7!<br />
A75646>8,?J]688!>8!/.7!4/J/94!>X!/.7!<br />
A7[6574!68!/.7!87/2>=]3!<br />
D8!J8!D>M!78[6=>8?78/L!/.7!5>8/=>;;68??9865J/6>8!<br />
>X! A7[6574! 68[>;[74! ?95.! 4784>=! 87/2>=]4! 26/.! 6809/! J8A!<br />
>9/09/!0>=/43!<br />
W.J/! A64/68M! X=>?! /=JA6/6>8J;! 87/2>=]!<br />
/75.8>;>,X>;A+! (6*! /.7! ^/.68X! A6[7=47!<br />
8J/9=7! 495.! J4! ?J5.687! 0J=/4L! 5>8/=>;! 986/4L! 5>??9865J/6>8!<br />
A7[6574LfL! (66*! /.7S! 78=7! 5>?0;7\!<br />
5>??9865J/6>8! 46?6;J=! J8A! 90! />! /.7! ;7[7;! >X! .9?J8!<br />
5>??9865J/6>83! a3! O=/.9=! [-]L! /.7! E/J8X>=A! 75>8>?64/! J8A!<br />
XJ/.7=! >X! ^685=7J4688A!K5>8>?S!J8A!7\7?0;6X674!6/!;6]7!/.64+!<br />
gf!6X!D_?!A=6[684!O82!>X!/=JXX65!2.7=7!?S!5J=!64!68!<br />
J! 5>8[7=4J/6>8! 26/.! /.7! 5J=4! J=>98A! 6/! /.J/! J=7! 68! J!<br />
5>8[7=4J/6>8! 26/.! 8A! 75>8>?S! 64! 5=7J/68=! 94! ,,,! 4;>2;SL! c967/;SL! J8A!<br />
4/7JA6;S!,,,!J!A6XX7=78/!2>=;Ah3!<br />
CA%D"4E-(*;#1%6'$'"(45%*#%B#+'(#'+%*F%<br />
35,#-$%<br />
D>M! ?7J84! 5>8875/68! /.7! D8/7=87/L! 6/! 5J8!:7! J8!<br />
>:@75/! >=! J! A7[6573! M.7! 5>??9865J/6>8! >X! A7[6574! >8! /.7!<br />
D8/7=87/,;6]7!87/2>=]!5J8!:7!>87!46A7A!>=!/2>!46A7A!68!/7=?4!<br />
>X! 87/2>=]68??9865J/6>8! 5J8! :7! J!<br />
46?0;7\! ]68A! >X! 5>??9865J/6>8! >=! A90;7\! 5>??9865J/6>83!<br />
M.7! D8/7=87/! .J4! :778! 9468X! /75.8>;>?! (5>??7=56J;*&! /.7! 5>??9865J/6>8!<br />
:7/2778! /.7! A7[6574! .J4! 7[>;[7A! J4! D8/7=87/! /75.8>;>07A3!M.7!872! 27:!A7[7;>0?78/!>8!/.7!27:!J=7!<br />
W7:! "! J8A! W7:! -! /75.8>;>;>;>=!A7[6574!<br />
/>! 5>??9865J/7! 26/.! /.7! 27:! A6=75/;SL! ?J]688,?J]6807=!<br />
J;=6/.?!/>!?J]7!J!A75646>8!J8A!J5/!J55>=A68=!684/J857L!5>846A7=!J8!J5/6[7!NdDY!5.60!7c96007A!26/.!<br />
J! =7;7[J8/! 4784>=&! /.64! NdDY! ;J:7;! 949J;;S! ?! /.7! 0>27=7A! 4784>=! 68! /.7! 5.60! J8A! 478A4! /.J/!<br />
AJ/J! />! /.7! 47=[7=! :9/! A>74! 8>/! /J]7! J8S! J5/6>83! V>27[7=L!<br />
2.78!27!??J8A4!68!5>?09/7=!;J8>47! J8!<br />
J00=>0=6J/7!J5/6>8!X=>?!/.7!X!J5/6>843!D8!/.64!5J47L!<br />
27! J=7! /J;]689/! ?J5.687,/>,?J5.687! (Q"Q*!<br />
5>??9865J/6>8! 2.7=7! /.7! ?J5.687! 5>??9865J/74! 26/.68!<br />
/.7?47;[74!J8A!07=X>=?4!J8!J00=>0=6J/7!J5/6>83!%<br />
D8A94/=6J;!A=6[68=574!68!/.7!D>M!5>8/7\/!!<br />
O55>=A68! D>M! A7[7;>0?78/L! D>M! :9468744! 7[>;[74!<br />
/.=>9! /7;7,<br />
>07=J/6>8! J8A! /7;7,0=747857! J4! J! =749;/! >X! /.7! 68/7=J5/6>8!<br />
X>=574! >X! :>/.! ?J=]7/! 09;;! J8A! /75.8>;>//680?78/!=>JA?J0!68A65J/74!/.J/!/.7!0=>;>8/689>94;S! 5>8/=6:9/7! />! /.7!<br />
A7[7;>0?78/! >X! D>ML! 2.6;7! /.7! 5>??7=56J;6`J/6>8! 68! /.7!<br />
?J=]7/!0;J57!64!J8>/.7=!]7S!64497!=7;J/68!/.7!0=>X!<br />
D>M3! M.7=7X>=7L! D>M! :9468744! 64! 0=>?>/7A! :S! 68A94/=6J;!<br />
A=6[68=574!>X!:>/.!/75.8>;>;>8!:76890?78/! >X! =7;J/7A! /75.8>;>0?78/!>X!D>M!/75.8>;>94!0>/78/6J;!/>!<br />
i=?4! 68! D>M! :94687443! d>=! 7\J?0;7L! ?J8S! 6?0>=/J8/!<br />
/75.8>;>9A!5>?09/68;>=! 87/2>=]! /75.8>;>?>/7A! /.7! A7[7;>0?78/!<br />
>X!D>M!:9468744!/>!J!872!;7[7;!(W.6/?>=7!7/!J;3![)]*3!<br />
E75>8AL!?J=]7/!09;;!64!A7i87A!J4!J8!688>[J/6[7!X>=57!:76807A! :S! i=?4! 68! =740>847! />! J8! 6A78/6i7A! J8A1! >=!<br />
0>/78/6J;! ?J=]7/! 877A! 5>8A6/6>8J;! >8! J! =?>94! ?J=]7/! A7?J8A! 68! D>M! :9468744!<br />
0=>[6A74! 980=757A78/7A! ?J=]7/! >00>=/986/674! />! i=?43! d>=!<br />
7\J?0;7L! D>M! J00;65J/6>84! 68! .>?7! J00;6J8574L! .7J;/.5J=7L!<br />
J8A! J9/>?>:6;74! J=7! 09;;7A! :S! 872! ?J=]7/! A7?J8A4! ;6]7!<br />
.>947.>;A!A7?J8A!X>=!.>?7!J9/>?J/6>8L!0J/678/!A7?J8A!X>=!<br />
594/>?6`7A! 47=[657L! J8A! 594/>?7=! A7?J8A! X>=! 68/7;;68J;! M7;75>??9865J/6>8! j86>8! (DMj*! X>=!<br />
684/J857! A7i874! /.7! D8/7=87/! >X! M.68=!/.7!D8X>=?J/6>8!E>567/SL!78J:;688875/688L! 7\64/68;[6807=J:;7! 68X>=?J/6>8!<br />
J8A! 5>??9865J/6>8! /75.8>;>
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$$%! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
/.7! /.68?7! 5>8875/7A! 68! /.7! D>M3! I/.7=!<br />
A7i86/6>84!X>594!>8!D8/7=87/,=7;J/7A!J4075/4!>X!/.7!D>ML!495.!<br />
J4! D8/7=87/! 0=>/>5>;4! J8A! 87/2>=]! /75.8>;>8!47?J8/65!5.J;;78M!=7;J/68!/.7!<br />
4/>=J=8! >X! ;J=;9?74! >X!<br />
68X>=?J/6>8!(O/`>=6!7/!J;3![%]*3!!<br />
W.6;7!/.7=7!64!8>!986[7=4J;!A7X686/6>8!X>=!/.7!D>ML!/.7!5>=7!<br />
5>8570/! 64! /.J/! 7[7=SAJS! >:@75/4! 5J8! :7! 7c96007A! 26/.!<br />
6A78/6XS682! /.7?! />! 5>??9865J/7! 26/.! >87! J8>/.7=! J8A!<br />
26/.!>/.7=!A7[6574!J8A!47=[6574!>[7=!/.7!D8/7=87/!/>!J5.67[7!<br />
4>?7! 947X9;! >:@75/6[73! M.7! 5>=7! 4688570/4! 98A7=;S68M!J=7!8>/!8723!d>=!S7J=4L!/75.8>;>=! 87/2>=]4! .J[7! :778! 947A! 68! 68A94/=6J;!<br />
J8A! ?J89XJ5/9=688/7\/4! X>=! /=J5]685]3!M.7!6A7J!>X!A6=75/!?J5.687,/>,!<br />
?J5.687!5>??9865J/6>8!64!J;4>!8>/!/.J/!872L!J4!6/!64!:J465!/>!<br />
/.7!6A7J!>X!/.7!D8/7=87/!68!2.65.!5;678/4L!47=[7=4!J8A!=>9/7=4!<br />
5>??9865J/7!26/.!7J5.!>/.7=3!W.J/!/.7!D>M!=70=7478/4!64!J8!<br />
7[>;9/6>8!>X!/.7!947!>X!/.747!7\64/68;>X!<br />
/.7! 89?:7=! J8A! ]68A4! >X! A7[6574! J4! 27;;! J4! /.7!<br />
68/7=5>8875/6>8! >X! 87/2>=]4! >X! /.747! A7[6574! J5=>44! /.7!<br />
D8/7=87/3!d>=!7\J?0;7L!?>4/!A7[6574!59==78/;S!>8!/.7!D8/7=87/!<br />
27=7!>=6X!/.7!D8/7=87/!J8A!.J[7!<br />
68/7574468=J=]! 5J0J:6;6/6743!<br />
M.747!A7[6574!685;9A7!47=[7=4L!A74]/>04! ;J0/>04L!/J:;7/4!J8A!<br />
4?J=/!0.>8743!W.J/!/.7!D>M!0=>0>474!64!/>!J//J5.!/75.8>;>! 7[7=SAJS! A7[6574L! 495.! J4! J9A6>1[6A7>! =7576[7=4L! 4?>]7!<br />
A7/75/>=4L! .>?7! J00;6J8574L! 7/53! J8A! ?J]688;687L!<br />
7[78!6X!/.7S!27=7!8>/!686/6J;;S!A746/.7=!?J@>=!7[>;9/6>8J=S!5.J8?647A!:S!/.7!<br />
D>ML!64!/.7!68/78!>X!87/2>=]4!/.J/!5>8/J68!/.747!A7[6574L!<br />
?J]689=!7\J?0;7L!NdDY!.J4!:778!947A!X>=!S7J=4!/>!/=J5]!0=>A95/4!<br />
/.=>9X!/.7!4900;S!5.J683!V>27[7=L!>857!/.7!<br />
0=>A95/! ;7X/! /.7! 4.7;X! >X! J! =7/J6;! >9/;7/L! /.7! ?J89XJ5/9=7=! _4!<br />
J:6;6/S!/>!/=J5]!/.7!>:@75/!2J4!;>4/3!!<br />
b6]72647L! 5>849?7=4! 27=7! 98J:;7! />! ! /.7!<br />
;6X75S5;7!68X>=?J/6>8!>X!0=>A95/4! /.7S!09=5.J47A3!aS! 9M! 0=>?6474! />! 78J:;7! 0=>A95/!<br />
/=J57J:6;6/S!/.=>99/!/.7!78/6=7!0=>A95/!;6X75S5;73!<br />
M.7! 7?7=;>X!A7[6574!J8A!<br />
/.7=7X>=7! /.7! >=68570/! >X! D>ML! .7857! 64! 7\/78A68!<br />
J?:678/! 68/7;;68>?>94! 5>8/=>;3! M>! AJ/7L! J!<br />
89?:7=!>X!/75.8>;>;[7A!68!D>ML!495.!J4!26=7;744!<br />
4784>=! 87/2>=]4! (WEF4*L! :J=5>A74L! 68/7;;684L! 5;>9A!<br />
5>?09/68=]!A7078A688!/.7!4/J8AJ=A6`J/6>83!<br />
Y7078A688! [J=6>94! /75.8>;>=! /.7!<br />
6?0;7?78/J/6>8L! /.7! A7X686/6>8! >X! /.7! D>M! [J=6743! V>27[7=L!<br />
/.7!X98AJ?78/J;!>X!D>M!6?0;674!/.J/!>:@75/4!68!J8!D>M!5J8!:7!<br />
6A78/6X67A! 986c97;S! 68! /.7! [6=/9J;! =70=7478/J/6>843! W6/.68! J8!<br />
D>ML! J;;! /.68! 7\5.J85744!AJ/J!J55>=A68!0=7A7X687A!45.7?743!<br />
GA%H9'4,F,4%3'45#*0*-,'$%"#1%I$"-'%<br />
O/!/.7!5>=7!>X!/.7!6A7J!>X!/.7!D8/7=87/!>X!M.68/6>8!/.J/!7[7=SAJS! ^/.68A4! 26;;! :7!<br />
7c96007A! 26/.! /=J5]688! 64! X9;;S! J5/9J;6`7AL! ^/.68! 5>8/J68! ?>=7!<br />
4>0.64/65J/7A!0=>574468=]68:@75/4!/>!98A7=4/J8A!/.76=!78[6=>8?78/4L!<br />
68/7=J5/!26/.!07>0;7!J8A!:75>?7!4?J=/3!b6]7!J8S!68X>=?J/6>8!<br />
4S4/7?L! /.7! D>M! 26;;! =7;S! >8! J! 5>?:68J/6>8! >X! .J=A2J=7L!<br />
4>X/2J=7! J8A! J=5.6/75/9=74L! J8A! /.=>9X/2J=7! ?>=7! J8A!<br />
?>=7!O=/6X656J;!D8/7;;68>?>94! J8A! 47=[7! J4! 68/7;;68,?J]68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $$C!<br />
!<br />
E784>=!87/2>=]4!J=7!A7[6574!/.J/!?>86/>=!5.J=J5/7=64/654!>X!<br />
/.7! 78[6=>8?78/! >=! >/.7=! >:@75/4! 495.! J4! /7?07=J/9=7L!<br />
.9?6A6/SL! ?>[7?78/L! J8A! c9J8/6/S3! W.78! ?9;/60;7! 4784>=4!<br />
J=7! 947A! />! J4! J!<br />
26=7;744! 4784>=! 87/2>=]! (WEF*3! W6=7;744! 4784>=! 87/2>=]4!<br />
5>8/J68! /.7! 4784>=4! /.7?47;[74! J8A! ?JS! J;4>! 5>8/J68!<br />
;;75/!AJ/J!X=>?!/.7!4784>=4!J8A!0J44!6/!>8!/>!J!<br />
47=[7=3!!<br />
W.6;7! 4784>=4! ^47847_! /.7! 4/J/7! >X! J8! 78[6=>8?78/! >=!<br />
>:@75/L!J5/9J/>=4!07=X>=?!J5/6>84!/>!JXX75/!/.7!78[6=>8?78/!>=!<br />
>:@75/!68!4>?7!2JS3!O5/9J/>=4!5J8!JXX75/!/.7!78[6=>8?78/!:S!<br />
7?6//6898AL! ;6! 2J[74! >=! 7[78! 4?7;;43! M.747!<br />
5J0J:6;6/674! J=7! >87! 2JS! /.J/! D>M! >:@75/4! 5J8! 5>??9865J/7!<br />
26/.! 07>0;73! O5/9J/>=4! J=7! X=7c978/;S! 947A! 68! 5>?:68J/6>8!<br />
26/.! 4784>=4! />! 0=>A957! 4784>=,J5/9J/>=! 87/2>=]43! I87!<br />
7\J?0;7!>X!/.7!947!>X!J5/9J/>=4!68!495.!J!87/2>=]!2>9;A!:7!<br />
/.7!947!>X!J!4784>=!/>!A7/75/!/.7!0=747857!>X!5J=:>8!?>8>\6A7!<br />
68!J!=>>?!J8A!/.7!947!>X!J8!J5/9J/>=!/>!0=>A957!J!;>9A!8>647!<br />
J;7=/680;7!/>!/.7!A7/75/6>8!>X!/.7!.J=?X9;!?:68J/6>8! >X! 4784>=4! J8A! J5/9J/>=4! 5J8! 78J:;7! >:@75/4! />!<br />
46?9;/J87>94;S! :7! J2J=7! >X! /.76=! 78[6=>8?78/! J8A! 68/7=J5/!<br />
26/.!07>0;7L!:>/.!J;4!>X!/.7!D>M3!<br />
D8!0=J5/65J;!68A94/=6J;!94JM!0=7478/;S!7?:=J574!<br />
NdDY,78J:;7A! 6A78/6X65J/6>8! J8A! /=J5]68;>[6A7! 49XX65678/! =7J;,/6?7!<br />
68X>=?J/6>8! J:>9/! /.68ML! 2.65.! J=7! [7=S! 947X9;! />!<br />
?J89XJ5/9=7=4L!A64/=6:9/>=4L!J8A!=7/J6;7=43!d>=!7\J?0;7L!NdDY!<br />
J00;65J/6>8! 68! 4900;S! 5.J68! ?J8J[7!<br />
68[78/>=S! ?J8J?7! 6A78/6X67A! JA[J8/J=! 5>4/L! 46?0;6X67A! :9468744! 0=>574474L! J8A!<br />
6?0=>[7A!7XX656785S3!<br />
d6[7!S7J=4!JL!6/!2J4!=70>=/7A!/.J/!-!07=578/!>X!Kj!:J47A!<br />
5>?0J8674! J=7! 9468![$#]*!<br />
:9/! 26/.! J! 4/=>82/.! 0>/78/6J;3! D8! NdDY,:J47A!<br />
J00;65J/6>84L! )'! 07=578/! >X! X6=?4! ! X>=! J55744! 5>8/=>;L! "k!<br />
07=578/! X>=! 4900;S! 5.J68L! ")! 07=578/! X>=! X=772JS! />;;4L! "B!<br />
07=578/! X>=! 4759=6/S! 5>8/=>;L! "$! 07=578/! X>=! 0=>A95/! 5>8/=>;L!<br />
J8A!$)!07=578/!X>=!J447/!?J8J;[7! [7=S! A6[7=46X67A! 40756X65J/6>84!<br />
68!/7=?4!>X!5>??9865J/6>8L!5>?09/J/6>8L! ?7?>=SL!J8A!AJ/J!<br />
4/>=J=! /=J84?6446>8! 5J0J56/6743! O8! D>M!<br />
J00;65J/6>8! 5>8464/4! >X! ?J8S! /S074! >X! A7[65743! O;;! /S074! >X!<br />
.J=A2J=7! A7[6574! 4.>9;A! :7! 27;;! >=9=]! J8A! :7! J557446:;7! [6J! J[J6;J:;7! 5>??9865J/6>83!<br />
MS065J;;SL! A7[6574! 5J8! :7! >=??9865J/6>8!09=0>47!>[7=!/.7!D8/7=87/3!<br />
D>M! 5J8! :7! J8! J94! 87/2>=]4L!<br />
495.! J4! WEF4L! 26=7;744! ?74.! 87/2>=]4L! ?>:6;7! 87/2>=]4L!<br />
J8A! W6=7;744! b>5J;;S! O=7J! F7/2>=]! (WbOF*3! M.747!<br />
87/2>=]4!.7;0!/.7!/.688,?J]68?09/J/6>8L! J8A! AJ/J! 7\5.J88L! /.7! =7;6J:;7! 5>??9865J/6>8! :7/2778! 8!26/.!=74075/!<br />
/>! D>M3! M.7! X! =98868?0;65J/7!<br />
>0/6?6`J/6>8! J;=6/.?! ;>5J;;S! :S! 7\0;>6/68=]!<br />
]8>2;7A?09/J/6>8J;! 5>?0;7\6/S! 64! 4.6X/7A! X=>?!<br />
/.68!/.7!0/6?J;!=>9/7!J8A!0J=J?7/7=!<br />
[J;974!X>=!/.7!:/J687A3!M.64!64!X7J46:;7!46857!<br />
/.7!46`7!>X!/.7!?J68!64!68!/.7!>=A7=!>X!J!X72!>X!/784!<br />
68!5>?0J=64>8!26/.!/.7!46`74!>X!/.68??9865J/6>8!<br />
=7c96=7?78/4! [J=S! X=>?! >87! A7[657! /S07! />! J8>/.7=3! M.7!<br />
/.68M! 5J8! .J[7! [7=S! A6XX7=78/! 5J0J:6;6/674! X>=!<br />
5>?09/J/6>8L! ?7?>=SL! 0>27=L! >=! 5>??9865J/6>83! d>=!<br />
684/J857L! J! 57;;9;J=! 0.>87! >=! J! /J:;7/! .J4! ?95.! :7//7=!<br />
5>??9865J/6>8!J8A!5>?09/J/6>8!5J0J:6;6/674!/.J8!J!40756J;,<br />
09=0>47! 7;75/=>865! 0=>A95/! 495.! J4! J! .7J=/! =J/7! ?>86/>=!<br />
2J/5.3!E6?6;J=;SL!/.68X! m9J;6/S! >X! E7=[657! (m>E*L! 68! 0J=/659;J=L! 68! /.7! J4075/4! >X!<br />
A7;JSL! 787=849?0/6>8L! J8A! =7;6J:6;6/S3! d>=! 7\J?0;7L!<br />
?686?6`68=! 5>??9865J/6>815>?09/J/6>8!<br />
09=0>474! 64! J! ?J@>=! 5>84/=J68/! X>=! /.7! :J//7=S,0>27=7A!<br />
A7[6574! 26/.>9/! 7XX65678/! 787=/! 5=6/65J;! X>=! /.7! A7[6574! 26/.! 0>27=!<br />
4900;S!5>8875/6>83!<br />
JA%D('"1+5%*F%8990,4"+,*#%8('"$%<br />
D>M! 64! J8! 68/7=5>8875/7A! 87/2>=]! >X! 4?J=/! >:@75/4! J8A!<br />
A7[65743! I:@75/4! 68! /.7! 2>=;A! >X! /.7! D8/7=87/! >X! M.684/;S! =7X7==7A! />! J4! 4?J=/! >:@75/4L! J=7! J9
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$$k! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
/.>47!0=>?0/7A!:S!/.7!87/2>=]68X!7[7=SAJS!>:@75/4h!(DMj!<br />
[']*3!<br />
M.7! D>M! 64! J! 5>8[7=X! [J=6>94! [646>84L! 7[>;[688! /.7! 4/J]7.>;A7=4_! X>594! >X! A7[7;>0?78/3!<br />
O/`>=6!7/!J;3![%]!6A78/6i74!/.=77!5;J4474!>X![646>84!26/.68!/.7!<br />
=747J=5.! 5>??986/S! >X! /.7! D>M+! M.68=678/7AL! D8/7=87/,<br />
>=678/7A! J8A! E7?J8/65,>=678/7A3! M.7! M.68=678/7A! [646>8!<br />
X>59474!>8!/.7!M.68X!M.68??9865J/6>84!(FdP*L!W6=7;744!E784>=!<br />
J8A! O5/9J/>=! F7/2>=]4! (WEOF*L! ?65=>0=>5744>=4L!<br />
7?:7AA7A! 4S4/7?4L! -Y! J8A! :J=5>A74! />=678/7A![646>8!64!5>857=87A!26/.!/.7!0=>/>5>;4!<br />
X>=!5>??9865J/6>8!87/2>=]43!E>?7!0=>/>5>;4!J=7!/.7!D8/7=87/!<br />
G=>/>5>;L! DG['! J8A! /.7! W7:! >X! M.68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $"#!<br />
!<br />
78>9!4900>=/!=>:94/!785=S0/6>83!M>!78J:;7!785=S0/6>8!>8!<br />
/.7!D>ML!J;=6/.?4!877A!/>!:7!?JA7!?>=7!7XX65678/!J8A!;744!<br />
787=849?688! 45.7?74!<br />
J=7!877A7A!(pJ8!J8A!W78![""]*3!!<br />
D8! JAA6/6>8! />! 785=S0/6>8L! 6A78/6/S! ?J8J=/J8/! 5>?0>878/! >X! J8S! 4759=6/S! ?>A7;! J8A! 986c97!<br />
6A78/6X67=4!J=7!74478/6J;!/>!D>M!A7[65743!M.747!6A78/6X67=4!?JS!<br />
:7! 947A! />! 74/J:;64.! 07=4>8J;! 6A78/6/674! J/! X68J856J;!<br />
684/6/9/6>84L!6A78/6XS!6;;7/.7=!X985/6>843!!<br />
M.94L!7849=68:@75/4!J=7!2.>!/.7S!4JS!/.7S!J=7!<br />
64!74478/6J;!/>!D>M!4955744!(QJ.J;;7!7/!J;3!["-]&!N>?J8!7/!J;3!<br />
["B]*3!<br />
G=6[J5S!!<br />
O4!?>=7!J8A!?>=7!>:@75/4!:75>?7!/=J57J:;7!/.=>9ML!<br />
/.=7J/4!/>!07=4>8J;!0=6[J5S!:75>?7!?>=7!47=6>943!D8!JAA6/6>8!<br />
/>! 4759=68! ?J]7! 49=7! /.J/! 6/! A>748_/! XJ;;! 68/>! /.7!<br />
2=>8X!AJ/J!>287=4.60!877A!/>!:7!JAA=7447A!<br />
68!>=A7=!/>!7849=7!/.J/!947=4!X77;!5>?X>=/J:;7!0J=/6560J/68M3!!<br />
M.94L!/.7!>287=4.60!>X!AJ/J!5>;;75/!7A!X=>?!4?J=/!>:@75/4!<br />
?94/!:7!5;7J=;S!74/J:;64.7A3!M.7!AJ/J!>287=!?94/!:7!J449=7A!<br />
/.J/! /.7! AJ/J! 26;;! 8>/! :7! 947A! 26/.>9/! .641.7=! 5>8478/L!<br />
0J=/659;J=;S! 2.78! /.7! AJ/J! 26;;! :7! 4.J=7A3! G=6[J5S! 0>;65674!<br />
5J8!:7!>87!J00=>J5.!/>!7849=68X!68X>=?J/6>83!<br />
E?J=/! >:@75/4! J8A! =7JA68M! 5J8! 7J5.! :7!<br />
7c96007A!26/.!0=6[J5S!0>;656743!W.78!/.7!>:@75/!J8A!=7JA7=!<br />
5>?7!68/>!5>8/J5/L!/.7S!5J8!7J5.!5.75]!/.7!>/.7=!_4!0=6[J5S!<br />
0>;65S!X>=!5>?0J/6:6;6/S!:7X>=7!5>??9865J/68?J8!7/!J;3!<br />
["B]*3!!<br />
LA%.4*#*/,4%D'#'F,+$%<br />
D8!J!49=[7S!>8!D>M!J8A!6/4!^75>8>?65!787=
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$"$! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
5>9;A!27;;!.J[7!0>46/6[7!75>8>?65!6?0J5/4!:9/!5>9;A!27;;!:7!<br />
87X!4759=6/S!:=7J5.74L!5S:7=!06=J5S!J8A!<br />
;>44! >X! 0=6[J5S! 0>/78/6J;;S! 68A95688>?65!<br />
AJ?J9;A! :7! 7[78! >9/! >X! 0=>0>=/6>8! 26/.! J!<br />
986[7=4J;! 45J;J:;7! J8A! 45J;68X! D>M3! O4! 68! /.7!<br />
75>8>?654!/=7J/?78/!>X!J556A78/!;J2!/.64!=7c96=74!J44744?78/!<br />
>X!J!A7;65J/7!:J;J857!>X!/=JA7>XX4!J8A!5>?0784J/6>83!!<br />
E9??J=6`68M! 0=>?6474! ?9;/60;7!<br />
75>8>?65!:787X6/4!495.!J4!!<br />
($*! ;>27=688! 5>4/4! X>=! [J=6>94! 68A94/=674! J8A!<br />
/.7! 78/6=7! 75>8>?SL! 6?0=>[68X! 0=>A95/4L! 47=[6574!<br />
X>=!5>849?7=4L!J8A!c9J;6/S!>X!;6X7!68!78[6=>8?78/J;!0=>/75/6>8!<br />
J8A!787=[646>8L!<br />
("*!78J:;68=?J857!68!J!:=>JA!<br />
47/! >X! [J;97! 8>?65! 2/.! J8A! 27;;! :76899/!5=7J/68:4!68!4>?7!68A94/=6743!<br />
MA%N;+;('%O,('4+,*#$%<br />
E6857! /.7! D>M! .J4! 8>/! S7/! :778! X9;;S! =7J;6`7A! s!<br />
5>8570/9J;;S!J8A!68!0=J5/65J;!6?0;7?78/J/6>84L!6/!?656>94! />! X>=75J4/! /.7! X9/9=7! A6=75/6>84L! /.7! 7\/78/! >X!<br />
68A94/=6J;! :=>JA7868X! D>M3! p7/L! /.7!<br />
?J89XJ5/9=68=! />8! /.7! W>=;A! K5>8>?65! d>=9?_4! (WKd*! "#$'! O889J;!<br />
Q77/684L! J;4>! 27! .J[7! .JA! J! 5>?0=7.7846[7!<br />
4.>25J47! J/! /.7! PKaDM! J8A! /.7! VJ88>[7=! D8A94/=6J;! dJ6=!<br />
("#$'L! "#$%*! 26/.! 5>?0J8674! 495.! J4! E67?784L! OaaL! D8/7;L!<br />
DaQL! V9J276L! tMKL! a>45.L! l9]JL! dJ895L! lJ2J4J]6! 68!<br />
0J=/87=4.60! 26/.! 4>X/2J=7! X6=?4! 4.>268@75/4L! J8A!<br />
7\0>468[7=! $##! 5J474! >X! =7J;,2>=;A! 6?0;7?78/J/6>84! >X!<br />
D8A94/=S!B3#3!d9/9=7![646>84!>X!/.7!D>M!26;;!JXX75/!6/4!59==78/!<br />
A7[7;>0?78/!J8A!?94/!/.7=7X>=7!:7!5>846A7=7A3!!<br />
I87! X9/9=7! [646>8! X>=! /.7! D>M! 64! /.7! W7:! >X! M.68X! M.680>474! /.7! 947! >X! 27:! 4/J8AJ=A4! />! X9;;S!<br />
68/7:@75/4! 68/>! /.7! W>=;A! W6A7! W7:! (WWW*3!<br />
j468;>=!A7[7;>07=4!/>!<br />
:96;A! J00;65J/6>84! 9468:@75/4! J8A! 7\64/68/>5>;4! 5J8! ?>=7! 7J46;S! 78J:;7! /.7! 68/7=>07=J:6;6/S! J8A!<br />
5>??9865J/6>8!>X!A6XX7=78/!A7[65743!O!?J4.90!64!J!W7:!"3#!<br />
5>8570/! 2.7=7! J8! J00;65J/6>8! 9474! AJ/J! J8A! X985/6>8J;6/S!<br />
X=>?!J![J=67/S!>X!27:!=74>9=5743!E>?7!=747J=5.7=4!0=>0>468X! M.68A7;! 49X!J00;65J/6>84!(R968J=A!7/!J;3![-"]*3!<br />
W6/.!/.7!X!/.7!W7:L!/.7!W>M!J8A!D>M!J=7!>8!<br />
/.7! 7AX! 7\07=678568AJSL! 27! J=7! >87!<br />
4/70! 5;>47=! />! /.64! [646>8! A97! />! ;J/74/! JA[J8574! 68! 27:!<br />
47=[6574L! 6A78/6i5J/6>8! /75.8>;>8[7==]4L! 2.65.! ?J]7! 5>??9865J/6>8! 5J0J:6;6/674!<br />
J8A! 0=>57446827=! J[J6;J:;7! 68! 685=7J468;[68! /.7! 4>,<br />
5J;;7A! ^W7:! >X! M.68M*L! J8! 78[6=>8?78/! 2.7=7!<br />
7[7=SAJS!A7[6574!495.!J4!/=JXi5!;6A6/674!J=7!=75>8/=>;;J:;7! [6J! /.7! D8/7=87/3! P7=/J68;SL! 4/J=/68?! J8! D8/7=87/! >X! 87J=;S! >87! /.>94J8A! ?6;;6>8! 5>?09/7=4L!<br />
/.7!W7:!8>2!/9=84!/>!:7!J8!D8/7=87/!>X!87J=;S!$##!:6;;6>8!>X!<br />
/.68=!A7[6574!0=74J?!J8!D>M!/>!J!W>M3!<br />
M.94L!68!J!W>M!0;J/X>=?!/.7!2>=];>JA!64!09/!J/!/.7!7\/=7?7!<br />
;7[7;!J8A!45J;J:6;6/S!64!J!5>?09;4>=S!=7c96=7?78/3!!<br />
M.64! A>74! 8>/! =7X7=! />! J8S! /75.8>;>=! J8S! 87/2>=]!<br />
4/=95/9=7L!:9/! >8;S! />! /.7! 6A7J! >X! 68/7=5>8875/688875/68?09/7=4! 26/.! /.7! D8/7=87/3! j47! >X!<br />
/.7! W7:! J4! /.7! 0;J/X>=?! .>4/684688875/7A!<br />
A7[6574! 5J8! :7! 7\0;J687A! :S! ?9;/60;7! /75.8>;>8>?65!:787i/4L!J!X72!>X!2.65.!685;9A7!A70;>S?78/L!.6X! 4/J8AJ=A6`7A!<br />
5>??9865J/6>8! 0=>/>5>;4! J8A! /.7! 75>4S4/7?! 5=7J/7A! /.J8]4!<br />
/>! W7:! "3#! 0J=JA6
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
! D8/7=8J/6>8J;!e>9=8J;!>X!a9468744!J8A!K5>8>?654!N747J=5.!"#$%&!'()*+!$$),$"-! $""!<br />
!<br />
899'#1,R2%H/"(+%!*/'%HE'0'+*#%O'$,-#%<br />
S%8#%B00;$+("+,7'%.R"/90'%%<br />
O! .>?7! :75>?74! 4?J=/! 6X! /.7! J55744! J8A! 5>8/=>;! >X! /.7!<br />
A7[6574!J8A!9/6;6/674!68!/.7!.>?7!4/J=/!/J]688!(27!<br />
J=7! 8>/! /J;]689/! O=/6X656J;! D8/7;;607=! 47/! >X!<br />
684/=95/6>84!64!!/.7?L!J8A!>8;S!6X!QJ5.687!/>!?J5.687!<br />
5>??9865J/6>8!64!.J007868!?J5.687!;7J=868446:;7! 6X! 4784>=! 87/2>=]4! 5>8/=>;! /.747! A7[65743!<br />
M75.8>;>/>5>;4L! 87/2>=]4! ;6]7! NdDYL! FdPL! t6>/.! J8A! [J=6>94! 4784>=4!<br />
(/7?07=J/9=7L! GDNL! 4?>]7L! J8A! .9?6A6/S*! 5J8! ?J]7! J!<br />
A75646>8,?J]682! 6;;94/=J/6[7;S!<br />
7\0;J68!/.7!5>??9865J/6>8!0=>57443!<br />
E?J=/!.>?7!87/2>=]3!<br />
W7!5J8!477!4?J;;!56=596/4!(4784>=4!J8A!4?J=/!56=596/4*!X6\7A!<br />
>8!J;;!/.7!A7[6574!J8A!J00;6J8574L!/.747!4784>=4!J8A!A75646>8,<br />
?J]68??9865J/7! 26/.! J55744! ! J! ;>5J;! =>9/7=! /.J/! 0=>[6A74!<br />
D8/7=87/! 5>8875/6>8! /.=>9!<br />
/.7!D8/7=87/3!<br />
E?J=/!.>?74!.J[7!/.7!X>;;>2688!?J]6827=!KXX65678/!O00;6J8574!<br />
V6!9/09/!<br />
J4! /.7! =7849?7! ?686?J;! 0>27=3!<br />
O00;6J8574! ;6]7! .627=! 4J[68! ?J]7! 4?J=/! 9/6;6`J/6>8! >X! 787=9/!5>?0=>?64688!5>?X>=/!;7[7;43!!<br />
-3!N7?>/7!O55744!<br />
M.7! ?J68! 09=0>47! >X! 4?J=/! .>?74! 64! />! J5.67[7! 5>8/=>;!<br />
>[7=!J00;6J8574!=7?>/7;S3!<br />
Q>4/!>X!/.7!/6?74!27!477L!07>0;7!X>=!426/5.!>XX!/.7!<br />
J00;6J8574!2.6;7!=94.68!2>=]!A97!/>!2.65.!/.7S!.J[7!/>!<br />
0JS! .628! >X!<br />
J00;6J857_4! A97! />! 7\57446[7! 94J! J! 986[7=4J;! 5>8/=>;J=7! J557446:;7! X=>?! J8S2.7=7!<br />
=7?>/7;S!26/.!>9=!4?J=/0.>874!J8A!5>?09/7=4L!/.7!947=!5J8!<br />
426/5.!/.7?!J4!.714.7!2J8/4!J8A!?>86/>=!/.76=!.>?73!!<br />
B3!E759=6/S!!<br />
!<br />
W7!:96;A!>9=!.>?7!/.7!2JS!27!2J8/L!J8A!27!J;4>!2J8/!6/!<br />
/>!4/JS!/.7!4J?7!2JSL!4>!4JX7/S!64!J;4>!/.7!0=6?J=S!5>857=8!68!<br />
J! 4?J=/! .>?73! d>=! 4JX7/S! X=>?! X6=7! 2J/7=! J8A! .60=6J/7! 4784>=4! 4.>9;A! :7! 684/J;;7A! />! J;7=/!<br />
94!68!5J474!>X!X6=7!:=7J]>9/!>=!2J/7=!;7J]J?7!<br />
5J8!J;4>!:7![9;87=J:;7!/>!=>::7=S3!M>!0=7[78/!>9=!.>947!X=>?!<br />
=A!/.7!X>>/J9A3! O;4>! />! =?7A! >8! J8S! 98J9/.>=6`7A!<br />
78/=674! J8A! 4940656>94! :7.J[6>=L! 68X=J=7A! J;J=?! 4S4/7?! (6/!<br />
5>9;A! :7! J! GDN! 4784>=! >=! :9=X!<br />
/.7! ;J/74/! /75.8>;>0;7*L! 4>! /.7! >07=J/6>8! >X!<br />
/.747!J00;6J8574! 4.>9;A!:7!7J46;S!98A7=4/J8AJ:;7!X>=!07>0;7!<br />
>X!J;;!J9:;74.>>/!>8!6/4!47;X3! M.7!68/7=XJ57!>X!5>8/=>;!4.>9;A!<br />
J;4>!:7!J4!46?0;7!J8A!X=678A;S!J4!0>446:;73!<br />
d>=! 6;;94/=J/6[7! 09=0>474L! 27! .J[7! A746A7;! />!<br />
A7?>84/=J/7! D8/7=87/! >X! M.68=! =7?>/7! ?>86/>=688/=>;;68X! J00;6J8574! =7?>/7;S! /.=>9X! /.64! ?>A7;! 64! />!<br />
0=>[6A7!8>80.S465J;!5>8/=>;!>[7=!J00;6J8574!J8A!/>!8! /.7! /7?07=J/9=7! J8A! .9?6A6/S! ;7[7;4! J/! .>?7!<br />
/.=>9?!J8S2.7=7!68!/.7!2>=;A3!M.64!<br />
]68A! >X! J! ?>A7;! .7;04! /.7! 0.S465J;;S! 5.J;;78! 5>8/=>;! J00;6J8574! J4! 6/! ?6=! /.7?! />! ?>[7! />! /.7! 426/5.! 7[7=S! /6?7! /.7S! 2J8/! />!<br />
5>8/=>;! 6/3! Q>=7>[7=L! /.7! 0=>0>47A! ?>A7;! 64! 5>4/! 7XX75/6[7!<br />
J8A!0>27=!7XX65678/3!!<br />
M.7! 5>8570/! >X! D8/7=87/! >X! M.68[7A! :S! 5=7J/6807=! 68/7=XJ57! J8A! A746??9865J/7! 26/.! /.7!<br />
D8/7=87/3!O[J6;J:;7!>A4!J8A!c9J8/6/S!5J8!:7!]8>28!68!J!<br />
468:@75/4! J8A! >A4! J=7! 98A!<br />
>:@75/4!5J8!:7!/=J5]7A!94688570/!:S!/=J5]680=6J/7! 4784>=4! X6\7A! />! /.7?3! O! ?>:6;7!<br />
J00;65J/6>8!5J8!:7!5=7J/7A!/>!5>8/=>;!/.747!J00;6J8574!X=>?!<br />
>9=! 4?J=/! 0.>8743! aS! J==J886/>=7A!X=>?!<br />
J8S2.7=7! :S! ;>;>94!J00;65J/6>84!X>=!<br />
7J4S!J557446:6;6/S3!<br />
6'F'('#4'$%<br />
u$v!<br />
u"v!<br />
u-v!<br />
uBv!<br />
K[J84L! G3! P3! J8A! O8898`6J/JL! Q3! ("#$"*3! D8A94/=6J;! D8/7=87/L!<br />
R787=J;!K;75/=65!(RK*!N70>=/L!F72!p>=]3!<br />
O;0JSA68L! K3! ("#$'*3! QJ5.687! b7J=868
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
HANS W. GOTTINGER<br />
$"-! VJ84!W7=87=!R>//68X!M.688>?65!J8A!D8A94/=6J;!Y6?7846>84! !<br />
!<br />
u)v!<br />
W.6/?>=7L! O3L! O=6L!b3L!D7=JL!O3!J8A!Q>=J:6/>L!R3!("#$#*3!hM.7!D8/7=87/!>X!<br />
M.68?09/7=!F7/2>=]4L!)B($)*L!"%C%s"C#)3!<br />
d68?JL!D3L!E6?07=;L!K3!J8A!V785.L!R3!("##k*3!gO!@>68/!=>JA?J0!<br />
X>=! E7?J8/65! /75.8>;>X! M.68;>=J8A6L! Y3L! E65J=6L! E3L! Y7! G7;;7X! /.688L! J00;65J/6>84! J8A! =747J=5.!<br />
5.J;;785!F7/2>=]4L!$#!(%*L!$Bk%s$)$'3!<br />
u$Bv! b6L!b3L!J8A!b69L!e3!("#$"*3!gO8!7XX65678/!J8A!X;7\6:;7!27:!<br />
47=[6574,:J47A! ?9;/6A64560;68J=S! A7460/6?64J/6>8!<br />
X=J?72>=]! X>=! 5>?0;7\! 788!ES4/7?4L!'(-*L!-B)s-%$3!<br />
u$)v! E5.JXX7=4L!V3L!l>?868>4L!F3L!GJ;;>/L!Q3L!M=>9447L!a3L!F6;44>8L!<br />
Q3L! J8A! I;6[76=JL! O3! ("#$$*3h! E?J=/! 56/674! J8A! /.7! X9/9=7!<br />
68/7=87/+! M>2J=A4! 5>>07=J/6>8! X=J?72>=]4! X>=! >078!<br />
688>[J/6>8hL! 68! Y>?68X! M.68=! J?:678/!<br />
J4464/7A! ;6[68577A68X! /.7! E7[78/.! D8/7=8J/6>8J;!<br />
P>8X7=7857! >8! D8X>=?J/6>8! M75.8>;>84!<br />
(DMFR*3!<br />
u$Cv! d;n84/=95/68??986,! 5J/6>84!<br />
68!5>?09/7=!J8A!68X>=?J/6>8!4567857!(Z>;3!-"*L!$'s"'3!a7=;68+!<br />
E0=68874L! e3! Y3! ("#$"*3! g7,<br />
a9468744! 4S4/7?4! 68/78+! J! 4S4/7?4! 07=4075/6[7hL!<br />
D8X>=?J/6>8!M75.8>;>84! =747J=5.! (IN*! 68! 47=[657!<br />
68A94/=674+! J! 5>?0=7.7846[7! =7[672hL! ES4/7?4! N747J=5.! J8A!<br />
a7.J[6>=J;!E567857L!-#!(-*L!-##s-)-3!!<br />
u""v! pJ8L!M3!J8A!W78L!m3!("#$$*3!ga96;A68X!M.68/>5>;!:J47A!>8!68X>=?J/6>8!<br />
/75.8>;,! >?09/3! E563! D8/7;;3! ES4/3! K8[6=>83! $#BL!<br />
$B-s$Bk3!!<br />
u"-v! QJ.J;;7L! G3L! aJ:J=L! E3L! G=J4JAL! F3! N3! J8A! G=J4JAL! N3!<br />
("#$#*3hDA78/6/S! ?J8J=]! />2J=A4! D8/7=87/! >X!<br />
M.68M*+!N>JA?J0!J8A!]7S!5.J;;78?J8L! N3L! FJ@7=JL! G3! J8A! b>07`L! e3! ("#$$*3hE759=68X!M.68?09/7=4L!BB(k*L!)$s)C3!<br />
u")v! d;7645.L! K3! ("#$-*3! gW.J/! 64! /.7! D8/7=87/! >X! /.68?!.//0+!1!1!2223!6?3!K/.`3!5.!1!<br />
7A95J/6>8!1!VE$#!1!OjMIDYbOaE,WG,aDtOGG,)-30AX3!!<br />
u"'v! R>//68X! R>[7=8J857L!<br />
F72!p>=]+!I\X>=A!j86[3!G=7443!<br />
u"kv! R>//68X! P6/6743!<br />
M.7! 68/7=87/! >X! K[7=S/.682! 27! ;6[7hL! 68!<br />
E0756J;!D4497!>X!d>=76X!M.68!/.7!W7:!>X!M.689=57,<br />
>=678/7A! J=5.6/75/9=7! J8A! :74/! G=J5/6574hL! 68! j5]7;?J88L! Y3L!<br />
VJ==64>8L! Q3! J8A! Q65.J.7;;74L! d3L! KA43! ("#$#*3! O=5.6/75/68X!M.68
HANS W. GOTTINGER<br />
1.8 INTERNET OF THINGS ECONOMIC AND INDUSTRIAL DIMENSIONS<br />
132
2.<br />
MARKETS<br />
AND<br />
COMPETITION
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
!"#$%&'&()*+),-(%.&(()/&(&0"#$)1)2*3456)7*48)<br />
9-:3%#0;%*.)
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"8!4='!B4,C8#!:#,A@#A,8)!I@"AB$8#8,!E4:!+B$,8::8'!5;!#"8!JA4D+#4#+98!'+4#+*=!*8!<br />
+=!,8:$8@#+98!+='A:#,+8:!SI@"8,8,?!.X30?.XX6T)!<br />
!<br />
G*,! "+-"U#8@"! +='A:#,+8:! E+#"! 4! -D*54D!
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!A!!<br />
!<br />
M80:! *$+! *&+5:! &,! 8&48G('%8! &,.0:(+&':! %$,(&,0$0:! &,,$9-(&$,! &:! (8'! ,-5'! $*! (8'! 4-5'! -:!<br />
%$5K'(&(&$,!&,!&,,$9-(&$,!.+&9':!(8'!$0(%$5':!&,!4+$S(8!1!S'@*-+'!%8-,4':!-,.!:0+9&9-@A!<br />
!<br />
V&(8$0(! &(! :(+0%(0+-@! 4+$S(8! &,! 5-+W'(:! S$0@.! ,$(! F'! K$::&F@'! S8'(8'+! &,! K+$.0%(!<br />
('%8,$@$4&':1!:'+9&%':!$+!$K'+-(&$,-@!K'+*$+5-,%'A!C,,$9-(&$,!F-:'.!*&+5!4+$S(8!S$0@.!F'!-!<br />
W')!&,.&%-($+!*$+!-!*&+5T:!:0:(-&,-F@'!*0(0+'A!<br />
!<br />
7*('+!+':$@9&,4!(8'!&::0'!8$S!-,.!S8)!@-+4'!*&+5:!-+'!-%(&9'@)!',4-4'.!&,!&,,$9-(&9'!-%(&9&()!-,!<br />
&5K@&%-('.! ,'L(! :('K! %$0@.! F'! 8$S! (8')! %$,(&,0'! ($! F'! &,,$9-(&9'! $+4-,&:-(&$,:A! 2,.'+! (8'!<br />
K+':05K(&$,!(8-(!@-+4'!*&+5:!F'%-5'!T@-+4'T!F'%-0:'!$*!T4+$S&,4!(8+$048!&,,$9-(&$,T!(8'+'!&:!<br />
(8'!X0':(&$,!8$S!(8')!W''K!$,!F'&,4!&,,$9-(&9'!*$+!(8'&+!@&*'(&5'A!<br />
!<br />
M8+$048!:(0.&':!$*!&,,$9-(&$,!&,!&,.0:(+&-@!$+4-,&Y-(&$,:!S'!@'-+,!(8-(!+$048@)!(S$G(8&+.!$*!<br />
&,,$9-(&$,:!-+'!$*!(8'!K+$.0%(!&5K+$9'5',(!()K'!S8&@'!$,'G(8&+.!&:!$*!(8'!K+$%'::!&,,$9-(&$,!<br />
()K'!=%$:(!+'.0%&,4!$+!K+$.0%(&9&()!',8-,%&,4>A!V'!5-W'!0:'!$*!(8&:!$F:'+9-(&$,!F)!-::&4,&,4!<br />
(8'!*&+:(!W&,.!($!K+&9-('!$+4-,&Y-(&$,:!=&,.0:(+&-@!*&+5:>!S8&@'!(8'!:'%$,.!($!-!K0F@&%!:'+9&%'!<br />
%',('+'.!$+4-,&Y-(&$,1!.':&4,'.!-,.!+'40@-('.!-:!:0%8A!<br />
!<br />
!*+,-././*012+345-.26.47!.74-23082/00*93./*02<br />
V'! *$%0:! 8'+'! $,! :$5'! W')! +':0@(:! *+$5! F'8-9&$+-@! '%$,$5&%! 5$.'@:! &,! &,.0:(+&-@!<br />
$+4-,&Y-(&$,=! CZ>! (8-(! @'-9'! *$$(K+&,(:! *$+! 'LK@$+&,4! (8'! &,,$9-(&$,! .+&9'! -,.! %-K-F&@&()! $*!<br />
*&+5:!&,!K-+(&%0@-+!&,.0:(+)!:'((&,4:A!!<br />
!<br />
C,!%8-+-%('+&Y&,4!-,.!K$+(+-)&,4!(8'!&,('+,-@!'%$,$5)!$*!@-+4'!*&+5:!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"!#"6!:*7'+#+*79!-*A6,7+7-!896>!47'!9*;6!:"47:6)!<br />
!<br />
B"6! :?499+:4?! $*+7#! *!<br />
*76! C*8?'! *596,A6! ;*,6! +77*A4#+*7! +7! #"*96! +7'89#,+69! #"4#! 4,6! :"4,4:#6,+O6'! 5@! 4! "+-"6,!<br />
'6-,66!*!4?#"*8-"!7*!:4894?+#@!9"*8?'!56!+;$8#6'!#*!#"+9!,6?4#+*79"+$)!<br />
!<br />
N*;6!;*'6?9!:*79+'6,!#"6!,6?4#+*79"+$!56#C667!$,*'8:#!;4,=6#!9#,8:#8,6!D+)6)!:*7:67#,4#+*7E!<br />
47'! +77*A4#+*7)! ! S7! #"6! :*7#6(#! *! #"6! $,*545+?+#@! *!C"67!+;+#4#+*7!+9!<br />
7*#!$*99+5?6>!47!+7:,6496!+7!#"6!78;56,!*
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!! .$5&,-(':! (8'! *&+5:R! .'%&:&$,A! Y8',! :0%%'::! &:! :($%8-:(&%1! 8$N'9'+1! (8'! (8+'-(!<br />
*+$5!(8'!+&9-@!&,,$9-(&,4!&:!@'::!-%0('A!C,!(8'!%-:'!$*!.+-:(&%!&,,$9-(&$,:1!(8'!%$5O'(&(&9'!(8+'-(!<br />
&:!(8'!:-5'!*$+!F$(8!*&+5:!-,.!&(!&:!(8'!O+$*&(!&,%',(&9'!=N8&%8!&:!@-+4'+!*$+!%8-@@',4'+:>1!,'Z(!($!<br />
.&9'+:'!,$,G5$,'(-+)!&,%',(&9':1!(8-(!.'('+5&,':!(8'!@'9'@!$*!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"?!4!,:?@9#A!4!D+':!;4,+:#8!*99!#":?:!#":*,+:?!?"4,:!#":!=*BB*7!<br />
4+B! *
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"! 94,! -,74#7,!<br />
7=*! #"7! 7:$"48+8! *!4!.P31>!N"4$)!2E)!<br />
!<br />
H,4'+#+*!*
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!A!<br />
!<br />
S8':'! %$,:&.'+-(&$,:! :044':(! (8-(! &(! &:! 0:'*0@! ($! .&:(&,40&:8! F'(L'',! .&**'+',(! '%$,$5&%!<br />
',9&+$,5',(:1! -%%$+.&,4! ($! (8'! +'@-(&9'! &5N$+(-,%'! $*! %$5N'(&(&$,! &,! &,,$9-(&$,! -,.!<br />
%$5N'(&(&$,!-(!(8'!N+$.0%(!5-+M'(!@'9'@!(8-(!%-,!F'!'TN'%('.!&,!(8'!5-+M'(A!S8'!@&('+-(0+'!$,!<br />
,'(L$+M!&,.0:(+&':!&:!N-+(&%0@-+@)!0:'*0@!&,!'TN@-&,&,4!L8)!-!5-+M'(!5-)!'T8&F&(!L&,,'+G(-M':G<br />
-@@!N+$N'+(&':A!<br />
!<br />
4,)5"%/2677,8)02<br />
U'(L$+M! '**'%(:! 5-)! F'! -,! &5N$+(-,(! *-%($+! &,! .'('+5&,&,4! (8'! %$5N'(&(&9'! ',9&+$,5',(A! 7!<br />
=N$:&(&9'>! .&+'%(! ,'(L$+M! 'T('+,-@&()! 'T&:(:! L8',! (8'! .'5-,.! *$+! -! :'+9&%'! &,%+'-:':! F)! -,!<br />
'T(+-! 0,&(A! C,! $+.'+! *$+! (8&:! '**'%(! ($! F'! O.&+'%(P! (8'! +'-:$,! *$+! (8'! &,%+'-:'! &,! .'5-,.! 50:(!<br />
%$5'!.&+'%(@)!*+$5!(8'!-..&(&$,-@!%$,:05N(&$,!$*!(8'!:'+9&%'!&,!V0':(&$,1!L&(8$0(!,''.!*$+!-!<br />
:(+-('4&%! +':N$,:'! F)!:0NN@&'+:!&,!(8'!:-5'!$+! +'@-('.! 5-+M'(A! E$+! 'T-5N@'1! (8'! .'5-,.! *$+!<br />
N0F@&%! ('@'N8$,)! :'+9&%':! =&A'A! (8'! .'5-,.! *$+! :0F:%+&N(&$,:! ($! ,'(L$+M:>! 5-)! &,%+'-:'! -:!<br />
5$+'!0:'+:!%$,:05'!(8&:!N+$.0%(!='A4A!',('+!&,($!:0F:%+&N(&$,!-4+''5',(:!L&(8!N0F@&%!,'(L$+M!<br />
$N'+-($+:>!:&5N@)!F'%-0:'!5$+'!0:'+:!-@@$L!-!4+'-('+!,05F'+!$*!N$(',(&-@!%$,,'%(&$,:!($!F'!<br />
5-.'!$*!(8'!,'(L$+MA!<br />
!<br />
;)! %$,(+-:(1! -! =N$:&(&9'>! &,.&+'%(! ,'(L$+M! 'T('+,-@&()! 'T&:(:! L8'+'! (8'! .'5-,.! *$+! -! :'+9&%'!<br />
:0NN@&'.!&,!$,'!+'@'9-,(!5-+M'(!8-:!-!:&4,&*&%-,(!'**'%(!$,!:0NN@)!$*!-,$(8'+!N+$.0%(!=()N&%-@@)!<br />
&,! -! .&**'+',(! +'@'9-,(! 5-+M'(>! :0%8! (8-(! (8&:! &,! (0+,! &,%+'-:':! .'5-,.! &,! (8'! *&+:(! +'@'9-,(!<br />
5-+M'(A!<br />
!<br />
7,!'T-5N@'!$*!&,.&+'%(!,'(L$+M!'**'%(:!&:!(8-(!$*!%$5N0('+!:$*(L-+'!*$+!-!N-+(&%0@-+!$N'+-(&,4!<br />
:):('5!N@-(*$+5A!7,!&,%+'-:'!&,!%$,:05N(&$,!&,!(8'!5-+M'(!*$+!:0NN@)!$*!$N'+-(&,4!:):('5:!&:!<br />
@&M'@)! ($! F+&,4! F','*&(:! ($! (8'! 5-+M'(:! *$+! :0NN@)! $*! %$5N-(&F@'! :$*(L-+'1! F)! 'TN-,.&,4! (8'!<br />
5-+M'(!:0%8!(8-(!-!4+'-('+!9-+&'()!$*!:$*(L-+'!&:!$**'+'.!$+!(8'!0,&(!N+&%'!.'%+'-:':!=:&,%'!*&T'.!<br />
%$:(:!$*!:$*(L-+'!.'9'@$N5',(!%-,!F'!:N+'-.!$9'+!-!4+'-('+!,05F'+!$*!N$(',(&-@!F0)'+:>A!C,!<br />
(0+,1!(8'!F','*&(:!($!(8'!:$*(L-+'!5-+M'(:!+',.'+!(8'!:-5'!$N'+-(&,4!:):('5!5$+'!9-@0-F@'1!<br />
-,.! (80:! &,%+'-:'! .'5-,.! *$+! &(A! ;0(! (8'! '**'%(! &:! &,.&+'%(! F'%-0:'! &(! +'@&':! $,! -! :(+-('4&%!<br />
+':N$,:'!$*!:$*(L-+'!:0NN@&'+:!($!(8'!&,%+'-:'!&,!(8'!%$,:05N(&$,!$*!$N'+-(&,4!:):('5:A!<br />
!<br />
9"%&:"+).12.+;2
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*""6,65!47'!<br />
56##6,! $,*'?:#9! '+9$84:6! *8'6,! *59*86#6! $,*'?:#9! =,*A! #"6! A4,F6#)! O7! A4,F6#9! 9?5R6:#! #*!<br />
;6,#+:48!'+==6,67#+4#+*7@!4!=+,A!B*,!A*,6!4::?,4#68!#"6!$4:6!*=!<br />
#6:"7*8*-+:48!4';47:6!A4I!6(+9#9@!*?#9+'6!>"+:"!<br />
=+,A9!:477*#!9?,;+;6!+7!#"6!A4,F6#C)!N"+9!6:*7*A+:!$,*:699!:47!-?+'6!*7!#"6!74#?,6!*=!A4,F6#!<br />
'+96!<br />
+77*;4#+*7!A4*,96V!!N"+9!:477*#!56!<br />
479>6,6'!4#!#"6!-676,+:!86;68@!-+;67!#"6!:?,,67#!86;68!*=!?7'6,9#47'+7-!*=!#"696!6==6:#9)!W47"+:"! A4
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
7+%8&9':!$*!;0:&,'::!! ($! -%%'::!<br />
'R('+,-@! :$0+%':! $*! M,$N@'.4'! ='A4A! :%&',(&*&%! M,$N@'.4'>S! 8-9&,4! -%%'::! ($! :$5'! (-%&(!<br />
M,$N@'.4'!N8&%8!&:!,$(!(+-,:*'+-E@'!=L'+8-L:!E'%-0:'!'5E$.&'.!&,!$+4-,&O-(&$,-@!+$0(&,':>S!<br />
-,.!8-9&,4!-%%'::!($!:$5'!:L'%&*&%!%$5L@'5',(-+)!+':$0+%':1!'A4A!:L'%&*&%!&,L0(!*-%($+:A!<br />
!<br />
*+,&)-./"0123'2-<br />
J8'!%-L-E&@&()!$*!-!*&+5!($!&,,$9-('!&,!L-+(&%0@-+!-+'-:!N&@@!E'!-**'%('.!E)!(8'!'R(',(!$*!(-%&(!<br />
M,$N@'.4'! :0LL$+(&,4! &(:! &,,$9-(&9'! -%(&9&()1! -,.! N8'(8'+! (8'! *&+5! 8-:! -%%'::! ($! (8&:!<br />
M,$N@'.4'!='A4A!-%T0&+'.!(8+$048!-!L+$%'::!$*!@'-+,&,4FE)F.$&,4>A!<br />
!<br />
J-%&(!M,$N@'.4'!%-,,$(!E'!%$.&*&'.1!(+-,:*'++'.!$+!&5&(-('.!E)!'R('+,-@!L-+(&':!E0(!&:!+-(8'+!<br />
'5E$.&'.!&,!+$0(&,':!-,.!:M&@@:!$*!(8'!*&+5A!J8'!('+5!U(-%&(!M,$N@'.4'K!&:!0:'.!($!',%-L:0@-('!<br />
(8'! &.'-! $*! M,$N@'.4'! (8-(! &:! *$0,.! ($! E'! 0:'*0@! ='A4A! &,! -,! $+4-,&O-(&$,-@! +$0(&,'! $+! -! :M&@@>!<br />
N&(8$0(! E'&,4! -+(&%0@-('1! -,.! (8'+'*$+'! N&(8$0(! E'&,4! .&+'%(@)! (+-,:*'+-E@'! ='A4A! :$@.! -:!<br />
&,*$+5-(&$,>A!<br />
!<br />
455678*465-859-:;8A!<br />
!<br />
Y'!:8$0@.!-%%'L(!(8-(!(8&:!.&:(&,%(&$,!&:!,$(!-E:$@0('A!/-%8!&,,$9-(&$,1!&,!'9'+)!5-+M'(1!N$0@.!<br />
4','+-@@)!&,9$@9'!E$(8!'@'5',(:!$*!&,9',(&$,!-,.!-.$L(&$,A!D'N!&,,$9-(&9'!&5L+$9'5',(:!N&@@!<br />
E'!4',0&,'@)!0,&T0'!($!(8'!&,,$9-(&,4!*&+5!:&,%'!(8')!N&@@!-@N-):!&,9$@9'!-!.'4+''!$*!&5&(-(&$,!<br />
-,.! -.$L(&$,! *+$5! $E:'+9-(&$,! $*! (8'! N$+@.! $0(:&.'! $*! (8'! *&+5A! 7,)! &,9',(&$,! N&@@! .+-N! ($!<br />
:$5'!'R(',(!$,!L-:(!&,,$9-(&$,:!-,.!M,$N@'.4'A!P$,9'+:'@)1!'9',!N8'+'!-!*&+5!:''5:!($!E'!<br />
.&+'%(@)! -.$L(&,4! &,,$9-(&$,:! -::$%&-('.! N&(8! &,9',(&9'! '**$+(! E)! $(8'+! $+4-,&O-(&$,:1! &(! N&@@!<br />
,''.!($!-.-L(!(8':'!($!&(:!$N,!%8-+-%('+&:(&%:!-,.!%&+%05:(-,%':1!N8&%8!+'T0&+':!-,!'@'5',(!$*!<br />
&,9',(&$,A!/9',!N8'+'!&,,$9-(&$,!&:!@-+4'@)!.+&9',!E)!&,9',(&9'!&5L+$9'5',(:!&,!(8'!:'+9&%':!<br />
L+$9&.'.!E)!:0LL@&'+:1!*&+5:!0:&,4!:0%8!:'+9&%':!5-)!,''.!:$5'!%+'-(&9'!'**$+(!($!&,%$+L$+-('!<br />
(8':'!&,($!(8'&+!L+$.0%(&$,!L+$%'::':!-,.!L+$.0%(:A!J8&:!9&'N!&:!0,.'+@&,'.!E)!%$55',(:!*+$5!<br />
Z'@:$,! =6[[I1! LAGBB>! *$@@$N&,4! -,! 'R(',:&9'! %+$::F%$0,(+)! '5L&+&%-@! :(0.)! $,! &,,$9-(&$,Q!<br />
\](8'!E0@M!$*!(8'!'**$+(!&,,$9-(&$,!,''.:!($!E'!.$,'!E)!(8'!*&+5:!(8'5:'@9':A!Y8&@'!(8')!5-)!<br />
.+-N! $,! $0(:&.'! .'9'@$L5',(:1! :&4,&*&%-,(! &,('+,-@! '**$+(! -,.! :M&@@! &:! ,''.'.! ($! %$5L@'5',(!<br />
-,.!&5L@'5',(!(8':'A^!<br />
Z$,'(8'@'::1!N'!%-,!%$,%'&9'!$*!.&:(&,40&:8&,4!(8'!'R(',(!($!N8&%8!(8'!&,,$9-(&$,!-::$%&-('.!<br />
N&(8!:0LL@)!$*!-!*0(0+'!5-+M'(!&:!-.$L(&9'!+-(8'+!(8-,!&,,$9-(&9'1!'9',!&*!E$(8!'@'5',(:!50:(!<br />
E'!L+':',(!($!:$5'!.'4+''A!<br />
!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 66!<br />
145
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"!4!?@,#";,!'+!
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
8+%9&:';!$*!(9'!M+&5-+)!5-+K'(?B!<br />
!<br />
8! %$55$,A)! %&('.! 'O-5MA'! &;! (9-(! $*! M+&,('+;! -,.! %$5M-(&FA'! &,KG+'*&AA! %-+(+&.4';1! L9'+'!<br />
.'5-,.! *$+! (9'! A-(('+! &;! .'+&:'.! ',(&+'A)! *+$5! %$,;05M(&$,! $*! (9'! *$+5'+B! D,! (9&;! %-;'1! (9'!<br />
;%$M'! *$+! %$5M'(&(&$,! &,! &,,$:-(&$,! .&+'%('.! ($! +'*&AA! %-+(+&.4';! L&AA! F'! -**'%('.! F)! L9'(9'+!<br />
%-+(+&.4'!5-,0*-%(0+'+;!%-,!&,('+-%(!;0%%';;*0AA)!L&(9!(9'!M+&,('+;!5-+K'(B!<br />
!<br />
8! A';;! $F:&$0;! 'O-5MA'! &;! 5-+K'(;! %$,%'+,'.! L&(9! -.:'+(&;&,4! >(9'! ;0MMA)! $*! -.:'+(&;&,4!<br />
;M-%'?B!D,!5$;(!%-;';1!(9'!'O&;(',%'!$*!(9'!-.:'+(&;&,4!5-+K'(!&;!.'M',.',(!$,!%$,;05M(&$,!&,!<br />
-!+'A-('.!%$,(',(!5-+K'(!>F'!(9-(!(9'!;0MMA)!$*!('A':&;&$,!;'+:&%';1!(9'!;0MMA)!$*!,'L;M-M'+;1!<br />
5-4-P&,';1!$+!(9'!;0MMA)!$*!%$,(',(!$,!(9'!D,('+,'(?B!8,-A);&;!$*!%$5M'(&(&$,!&,!(9'!&,,$:-(&:'!<br />
-%(&:&()!+'A-('.!($!',(+)!&,!(9'!-.:'+(&;&,4!5-+K'(;!50;(!M-)!-((',(&$,!($!(9'!,''.!*$+!(9'!*&+5!<br />
($! ;&50A(-,'$0;A)! $M'+-('! ;0%%';;*0AA)! &,! (9'! %$,(',(! 5-+K'(;B! #$! (9'! +'A-(&$,;9&M! F'(L'',!<br />
-.:'+(&;&,4!+'A':-,(!5-+K'(;!-,.!%$,(',(!+'A':-,(!5-+K'(;!*-AA;!0,.'+!(9'!%-('4$+)!$*!-.N0,%(!<br />
5-+K'(;B!<br />
!<br />
Q*!M-+(&%0A-+!&5M$+(-,%'!&;!L9'+'!-!%$5M'(&(&:'!M+&5-+)!5-+K'(!&;!-;;$%&-('.!L&(9!-,!-.N0,%(!<br />
5-+K'(!&,!L9&%9!(9'+'!&;!5$,$M$A)B!R90;!*&+5;!%$5M'('!&,!-!F+$-.!M+&5-+)!5-+K'(!-,.!'-%9!<br />
-%9&':';!5$,$M$A)!;0MMA)!&,!-!,-++$L!-.N0,%(!5-+K'(!>M+&,('+;!-,.!+'*&AA!%-+(+&.4';?B!!D,!;0%9!<br />
%-;';!L'!5-)!'OM'%(!*&+5;!($!0;'!-,)!5-+4&,-A!&,%+'-;'!&,!M+$*&(;!+'-A&P'.!&,!(9'!;'%$,.-+)!<br />
5-+K'(! ($! *0,.! -! .'%+'-;'! &,! M+&%'! &,! (9'! %$5M'(&(&:'! M+&5-+)! 5-+K'(! -,.! (9'+'F)! &5M+$:'!<br />
(9'&+! +'A-(&:'! M$;&(&$,! -4-&,;(! %$5M'(&($+;B! S$+'$:'+1! L'! L$0A.! 'OM'%(! (9'+'! ($! F'! %9'%K;!<br />
-4-&,;(!%'+(-&,!()M';!$*!-,(&G%$5M'(&(&:'!F'9-:&$+!&,!(9'!5$,$M$A)!A&,K'.!5-+K'(;!.0'!($!(9'!<br />
(+-,;*'+!$*!%$5M'(&(&:'!'**'%(;!*+$5!(9'!M+&5-+)!5-+K'(B!<br />
!<br />
R90;! &,! -.N0,%(! 5-+K'(! %-;';1! (9'! '**'%(! 5-)! F'! ;0%9! (9-(! &,,$:-(&:'! -%(&:&()! +'A-('.! ($! (9'!<br />
;'%$,.-+)! 5-+K'(;! L&AA! .'M',.! $,! (9'! -F&A&()! $*! (9'! *&+5! ($! ',('+! (9'! M+&5-+)! 5-+K'(!<br />
;0%%';;*0AA)T! $+! ! (9'! -F&A&()! $*! (9'! *&+5! ($! U4-&,! -%%';;V! ($! (9'! ,'%';;-+)! A&,K;! F'(L'',! (9'!<br />
M+&5-+)!-,.!!A&,K'.!5-+K'(;B!<br />
!<br />
.6)7"%804996:);0<br />
R9'+'! ;''5! ($! F'! %A'-+! ;&5&A-+&(&';! F'(L'',! -.N0,%(! 5-+K'(;! -,.! ,'(L$+K! '**'%(;1! F0(! (9'!<br />
*$+5'+!&;!.'*&,'.!-F$:'!-;!L9'+'!%$,;05M(&$,!$*!-!M+$.0%(!&,!$,'!5-+K'(!%+'-(';!.'5-,.!&,!<br />
-,$(9'+1! L9&A'! ,'(L$+K! '**'%(;! -+'! 5$+'! +'A-('.! ($! 9$L! %$,;05M(&$,! $*! -! M+$.0%(! &,! $,'!<br />
5-+K'(!&,%+'-;';!.'5-,.!*$+!'&(9'+!>&?!(9-(!M+$.0%(!$+!>&&?!-!M+$.0%(!&,!-!.&**'+',(!5-+K'(!(9-(!<br />
&,! (0+,! 5-)! A'-.! ($! &,%+'-;';! &,! .'5-,.! *$+! (9'! $+&4&,-A! M+$.0%(! >;''! W$((&,4'+! >HII7?! *$+!<br />
.&;%0;;&$,!$*!,'(L$+K!'**'%(;?B!<br />
84-&,!,'(L$+K!'**'%(;!9-:'!&5M$+(-,(!&5MA&%-(&$,;!*$+!(9'!%-M-F&A&()!$*!*&+5;!($!;0MMA)!*0(0+'!<br />
5-+K'(;! -;;$%&-('.! L&(9! &,,$:-(&$,B! #0MMA)! ($! (9'! *0(0+'! 5-+K'(! L&AA! +'X0&+'! -! ;(+-('4)! *$+!<br />
&,('+-%(&,4!L&(9!(9'!,'(L$+K!'**'%(;!&,!;0%9!-!L-)!-;!($!5-K'!(9'!',(+)!;0%%';;*0A!-,.!;0;(-&,!-!<br />
M$;&(&$,! &,! (9'! 5-+K'(B! D,! M-+(&%0A-+1! ,'(L$+K! '**'%(;! %-,! 4','+-('! &5M$+(-,(! F','*&(;! *$+!<br />
&,%05F',(! ('%9,$A$4&';! -,.! (90;! %+'-('! &,('+G('5M$+-A! A&,K;! F'(L'',! %0++',(! -,.! *0(0+'!<br />
5-+K'(;1!-;!.&;%0;;'.!&,!(9'!*$AA$L&,4!;'%(&$,B!<br />
!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 67!<br />
147
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
HANS W. GOTTINGER<br />
8+%9&:';!$*!$+!&(;!.';%',.-,(?!&,!(9'!*0(0+'!5-+O'(B!<br />
!<br />
D,!(9'!%-;'!$*!&,.&+'%(!,'(L$+O!'**'%(;1!(9'!+'A':-,(!A&,O!F'(L'',!5-+O'(;!&;!&,.&+'%(!F'(L'',!<br />
(9'! %0++',(! +'A':-,(! 5-+O'(! -,.! (9'! *0(0+'! 5-+O'(1! F'%-0;'! &(! 'P&;(;! (9+$049! -! 9$+&S$,(-AA)!<br />
+'A-('.!5-+O'(B!<br />
!<br />
#0%9!+'A-(&$,;9&M;!-+'!(9'+'*$+'!&5M$+(-,(!&,!-!A-('+!;(-4'1!-,.!.&;%0;;'.!*0+(9'+!F'A$LB!K9'!<br />
*$AA$L&,4! (-FA'! ;044';(;! M$;;&F&A&(&';! *$+! &,('+G('5M$+-A! A&,O;! F'(L'',! 5-+O'(;! (9-(! 5-)! M0(!<br />
*&+5;!L9&%9!%0++',(A)!;0MMA)!-!+'A-('.!%0++',(!5-+O'(!-(!-!;&4,&*&%-,(!%$5M'(&(&:'!-.:-,(-4'!&,!<br />
(9'!%-M-F&A&()!($!;0MMA)!(9'!*0(0+'!5-+O'(!(9-,!*&+5;!(9-(!.$!,$(B!<br />
!<br />
*+,+%+-.+/0<br />
849&$,1!TB!-,.!=BU&**&(9!>HII7?!R$5M'(&(&$,!-,.!U+$L(91!R-5F+&.4'1!Q-B!QDK!T+';;!<br />
8++$L1!3BVB!>6WJH?1X/%$,$5&%!N'A*-+'!-,.!(9'!8AA$%-(&$,!$*!=';$0+%';!*$+!D,:',(&$,X1!&,!(9'!=-('!-,.!Y&+'%(&$,!$*!<br />
D,:',(&:'!8%(&:&()Z!/%$,$5&%!-,.!#$%&-A!E-%($+;1!T+&,%'($,Z!T+&,%'($,!2,&:B!T+';;1!JIWGJH7!<br />
>6W[\?1!=&:-A+)!&,!6W[W?1_/5M&+&%-A!#(0.&';!$*!D,,$:-(&$,!-,.!Q-+O'(!#(+0%(0+'_1!&,!`-,.F$$O!$*!<br />
D,.0;(+&-A!a+4-,&S-(&$,1!@$AB!DD1!=B!#%95-A',;''!-,.!=BYBN&AA&41!'.;1!85;('+.-5Z]$+(9!`$AA-,.!<br />
Y$;&1!RB!>6W[[?1!_#$0+%';1!T+$%'.0+';!-,.!Q-%+$'%$,$5&%!/**'%(;!$*!D,,$:-(&$,X1!<br />
V$0+,-A!$*!/%$,$5&%!^&('+-(0+'!HJ>b?1!66HIG66\6!<br />
/-;A')1!YB!-,.!VB3A'&,F'+4!>HI6I?1!]'(L$+O;1!R+$L.;!-,.!Q-+O'(;1!R-5F+&.4'Z!R-5F+&.4'!2,&:B!T+';;!<br />
/:-,;1!YB1!`-4&01!8B!-,.!=B!#%95-A',;''!>HIIJ?1!D,:&;&FA'!/,4&,';1!`$L!#$*(L-+'!TA-(*$+5;!Y+&:'!D,,$:-(&$,!-,.!<br />
K+-,;*$+5!D,.0;(+&';1!R-5F+&.4'1!Q-B!QDK!T+';;!<br />
!<br />
!"#$%&'()!"!#$%&'()!*$+!#%&',%'!-,.!/.0%-(&$,1!2,&('.!3&,4.$5! 67!<br />
149
HANS W. GOTTINGER<br />
2.1 INNOVATION, DYNAMICS OF COMPETITION AND MARKET DYNAMICS<br />
!"##$%&'()*+,*-,*./0123,*4%%"56#$"%)*78%69$:;*"**'!B.C32D:!7*E'9':!F$+?'E9':!4?'!;G#+E4#9':!A#"!+?!#"9!=E*54E!A!M?'G8#,+4E!^,-4?+_4#+*?:!;4,W9#!7*,
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3)4,-567)8,59)<br />
:2;-*&$%9K*!L:$E3>) M#:>! 78A$9:! G:9K>8&8D%:4*! N>>8O3;%8>*! P:;Q8$R! @98>8E%94*! S%9:>4%>D*! NG!<br />
N>?A4;$?!T:O:&8#E:>;!UHVTW!%4!3!9$A9%3&!#K:>8E:>8>!X8;K!=$8E!;K:!#8%>;!8=!O%:Q!8=!;K:!<br />
%>?%O%?A3&!=%$E!3>?!;K:!:98>8E:=%;4!;8!3!=%$E!=$8E!<br />
%>>8O3;%>D!3$:!&%R:&;!=$8E!;K:!489%3&!X:>:=%;4'!N>!;K:!3X4:>9:!8=!3>%4E!<br />
#$:O:>;%>D!%;*!;K:!X:>:=%;4!;8!3>!%>>8O3;%>D!=%$E!3$:!&%R:&;$D!=%$E4'!N>!4A9K!3!49:>3$%8*!=%$E4!3$:!A>&%R:&>8O3;:'!GKA4!3998$?%>D!;8!<br />
98>O:>;%8>3&! ;K%>R%>D*! =%$E4! >::?! ;8! K3O:! 48E:! 48$;! 8=! $:Q3$?! =8$! %>>8O3;%>D'! N>;:&&:9;A3&!<br />
#$8#:$;;4!3>?!98#43;%8>'!"!X%D!#8$;%8>!8=!;K:!<br />
HVT!&%;:$3;A$:!K34!=89A4:?!8>!;K:!8#;%E3&!#3;:>;4[!?A$3;%8>!3>?!X$:3?;K!3>?!;K:!%>9:>;%O:!8=!<br />
=%$E4!;8!%>>8O3;:'!<br />
!<br />
Y8Q:O:$*!3!?%==:$:>;!;$:>?!K34!:E:$D:?!;K:4:!?3D!#$8&%=:$3;%8>!<br />
8=! K%\;:9K! U>:;Q8$RW! %>?A4;$%:4'! N>4;:3?! 8=! ;$D! ;8! D:;! :]9&A4%O:! 8Q>:$4K%#! $%DK;4*! 3>!<br />
%>9$:34%>D!>AEX:$!8=!=%$E4!3$:!E3R%>D!;K:%$!;:9K>8&8D8&8D8!&8>D:$!#$8#$%:;3$!48A$9:_!UI3>:*!@98>8E%4;*!(++/W'!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!<br />
!<br />
N>! ;K:! 48\93&&:?! X$8Q4:$! Q3$*! %>! ;K:! ,``+4*! Q:! K3O:! Q%;>:44:?! %>;:>4:! 98E#:;%;%8>! %>! ;K:!<br />
E3$R:;! =8$! %>;:$>:;! X$8Q4:$4! X:;Q::>! J%9$848=;! 3>?! P:;493#:'! P:;493#:! K3?! 3! E358$! K:3?!<br />
4;3$;!8>!J%9$848=;*!98>;$8&&%>D!`+a!8=!;K:!X$8Q4:$!E3$R:;!XD!%>!;K:!E3$R:;'!b%;K!;K:!:>;$D3D:?!%>!3!$39:!;8!<br />
K3O:!;K:!X:4;!3O3%&3X&:!#$8?A9;'!!<br />
!<br />
L%O:>!;K:!%>;:>4:!98E#:;%;%8>!X:;Q::>!;K:!;Q8!=%$E4*!X?!8=!,``-!J%9$848=;!Q34!#$%9%>D!<br />
;K:!N>;:$>:;!@]#&8$:$!=$::'!N>!98>;$34;*!P:;493#:!Q34!9K3$D%>D!98$#8$3;%8>4!&%9:>4%>D!=::4!=8$!<br />
!<br />
!<br />
!<br />
@,AB/*C09:!3>?!@?A93;%8>*!B>%;:?!C%>D?8E!<br />
!<br />
,F/!<br />
151
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
2(*%G) A0'*/) K/,L('/6) MB) A0') '%") ,C) ?NN:8) O*&/,(,CA) 0$") (A,-'%) $) -$/G') &02%P! ,C) Q'A(&$;'R()<br />
H$/P'A)(0$/'6)Q'A(&$;')'#'%A2$--B)C,--,L'")(2*A)$%")(A$/A'")G*#*%G)$L$B)*A()K/,L('/)C/''6)<br />
)<br />
S0') 'TA'%"'") K$AA-') K'AL''%) O*&/,(,CA) $%") Q'A(&$;') 0$") *A() A,--) ,%) A0') ;/,C*A() ,C) K,A0)<br />
&,H;$%*'(6)9%)?NNU)Q'A(&$;')&$H')2;)L*A0)$)%'L)(A/$A'GB)$%")"'&*"'")A,)/'-'$(')*A()(,2/&')<br />
&,"'8)A0')$&A2$-)-*%'),C);/,G/$HH*%G)-$%G2$G'8)C,/)A0')Q'A(&$;')V,HH2%*&$A,/6)S0*()$--,L'")<br />
2('/()$%")"'#'-,;'/()A,)-,,P)*%(*"')A0')L,/P*%G(),C)A0')K/,L('/8)A,)H,"*CB)A0')(,CAL$/')$%")<br />
'#'%)A,)/'"*(A/*K2A')A0')%'L)#'/(*,%)2%"'/)A0'*/),L%)K/$%")%$H'8);/,#*"'")A0$A)A0')H,"*C*'")<br />
(,2/&')&,"')L$()$-(,)C/''-B)$#$*-$K-'6)S0')L0,-')*"'$)*()A,)A2/%)A0')'%A*/')*%A'/%'A)&,HH2%*AB)<br />
*%A,)$)#$(A)/'('$/&0)"*#*(*,%)C,/)Q'A(&$;')K/,L('/6))<br />
)<br />
S0') A'/H) W,;'%) (,2/&') (,CAL$/'X) 0$() K''%) L*"'-B) 2('") *%) A0') ;,;2-$/) $%") ;/,C'((*,%$-)<br />
-*A'/$A2/') 35$/*$%) $%") +0$;*/,8) =>>YZ) ['/%'/) $%") S*/,-'8) =>>746) 9%(A'$") ,C) P'';*%G) A0'*/)<br />
A'&0%,-,GB) ;/,;/*'A$/B8) A0') C*/H() L*--) "*(A/*K2A') *A) C/''-B6) 9A) *() A0*() ;0'%,H'%,%) A0$A) A0*()<br />
&0$;A'/) $AA'H;A() A,) 'T;-,/'6) \') L*(0) A,) (A2"B) A0') "'&*(*,%() ,C) C*/H() L0'A0'/) A,) P'';) A0'*/)<br />
A'&0%,-,G*'();/,;/*'A$/B),/)%,A6)<br />
)<br />
D#'%)A0,2G0)A0')L0,-')2%,/A0,",T),;'%)$;;/,$&0)H$B)(''H)&,2%A'/*%A2*A*#'8)Q'A(&$;')L$()<br />
%,A)A0'),%-B),%')L0,)'H;-,B'")*A6)!;$&0'8)$);/,G/$H)C,/)('/#*%G)L,/-")L*"')L'K)(*A'(8)$%")<br />
+'%"H$*-8) $) ;/,G/$H) A0$A) /,2A'() $%") "'-*#'/() *%A'/%'A) '-'&A/,%*&) H$*-8) $/') 'T$H;-'() ,C) C/'')<br />
,;'%)(,2/&');/,G/$H()A0$A)",H*%$A')A0')H$/P'A6)];'%)(,2/&')$;;/,$&0'()0$#')K''%)'T;$%"'")<br />
A,) A0') K*,A'&0%,-,GB) $%") 0'$-A0) &$/') *%"2(A/*'(8) D&,%,H*(A) 3=>>7$46) [*%2T8) $%) *%&/'$(*%G-B)<br />
;,;2-$/),;'/$A*%G)(B(A'H)&/'$A'")*%)?NN?)*()$%,A0'/)&-$((*&)'T$H;-'),C)(2&&'((C2-),;'%)(,2/&')<br />
(,CAL$/'6) O$%B) ,C) A0') ;/,G/$HH'/() $%") (,CAL$/') "'(*G%'/() $"#,&$A*%G) ]++) H$B) (0$/') $)<br />
2A,;*$%)#*(*,%),C)(,CAL$/')"'#'-,;H'%A8),/)A0'B)H$B)(*H;-B)L$%A)A,);/,#')A0'H('-#'()A,)K')<br />
K'AA'/) A0$%) (,CAL$/') G*$%A) O*&/,(,CA6) ^,L'#'/8) A0') L0,-') *"'$) ,C) ]++) H$B) %,A) K') (,) $%A*<<br />
&$;*A$-*(A*&) $() *A) (''H(6) 9A) *() 0$/") A,) K'-*'#') A0$A) ;/,C*A
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
&,%(2H'/) J$('6) K0$A) *() J'&$2(') A0') ",L%(A/'$H) (,CAL$/') &,H;$%*'() A02() &$%) A$;) *%A,) A0*()<br />
-$/G'/)%'AL,/M),C)&2(A,H'/(6)K0*()*H;/,#'()A0')N2$-*AB),C)A0')O+6)K0*()*()A0')*%"*/'&A)'CC'&A),C)<br />
%'AL,/M),%)&,%(2H'/(P)2A*-*A*'(6)<br />
)<br />
O;'%)(,2/&')(B(A'H)$;;/,$&0'()'%A$*-)(,H')A/$%(*A*,%$-);/$&A*&$J-')"*($"#$%A$G'()$G$*%(A)A0')<br />
;/'#$*-*%G)(A$%"$/"6)<br />
)<br />
Q'&$2('),C)J2*-"*%G)2;)$)%'AL,/M)A0'/')L*--)J')C'L'/)$;;-*&$A*,%()L0*&0)$/*(')$()/,$"J-,&M()<br />
A,L$/"() C2/A0'/) "*(('H*%$A*,%8) C,/) 'R$H;-'8) $() *%) A0') &$(') ,C) S*%2R8) ,%-B) $) (*%G-') "*G*A)<br />
;'/&'%A$G') L,/M() C,/) &,H;2A'/) G$H'() 3D&,%,H*(A8) =>>7J46) 9%&,H;$A*J*-*AB) *((2'() L*A0) A0')<br />
",H*%$%A),;'/$A*%G)(B(A'H)$/')$-(,),%'),C)A0')J*GG'(A),J(A$&-'()C,/)C2/A0'/)"*(('H*%$A*,%8)*%)<br />
;$/A*&2-$/8) C,/) -$/G') &,/;,/$A*,%() L0,) L,2-") C$&') (*G%*C*&$%A) (L*A&0*%G) &,(A() L*A0) ;,((*J-')<br />
,;'/$A*,%$-)"*(/2;A*,%(6)O%)A0')J'%'C*A)(*"'8)0,L'#'/8),;'%)(,2/&')(B(A'H()H$B)*%"2&').TU)*%)<br />
&,H;-'H'%A$/B);/,"2&A()$%")('&A,/()J'&$2('),C)-,L'/)-*&'%(')C''(8)A02()-,L'/)&,(A(),C).TU8)<br />
A0'')$/')H,/')*%&'%A*#'()A,)'%-$/G')$%")'R;$%");/,"2&A)"'#'-,;H'%A)$%")$;;-*&$A*,%()J'B,%")<br />
&,/') *%%,#$A*,%(6) Q'&$2(') ,C) *A() &,%A*%2,2(-B) ,%G,*%G) &0$/$&A'/) *A) *() H,/') -*M'-B) A,) J')<br />
*%&/'H'%A$-) A0$%) ('H*%$-) L0*&0) ;,('() N2'(A*,%() ,C) A0'*/) ,L%) $() A,) L0'A0'/) ,;'%) (,2/&')<br />
(B(A'H()&$%)G'%2*%'-B)C,(A'/)*%%,#$A*,%6)V,/)"'"*&$A'")$;;-*&$A*,%()$()*%*A*$--B)J2*-A)2;)C/,H)<br />
(&/$A&0)(2&0)$()('$/&0)A0/,2G0)W,,G-')*A)H$B)B*'-")(,H')J*G)$"#$%A$G'()8)%,A)A0')-'$(A)J'&$2(')<br />
,C)&,H;$/$A*#'-B)-,L)*H;-'H'%A$A*,%)$%"),;'/$A*,%$-)&,(A(6)<br />
)<br />
K0*();$;'/)*(),/G$%*X'")$()C,--,L(Y)K0')%'RA)('&A*,%)-,,M()$A)A0')-*A'/$A2/')/'-$A'")A,)A0*()L,/M6)<br />
9%)+'&A*,%)Z6)L');/'('%A)A0')H,"'-6)9%)+'&A*,%)78)L')-,,M)$A)A0')'N2*-*J/*2H)/'(2-A()$%")G*#')<br />
*%A'/;/'A$A*,%(6)+'&A*,%)[)*%"*&$A'()'RA'%(*,%()A,),2/)H,"'-)$%")"/$L()&,%&-2(*,%(),%)C2/A0'/)<br />
2('6)<br />
*<br />
+,)-.)&.'/*0&1-.2&.'*,.3*4#-.*5"6%1-*7-1(."8"'&-2*<br />
K0') *((2'() &,#'/'") *%) A0*() ;$;'/) /'-$A') A,) L,/M) ,%) %'AL,/M) 'RA'/%$-*A*'(8) .TU8) '%A/B)<br />
"'A'//'%&')$%")-*&'%(*%G6)<br />
)<br />
.TU) *() $%) 'RA'%(*#'-B) /'('$/&0'") $/'$) *%) *%"2(A/*$-) ,/G$%*X$A*,%) (A$/A*%G) L*A0) !//,LP()<br />
;*,%''/*%G)$/A*&-')L0'/')0')$(M'"Y)\]0$A)*()A0')G$*%)C/,H)*%%,#$A*,%)A,)$)C*/H)A0$A)*()A0'),%-B)<br />
,%')A,)2%"'/A$M').TU8)G*#'%)A0$A)*A()*%%,#$A*,%)*();/,A'&A'")JB)$);$A'%A),C)2%-*H*A'")"2/$A*,%^_)<br />
3!//,L8)?`I=46))+*%&')A0'%)A0'/')0$()J''%)$)(;$A'),C)/'('$/&0),%).TU)&,#'/*%G)*((2'()(2&0)$()A0')<br />
*%&'%A*#'() A,) *%%,#$A'8) ;$A'%A) /$&'(8) L'-C$/') *H;-*&$A*,%() ,C) .TU8) &0,*&') $%") $",;A*,%) ,C)<br />
A'&0%,-,G*'(6)a,L'#'/8)H,(A),C)A0*()L,/M)0$()C,&2('"),%)&,%#'%A*,%$-)H$/M'A()/$A0'/)A0$%),%)<br />
H$/M'A()L*A0)%'AL,/M)'RA'/%$-*A*'()*%)L0*&0)"B%$H*&),/)b+&02H;'A'/*$%c)&,H;'A*A*,%)'#,-#'()<br />
3D#$%()$%")+&0H$-'%(''8)=>>?46)]')'R;-*&*A-B)-,,M)$A)A0')'CC'&A),C)%'AL,/M)'RA'/%$-*A*'(),%).TU)<br />
&,H;'A*A*,%)$%")*%A/,"2&')A0');,((*J*-*AB),C)C*/H()%,A)L$%A*%G)A0'*/)A'&0%,-,GB)A,)J');/,;/*'A$/B)<br />
L0*&0)*()%,A)/'&,G%*X'")JB)A0')A/$"*A*,%$-).TU)-*A'/$A2/'6)<br />
)<br />
]*A0*%)A0')A,;*&),C).TU8)A0'/')0$()J''%)(,H')L,/M)"'#,A'")A,)-*&'%(*%G6)K0*()-*A'/$A2/')A$M'()$()<br />
A0')(A$/A*%G);,*%A),%'),/)H,/')C*/H()0$#*%G)$);$A'%A6)S*&'%(*%G)*()A0'%)$)H'$%(),C)"*(('H*%$A*%G)<br />
$%) *%%,#$A*,%6) !H,%G) A0') *%&'%A*#'() C,/) -*&'%(*%G) $/') ;/,"2&A) H$/M'A) &,H;'A*A*,%) L0*&0)<br />
&/'$A'() *%&'%A*#'() C,/) H$%$G'/() L0,) L,2-") ,A0'/L*(') 'R;-,*A)A0'*/) H,%,;,-B) ;,(*A*,%(8) &,(A)<br />
($#*%G() A,) A0') -*&'%(''() L0*&0) &,2-") J') $;;/,;/*$A'") JB) A0')-*&'%(*%G) C*/H8) $%") -,L'/*%G)<br />
/*#$-(P)*%&'%A*#'()A,)*%#'%A)$/,2%")A0')*%%,#$A*,%6)F$AX)$%")+0$;*/,)3?`d[4)-,,M)$A)A0')<br />
*%&'%A*#'()A,)'%G$G')*%)-*&'%(*%G),%&')$%)*%%,#$A*,%)*()"'#'-,;'")*%)$)L,/-"),C);'/C'&A);$A'%A(6))<br />
)<br />
154<br />
!"#$%&'())@)+,&*'AB)C,/)+&*'%&')$%")D"2&$A*,%8)E%*A'")F*%G",H)<br />
)<br />
?I:)
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;-! 5#,4#:-;! *B! 4! ,:5:4,?"! ?:>#+@:5!#*!'+55:J+>4#:!+>B*,J4#+*>!J4;!C:!#**!':!G.HH3I!<br />
:($!#"+5!B45"+*>!?4>!+>?,:45:!$,*B+#5!C;!<br />
-+@+>-!?*>5LJ:,5!4>!4'':'!!"#$"%!&$'#*!+>?L,!#":!5:#L$!?*5#!*B!$L,?"45+>-!#":!J4+>!?*J$*>:>#)!!<br />
9":!,:5L!*$:>!5;5#:J!+5!!+#!<br />
+5!J*,:!:!#":!5"4,:! *B!!<br />
#":!J4+>!?*J$*>:>#!+>!#":!#*#4! ,:@:,5:'! ?*J$4,:'! #*! #":! #,4'+#+*>45+>-! 5#:4'! *B! 5+>-! C:+>-! L5:'! 45! 4! J:4>5! *B! '+BBL5+>-! 4>! +>>*@4#+*>! ()%$*' +#! "45!<br />
*??L,,:'A!+>!*L,!J*':5+>-!4?#L4:,! G.H3TA! .H38I! 4>'! D4#E! 4>'! F"4$+,*! G.H3TA.H38I! 4,:! #":! $+*>::,+>-!K*,=5!<br />
:($-! #":! +J$5! *B! >:#K*,=! :(#:,>4! +>'L5#,+4+E4#+*>)! S4,,:'! 5+':! 4>'! 5"*K! #"4#! #":! :(+5#:>?:! *B! >:#K*,=!<br />
:(#:,>4!$,*C'!#"L5!#*!4!JL! *B! 4! >:K! 5#4>'4,'! J+-"#! C:! #**! 5'!F4:,M5!K*,=5!C;!+>?-! C#)!9":;!L5:!#":+,!J*':B*,?:!#":!+J$*,#4>?:!*B!?*>5LJ:,!<br />
:($:?#4#+*>5!+>!J4,=:#5!K+#"!>:#K*,=!:(#:,>4-!?*J$4#+C+'!K":#":,!5+':!$4;J:>#5!4,:!B:45+C'!F"4$+,*!G.H38I!5"*K!<br />
#"4#! +>! #":! $,:5:>?:! *B! >:#K*,=! :(#:,>45!L>':,!K"+?"!B+,J5!J4;!L5:!?*J$4#+C+-!<br />
?*J$:#+#+*>)!<br />
!<br />
W"+:#K*,=! :(#:,>4:K! $,*'L?#! K"+?"! :JC*'+:5! #:?">*!:UL+:K! $,*'L?#! +5! +>#,*'L?:'! #**! 5**>! ?*J$4,:'! #*! #":!5*?+45!<br />
C:?4L5:! #":! 5**>:,! #":! $,*'L?#! +5! +>#,*'L?:'A! #":! 5**>:,! #":! B+,J! ?4>! 5#4,#! CL+-!+#5!C:>:B+#5)!["*+!G.HH0I!!4!<br />
J4,=:#!K+#"!>:#K*,=:(#:,>4
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
K')(A2"*'()A0')*%&'%A*#'),C)A0')H,%,;,-*(A)A,)*%A/,"2&')$%)*%&,H;$A*L-')*H;/,#'");/,"2&A)*%)<br />
A0') ;/'('%&') ,C) %'AM,/N) 'OA'/%$-*A*'(6) F/*(A*$%('%) 3?JJI4) (A2"*'() A0') &,%('P2'%&'() ,C) %'AM,/N)<br />
'OA'/%$-*A*'() ,%) A0') /*(N*%'(() ,C) .QR) ;/,S'&A() &0,('%) LB) $%) '%A/$%A) $%") $%) *%&2HL'%A6) K')<br />
(0,M() A0$A) A0') *%&2HL'%A) &0,,('() $) A,,) /*(NB) ;/,S'&A) A0$A) A,,) ,CA'%) -'A() $) %'M) C*/H)M*A0) $%)<br />
*%&,H;$A*L-')A'&0%,-,GB)'%A'/)$()&,H;$/'")A,)A0')(,&*$-),;A*H2H6)9%)$""*A*,%8)A0')'%A/$%A)<br />
0$() $%) *%&'%A*#') A,) &0,,(') H,/') &'/A$*%) ;/,S'&A() A0$%) $/') (,&*$--B) ,;A*H$-) $%") A0'(')<br />
(A/'%GA0'%) A0') ;,((*L*-*AB) ,C) $",;A*,%) ,C) $%) *%&,H;$A*L-') A'&0%,-,GB6) .'G*L'$2) $%") .,&N'AA)<br />
3?JJT4)$%$-BU')&,H;$A*L*-*AB)&0,*&'(),C)AM,)C*/H()M0*&0)H2(A)$-(,)"'&*"')M0'%)A,)*%A/,"2&')<br />
A0'*/) G,,"() *%) $) H$/N'A) &0$/$&A'/*U'") LB) %'AM,/N) 'OA'/%$-*A*'(6) V0'B) (0,M) A0$A) A0') C*/H(W)<br />
*%&'%A*#'() A,) $&0*'#') &,H;$A*L*-*AB) "';'%") &/2&*$--B) ,%) A0') A*H') $A) M0*&0) A0') "'G/'') ,C)<br />
&,H;$A*L*-*AB) H2(A) L') &0,('%6) V0') &2//'%A) ;$;'/) *() $-(,) /'-$A'") A,) A0') -*A'/$A2/') ,%)<br />
*%#'(AH'%A)$()$)H'$%(),C)'%A/B)"'A'//'%&'6)X*/H()&,H;'A')%,A),%-B)M*A0)'O*(A*%G)C*/H(8)$()*()<br />
A/$%(;$/'%A8)L2A)$-(,)M*A0);,A'%A*$-)'%A/$%A(6)9%)$%)'OA'%(*,%),C)A0')+;'%&')H,"'-)R*O*A)3?JT>4)<br />
(0,M() A0$A) $%) *%&2HL'%A) C*/H) H$B) H$N') *//'#,&$L-') &,HH*AH'%A) ,C) *%#'(AH'%A) *%) ,/"'/) A,)<br />
$-A'/)A0')*%*A*$-)&,%"*A*,%(),C)A0');,(A)2%-*N')A0')'%A/$%A)3OD()(>4*([')&,%(*"'/)$)AM,)(A$G')G$H')*%)$)H$/N'A)<br />
M*A0)%'AM,/N)'OA'/%$-*A*'(6)V0');/,"2&A(),C)A0')C*/H()$/')*%&,H;$A*L-')3*6'68)A0'B)0$#')(';$/$A')<br />
%'AM,/N(46)<br />
)<br />
9%) A0') C*/(A) (A$G'8) C*/H() &0$/G') $) -*&'%(') C'') 3+4) A,) ,A0'/) ",M%(A/'$H) C*/H() C,/) 2(') ,C) A0'*/)<br />
;/,"2&A)3C,/)'O$H;-'8)\*&/,(,CA)-*&'%('()A0'*/)[*%",M(),;'/$A*%G)(B(A'H)A,),A0'/)&,H;$%*'()<br />
A,)"'#'-,;)$;;-*&$A*,%()(,CAM$/'46)V0')%2HL'/),C)",M%(A/'$H)C*/H()M*--*%G)A,)M,/N),%)$)C*/HW()<br />
;/,"2&A)3H*4)"';'%"(),%)A0')-*&'%(')C'')&0$/G'")$%")$-(,)A0')*%*A*$-)%'AM,/N),C)A0')C*/H)3A0')<br />
&,%(2H'/)L$('),%)C,/)C*/H)*46)<br />
(<br />
-(%)(.(/(,%(0(+%8)(.(1(2( 3?4)<br />
)<br />
.(%3($)H'$(2/'),C)A0');,A'%A*$-)H$/N'A)(*U'),C)",M%(A/'$H)C*/H()*%"';'%"'%A),C)A0')-*&'%(')C'')<br />
$%")*%*A*$-)%'AM,/N6)<br />
)<br />
V0')(H$--'/)A0')-*&'%(')C'')&0$/G'"8)A0')H,/')C/''-B)A0')C*/H)"*(A/*L2A'()*A()A'&0%,-,GB)$%")*%),2/)<br />
A'/H*%,-,GB8) A0') H,/') ],;'%^) A0') C*/H) *(6) [') $((2H') %,) &,H;'A*A*,%) *%) G'AA*%G)<br />
",M%(A/'$H) C*/H() A,) L2B) A0') C*/H(W) -*&'%('(6) -%) "'&/'$('() $() +*) *%&/'$('() L'&$2(') $() A0')<br />
-*&'%('() L'&,H') H,/') 'O;'%(*#') C'M'/) ",M%(A/'$H) C*/H() L2B) A0'H6) V0*() *() S2(A) A0')<br />
(A$%"$/") $/G2H'%A) C,/) $) ",M%M$/") (-,;*%G) "'H$%") *%) ;/*&'6) !() C,/) A0') /'-$A*,%(0*;)<br />
L'AM''%( -%) $%") ,%(4) (*%&') A0') ",M%(A/'$H) C*/H() $/') "'#'-,;*%G) ;/,"2&A() A0$A) &,2-") L')<br />
2('")*%)&,%S2%&A*,%)M*A0)A0')C*/HW();/,"2&A)3-*N')&,H;-'H'%A(48)A0')-$/G'/)A0')*%(A$--'")<br />
L$('),C)&2(A,H'/()A0$A)A0')C*/H)0$(8)A0')-$/G'/)*()A0');,A'%A*$-)"'H$%")C,/)A0')",M%(A/'$H)<br />
C*/H(W);/,"2&A6))V02(8)A0')H,/')A0')",M%(A/'$H)C*/H()$/')M*--*%G)A,);$B)C,/)A0')-*&'%('6)<br />
)<br />
_'A)2()"'C*%'`))9C)))+% ( !a)+5 ( !A0'%)C*/H)*)&0,,('()A,)L')H,/'),;'%)A0$%)C*/H)S6)3!-A'/%$A*#'-B(C*/H)S)<br />
&0,,('()A,)L')-'((),;'%)A0$%)C*/H)*4*(<br />
)<br />
)<br />
156<br />
!"#$%&'())@)+,&*'AB)C,/)+&*'%&')$%")D"2&$A*,%8)E%*A'")F*%G",H)<br />
)<br />
?IJ)
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;##*! 5:@@! *?#$?#)! N:9;:! D:! ,:5#,+;#! #":!<br />
:9#,49#!#*!;"**5+9-!4!@+;:95:!*?,9*#!-4=:)!B"?5H!#":!X'*D95#,:4=!
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
!()*%),A0'/)AI,
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;!'9:+5+*;!*?!?+,@!A!4;'!#"9!9BC+D+E,+C@!D+:9;59!<br />
?995!?*,!#"9!#!+;!#"9!H*C,;*#!5CE-4@9G!#*!#"9!'+??9,9;#!D+:9;59!<br />
?995!#"4#!#"9!+;:C@E9;#!:4;!:"4,-9)!I*,!$,**?5!$D9459!,9?9,!#*!#"9!4$$9;'+()!89!:4;!5"*!*C#!*?!#"9!@4,J9#!4;'!?*,!"!##>##" ! = #+#!+5!#*!9;#9,)!K"9,9!9(+5#5!"! ? ##>##" ! = !<br />
5C:"!#"4#!+;!#"9!+;#9,L4D!M"!# = #6#"! ? N##"9!E95#! ,95$*;59! *?! #"9! 9;#,4;#! +5! #*! :"4,-9! #"9! D4,-95#!<br />
4::9$#4ED9! D+:9;59! ?99! #"4#! @4J95!*$#+@4D! (A! 9(4:#D>! /G! 4;'! ?*,! ƒO! P! ƒ! ? ! +#! +5! *$#+@4D! ?*,! #"9!<br />
9;#,4;#!#*!9;#9,!4;'!:"4,-9!+#5!C;:*;5#,4+;9'!*$#+@4D!D+:9;59!?99!,95CD#+;-!+;!4!$*5+#+L9#<br />
@A#<br />
%<br />
&'()*(+%,%&'(%-(.(-/%)0%12 ) 3%4%567%8%7(9(*+:6(%9'(%(69*;%7(+%5/%0)--)@/$%<br />
.) A;#,>!+5!ED*:J9'!+?!Q.!R!7!S!(O! * !"!T!R!7J)!<br />
7) I*,!T!R!7J!"!Q.!R7!S!(O! * !"!TRJ&7G!9;#,>!+5!,95#,+:#9'!)!#"9!9;#,4;#!<br />
9;#9,5!#"9!@4,J9#)!Z*!?995!4;'!@4J9!+#5!@*;*$*D>!$,*?+#5!! "+-"! #"9;! #"9! +;:C@E9;#!:4;! ,95#,+:#!<br />
9;#,>!4;'!@4J9!@*;*$*D>!$,*?+#5)!O?!+#5!4'L4;#4-9!+5!"*!$,*?+#5)!K"C5!<br />
+;!#"+5!:459!#"9!+;:C@E9;#!+5!4ED9!#*!4:#!45!4!@*;*$*D+5#!+;!#"9!*C#$C#!@4,J9#!EC#!;*#!+;!D+:9;59!<br />
?995)!<br />
!<br />
O?!#"9!+;+#+4D!;9#
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
!()&,/,--$/*'()A,)J/,;,(*A*,%)=)K')G'A)A0')/'(2-A)A0$A)A0'/')*()%,)'%A/B)"'A'//'%&')$%")K')$-(,)G'A)$)<br />
&,H;$/*(,%) L'AK''%) A0') H,%,;,-B) -*&'%(') C''() ,C) A0') *%&2HL'%A) $%") *A() -*&'%(')C''() K0'%) $)<br />
;,A'%A*$-)'%A/$%A)'M*(A(6)<br />
)<br />
!"#"$$%#&'(')*+#+',-'."'+./#&'0+/+##+.1+2'<br />
!A)'!"# $# 'A0')'%A/$%A);/,"2&'()%,A0*%G)$%")A0')*%&2HL'%A)N2(A)H$O'()A0');/,C*A()*A)K,2-")$()$)<br />
H,%,;,-*(A) *%) A0') ('&,%") (A$G') ,2A;2A) G$H'6) +*%&') !"# $# ' *() A0') -$/G'(A) -*&'%(') C'') C,/) A0')<br />
*%&2HL'%A)A0$A)"'A'/()'%A/B8)*C)#!"# %# !#!"# $# 'A0'%)A0'/')*()L-,&O'")'%A/B6)9C)!" % #&#!"# $# '0,K'#'/8)A0'%)<br />
'%A/B)*()%,A)L-,&O'"6)P2A)*A)$-(,)H'$%()A0$A)A0')*%&2HL'%AQ();/,C*A()H2(A)L')*%&/'$(*%G)$A)!"# $# '<br />
3(*%&')A0')H,%,;,-B);/,C*A()$/')&,%&$#')*%)!" # '$%")C/,H)R'HH$)!6=)K')O%,K)A0$A)*%)A0')*%A'/#$-)S!"#<br />
$ ''!"# (# ]#'#A0')*%&2HL'%A)H$O'()H,%,;,-B);/,C*A()46)T'%&')A0')*%&2HL'%A)K*--)%,A)C*%")*A);/,C*A$L-')<br />
A,)&0$/G')!"# $# ',/)-'((6)U02()'%A/B)*()%,A)"'A'//'"6))<br />
)<br />
!()$)&,/,--$/B)A,)U0',/'H)=)K')G'A)$)&,H;$/*(,%)L'AK''%)A0')H,%,;,-B)-*&'%(')C''(),C)A0')<br />
*%&2HL'%A)$%")*A()-*&'%(')C''()K0'%)$);,A'%A*$-)'%A/$%A)'M*(A(6)<br />
)<br />
!"#"$$%#&'3'4/'/*+'+56,$,7#,68'06"9"$&'$,1+.-+':++':;
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;9:!#":!;:#A*,C!#:,=H!A:!5::!#"4#!#":!+;9>=D:;#!9"**5:5!#*!D:!:L:;!=*,:!*$:;!A"+8:!<br />
#":!:;#,4;#!9"**5:5!#*!D:!8:55!*$:;)!!<br />
!<br />
O,*=!#":!P$$:;'+(!QP!)./@!A:!5::!#"4#!5#4,#+;-!9#+*;!+;!!%"# +;!<br />
*,':,!#*!D,+;-!#":!=4,-+;48!:8#!+#!+5!<br />
,:!#":!9:;#,48!+':4!*#>,:!<br />
,:5:4,9"!:':'! #":! #,4'+#+*;48! M#:9";*8*-?! 5#:48+;-M! :=:,5H!;4=:8?!#"4#!9*;5>=:,5!A+88!$:,9:+L:!#":+,!<br />
$,*'>9#! #*! D:! ;*#! 45! 4##,49#+L:! 45! #":! +;9>=D:;#J5H! :L:;! +48+#?)! E#!<br />
A*>8'!5#+88!D:!+;#:,:5#+;-!#*!5::%#":!,:5>8#!*-"!#*!4!8:55:,!:(#:;#)!I":!<br />
:R>+8+D,+>=!;*A!;::'5!#*!5#,+C:!4!D484;9:!D:#A::;!#":!#A*!*$$*5+;-!:8'! :($:9#! #":! ,:5>8#! #"4#! #":! +;9>=D:;#! -*:5! =*,:!<br />
*$:;!#*!D:!,:+;9:!9*=$:#+#+*;!+;!#":!
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
$%) *"'%A*&$-) $%") *%"';'%"'%A) "'H$%") C,/) *A() -*&'%('(6) J0*() &,2-") K') L2(A*C*'") *C) A0'/') $/')<br />
%2H'/,2() ",M%(A/'$H) C*/H() -,,N*%G) C,/) ;/,(;'&A*#') A'&0%,-,G*'() A,) "'#'-,;6) !) H,/')<br />
/'$-*(A*&)('AA*%G)H$B)K'),%')M0'/')K,A0)A0')*%&2HK'%A)$%")A0')'%A/$%A)$/')&,H;'A*%G)*%)<br />
-*&'%(') C''() C,/) ",M%(A/'$H) C*/H() A,) K2B) A0'*/) -*&'%('(6) D$&0) ",M%(A/'$H) C*/H) K2B() $A)<br />
H,(A),%')2%*A),C)A0')-*&'%('6))<br />
)<br />
O,M'#'/8)A0*()(0,2-")%,A)&0$%G'),2/)/'(2-A()"/$(A*&$--B)$%")M')(0,2-")(A*--)('')P,;'%P)$()$%)<br />
'Q2*-*K/*2H)(A/$A'GB6))<br />
)<br />
9%)C$&A8)M*A0)&,H;'A*A*,%8)K,A0)A0')*%&2HK'%A)$%")A0')'%A/$%A)$/')-*N'-B)A,)G,)H,/'),;'%)<br />
K'&$2(')%,M)A0'B)%''")A,)&,H;'A')A,)('--)A0'*/)-*&'%('(6)<br />
)<br />
R')&,2-")*%A/,"2&')&,(A()*%A,)A0')A'&0%,-,GB),C)A0')C*/H(6)9%),2/)&2//'%A)H,"'-8)M')0$#')<br />
%,/H$-*S'")K,A0)C*/H(T)&,(A()A,)S'/,6)U*CC'/'%&'()*%)A0')A'&0%,-,G*'(),C)A0')C*/H()H$B)(M$B)<br />
,2/)/'(2-A()'*A0'/)M$B)"';'%"*%G),%)M0,)0$()A0')K'AA'/)3-'(()&,(A-B4)A'&0%,-,GB6)9A)M,2-")<br />
$-(,)K')*%A'/'(A*%G)A,)('')0,M)A0')&,2;-*%G),C)A0')PA'&0%,-,GB)(A'$-*%GP)'CC'&A)$%");,(*A*#')<br />
&,(A()H$B)M,/N6)V,/)'W$H;-'8)*C)A0')*%&2HK'%A)C/''-B)"*(A/*K2A'()*A()A'&0%,-,GB)A0'%)A0')<br />
&,(A() C,/) A0') '%A/$%A) H$B) C$--) K'&$2(') *A) &$%) %,M) P(A'$-T) A0') A'&0%,-,GB) ,C) A0')<br />
*%&2HK'%A6)<br />
)<br />
9%),2/)H,"'-8)M')",%TA)0$#')'%A/B)"'A'//'%&')K2A)M')0$#')/'(A/*&A'")'%A/B)*%)M0*&0)A0')<br />
'%A/$%A)('--()-*&'%('()*%)A0')C*/(A)(A$G'8)K2A);/,"2&'()%,A0*%G)*%)A0')('&,%")(A$G'6)J0*()<br />
H$B) (''H) $) K*A) 2%/'$-*(A*&6) R*A0) ;,(*A*#') C*W'") &,(A(8) M') (0,2-") K') $K-') A,) 0$#') A0')<br />
(A$%"$/")'%A/B)"'A'//'%&')/'(2-A6)R')M,2-")(A*--)'W;'&A)/'(A/*&A'")'%A/B)K2A),%')M0'/')A0')<br />
'%A/$%A)*()T/'(A/*&A'"P)A,)(,H');,(*A*#'),2A;2A)-'#'-6)<br />
)<br />
!%,A0'/);,((*K-')'WA'%(*,%)*()H$N*%G)*%%,#$A*,%)(A,&0$(A*&8)*6'6)/';-$&*%G)A0')-*%N)!"#$%#&#$#'(%)#<br />
M*A0) $%#) X)$'(%*+%)# *#M0'/') + % #*() (A,&0$(A*&6) R') $/G2') A0$A) A0*() $G$*%) (0,2-") %,A) &0$%G') ,2/)<br />
/'(2-A() Q2$-*A$A*#'-B) (,) -,%G) $() A0') /'(2-A*%G) "*(A/*K2A*,%() C,/) Q2$-*A*'() 'W0*K*A) ('&,%") ,/"'/)<br />
(A,&0$(A*&) ",H*%$%&')
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;/:N0/ONN2!<br />
EK*F*I+'B5)!A)!=.>>8?!@S"B!EK*F*I+K5!*U!AB#C*,D5M:!-)0#()*01&)*+$%&'()*+$&,$-).'/0(1*+$5(6*)17*01&)$.0:!133O3//!<br />
S"B!EK*F*I+5#!=7//048$9:)$5;#)3N?!@EK*F*I+K!V55YB5!+F!Z#4F'4,'+`4#+*FM:!+F!;4IB5!\+GGB,:!B'):!!#+#3&44')13*01&)/$*).$<br />
2G'10HI$J&+13H$K#/#*(3"$-//'#/:!A*,#"!a*GG4F':!ABC!b*,D:!.8NO.13$<br />
^4,,BGG:!;!4F'!Z4G*FB,:!_)!=.>38?!@VF5#4GGB'![45B!4F'!9*I$4#+J+G+#H%!VFF*W4#+*F:!3N?!@LF!#"B!c+KBF5+F-!*U!VFF*W4#+*F5M)!K*).$%&'()*+$&,$23&)&413/$.8:!N/0ON7/!<br />
R4#`:!\)!4F'!Z"4$+,*:!9)=.>38?!@>7?!!d8?!@TeP!+F!\4,DB#5!C+#"!AB#C*,D!E(#B,F4G+#+B5M:!-)0#()*01&)*+$%&'()*+$&,$-).'/0(1*+$5(6*)17*01&)$<br />
.0:!N2.ON01!<br />
cB,FB,:;)!4F'!;)!S+,*GB!=7//0?:!S"B!EK*F*I+K5!*U!SBK"F*G*-H!Z"4,+F-%!L$BF!Z*Y,KB!4F'![BH*F'M:!a4,W4,'![Y5+FB55!ZK"**G!]*,D+F-!<br />
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
*++,-./01<br />
+%""21"213(4"%45161<br />
I0'%)A0')H,%,;,-*(A)0$()$);/,"2&A),C)J2$-*AB)!"#$%")&0$/G'()$);/*&'),C);9)A0'),%-B)&,%(2H'/()<br />
K0,) L2B) C/,H) *A) $/') A0') ,%'() C,/) K0,H) $# %# !"# %# !# &" # ' # (# )"# # !) >6) M0'/'C,/') "'H$%") C,/) A0')<br />
H,%,;,-*(AN();/,"2&A)'J2$-()*#%#!"#%#!#&" #' #(#)"#+##9AN()/'#'%2'3.4)'J2$-())"#,*#%#!"#%#!#&" #' #(#)"-+6)M0')<br />
,;A*H$-);/*&')C,/)('--*%G)*A(),2A;2A)*()A02()G*#'%)LBO)<br />
)<br />
" .P" ; 9 ) Q )!)R)J9)R)#)S9 ), ))<br />
)<br />
,/))))))))))); 9 T QS9 T Q3)!)R)J9)R)#)S9 ), 4P)=))))))))))))))))))))))))))))3!?4)<br />
)<br />
M0')H,%,;,-*(AN();/,C*A)*()A0'/'C,/')<br />
$9 T) Q)[)3)!)R)J9)R)#)S9 ), 4P)=] )= )R)C9)3U9)R)S9 ), )
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;B,>45>5!#">!<br />
$,*=+#!=,*:!#">!=+,5#!#>,:!D5+;B>!=,*:!#">!E*9,;*#!-4:>A!"#!+5!'>B,>45+;-!+;!ƒ+$%!8#!#">!*$#+:4?!ƒ+!<br />
#"+5! :95#! ! >(4B#?C! *==5>#! !>F9+?+,>45!#">!=+,:!+H5!,>@>;9>!<br />
=,*:!?+B>;5+;-!+5!9;4==>B#>'!;!4!5:4??!+;B,>45>!+;!ƒ+!=,*:!+#5!<br />
$,>@+*95!*$#+:4?!?>@>?!+;B,>45>5!#">!,>@>;9>!=,*:!#">!?+B>;5>!=>>5!!'>B,>45>!+;!,>@>;9>!=,*:!'+,>B#!54?>5!+5!?4,->,!5+;B>!"#!+5!?4,->,)!I"95!G+#"!<br />
#">!?4,->,!ƒ6!$,*=+#5!4,>!'>B,>45+;-!+;!+H5!?+B>;5>!=>>5!4#!#">!$,>@+*95?C!*$#+:4?!ƒ+%%!K+@>;!#">!<br />
B*;B4@+#C!*=!#">!$,*=+#!=9;B#+*;!+;!4!=+,:H5!*G;!?+B>;5>!=>>!#"+5!,>F9+,>5!4!?*G>,+;-!*=!ƒ+!+;!*,'>,!#*!<br />
,>4B"!>F9+?+;B>!#">!,>4B#+*;!=9;B#+*;!+5!'*G;G4,'!5?*$+;-)!<br />
!<br />
L>::4! 8.)! I">,>! >(+5#5! =M N ! =*,! #">! +;B9:;#! 59B"! #"4#! =*,! 4??! =M! O! =M N !!#">! *$#+:4?!<br />
,>5$*;5>!*=!#">!>;#,4;#!+5!#*!5#4C!*9#!*=!#">!:4,P>#!4;'!=*,!=M!Q!=M N! +#!+5!#*!>;#>,)!=M N !+5!#"95!<br />
#">!?4,->5#!?+B>;5>!=>>!=*,!#">!+;B9:;#!#"4#!'>#>,5!>;#,C)!<br />
!<br />
R,**=&%L>#!!'( ) !!#">!?+B>;5>!=>>!=*,!G"+B"!'*!+!/!+5!#">!5:4??>5#!?+B>;5>!=>>!#"4#!:4P>5!"*!,'*&!'( ) $!S!/)!<br />
T">;!'*!+!-%!"*!S!/A!4;'!'(!+!'( ) !#">;!!*!S!/!4;'!"!*!/!"ƒ*!+!/!U!P!Q!/)!I"95!,>'9B+;-!'*!G+??!<br />
*;?C!:4P>!!*!!;>-4#+@>!G"+?>!4!,!'*!+5!;*#!4;!4BB>$#4!?+B>;5>!=>>)!V*!>;#,C!+5!'>#>,,>'!4#!<br />
'( ) %!<br />
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!<br />
I*!B4?B9?4#>!'( ) !G>!5>#!!"*!>F94?!#*!W>,*!4;'!'*!+!-!G+#"!+#)!<br />
T>!#">;!->#!<br />
!<br />
8U7DP!!!ƒXY!!!DP!U!(M! 0 !!!!=M N Y!!!"(M! * !S!/!!!!!!!!!!!!!!D8ZY!<br />
!<br />
I">,>=*,>A!=M N !S!D.!U!"Y!(M! !!!8!!!P!!!!!!!!!!!!!!!!!!!!!!!!!!!D81Y!<br />
0<br />
!<br />
X;#,C!+5!'>#>,,>'!4#!4;C!ƒ(!.!'( ) !!#**A!5+;B>!#">;!#">!5:4??>5#!'*!!#"4#!:4P>5!"*!,'*&!'($!S!/)!:95#!!<br />
;>-4#+@>!A5>>!D8ZY)![*,!'(!/!'( ) !#">!5:4??>5#!!' * !#"4#!:4P>5!"*!,'*&!'($!S!/)!+5!$*5+#+@>A!#"95!#">,>!'*!<br />
>(+5#! 4BB>$#4! ?+B>;5>! =>>5! #"4#! -+@>! #">! >;#,4;#! $*5+#+@>! $,*=+#5)! J>;B>! #">! >;#,4;#! G+??!<br />
>;#>,!4;'!5*!#">,>!+5!;*!>;#,C!'>#>,,>;B>)!I"95!'( ) !#0!+;'>>'!#">!?4,->5#!?+B>;5>!=>>!#"4#!#">!<br />
+;B9:;#!B4;!B"4,->!4;'!5#+??!'>#>,!>;#,C)!<br />
!<br />
L>::4!87&%%I">,>!>(+5#5!=M \ !Q!=M N !, 59B"!#"4#!+;!#">!+;#>,@4?![!=M N A!=M \ ]!!+#!+5!*$#+:4?!=*,!=+,:!X!#*!<br />
B"4,->! #">! ?4,->5#! 4BB>$#4! ?+B>;5>! =>>! =X!!#"4#! :4P>5! (X! S! /! 4;'! =*,! =M! Q! =M \ ! #">! *$#+:4?!<br />
,>5$*;5>!=*,!=+,:!X!+;'9B>5!+#!#*!$,*'9B>!4!$*5+#+@>!4:*9;#)!<br />
!<br />
R,**=&%[,*:!>F94#+*;!D8]Y!G>!P;*G!#"4#!#">!5#!,>5$*;5>!=*,!#">!>;#,4;#!#*!=M \ +5!-+@>;!;!=M!S!=M \ A!!= \ X!!:95#!!45!-+@>;!!(H5)!#">;!G">;!=+,:!M!B"*5>!'( 1 !4;'!=+,:!X!B"*5>!+#5!?+B>;5>!<br />
=>>!*$#+:4??CA!! +#5! ?+B>;5>! =>>! *$#+:4??CA! #">;! +;! #">!<br />
59F9>;#!*9#$9#!-4:>!=+,:!X!G*9?'!>;'!9$!5>??+;-!4!$*5+#+@>!4:*9;#)!M=!!"*G>@>,!=+,:!M!<br />
B"*5>!'(!;!+;!#">!<br />
59F9>;#!*9#$9#!-4:>!=+,:!X!G*9?'!!,>F9+,>'!#*!5>??!4!;>-4#+@>!4:*9;#)!V+;B>!#"4#!+5!<br />
!<br />
!"#$%"##$%&&'()'*+)*,-&./).0123&455,6)01)72.7)% .13!<br />
!<br />
165
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
166<br />
5,-678)9((2'):)!;/*-?:)<br />
)<br />
%,A);,((*J-')D)K2(A)('A()*A()-*&'%(')C'')$A)A0')-'#'-)A0$A)K2(A)H$L'()*A(),2A;2A)*%)A0')%'MA)(A$G')<br />
N'/,6)O'%&') P0'%) A0'/') *() %,) '%A/B) J-,&L8) *%) A0') *%A'/#$-)[!"# $ %!"# & ]!!C*/H) D) &0$/G'() A0') -$/G'(A)<br />
$&&';A$J-')-*&'%(')C'')"'!A0$A)H$L'()('!)!*+!Q*/H)9)H$L'();/,C*A()$&&,/"*%G)A,)A0')H,%,;,-*(AR()<br />
;/,C*A)C2%&A*,%6)S0*()*()P0'/')A0')*%'T2$-*AB)M*)!)>)C,/)A0')'%A/$%A)*()J*%"*%G6)<br />
)<br />
Q,/)"#!,!"# & !A0')'%A/$%A)K2(A);-$B()*A()2%&,%(A/$*%'")J'(A)/'(;,%(')*%)A0')+A$&L'-J'/G)G$H'*+!"# & )*()<br />
A0') (H$--'(A) -*&'%(') C'') C,/) A0') *%&2HJ'%A) $J,#') P0*&0) A0') '%A/$%A) ;-$B() *A() 2%&,%(A/$*%'")<br />
,;A*H$-)C''()$%"),2A;2A)*%)'T2*-*J/*2H*+<br />
+<br />
,%""-+"-+.(/"%/0+1+<br />
U');/,#')A0')A0',/'H)*%)$)C'P)(A';(6)<br />
)<br />
-.//0!12+!'3456!78!9:;!7"!?@!A!B!C!(# ; "!1!A!B=+!<br />
)<br />
U')C*%")A0')(H$--'(A)"'!A0$A)G*#'()2()('!?"'%!"# D C!V)>!<br />
Q/,H)3!W4)P')G'AX)<br />
)<br />
!)Y)=3L")ƒD4)")3L)Y)M9 > )")C9 Z 4)")#M9 , )V)>)<br />
)<br />
+2J(A*A2A*%G)A0')'M;/'((*,%)C,/)"# D !f C/,H)'T2$A*,%)3!=4!*%A,)A0')$J,#')G*#'()2(X)<br />
)))))))))))))))))))))))))))))))))))))))))))))))))))<br />
ƒD $) V)!)Y=L)")3?)Y)=#4M9 , [\)<br />
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))<br />
ƒD $ !78!A0')-$/G'(A)$&&';A$J-')-*&'%(')C'')C,/)A0')'%A/$%A)P0'%)A0')*%&2HJ'%A)&0$/G'()"# D )<br />
U0'%)ƒD $ %)>)8)#'!)$A)ƒD $ )*()%,%)>6!<br />
)<br />
]*#'%)A0')&,%&$#*AB),C)#'!)*%)ƒD)))H'$%()A0$A)*C)!ƒD %)>)8)<br />
$<br />
A0'/') *() %,) $&&';A$J-') -*&'%(') C'') C,/) A0') '%A/$%A) A0$A) G*#'() *A) $) ;,(*A*#') ;/,C*A6)O'%&') '%A/B) *()<br />
J-,&L'")*C)!ƒD %)>)8)*6'6)*C)<br />
$<br />
)<br />
!Y=L^)3?Y=#4M9 ,)<br />
)<br />
ƒD V) >) *() A0') -$/G'(A) $ ƒD ) C,/) P0*&0) '%A/B) *() J-,&L'") (*%&') *C) $ ƒD ) >) >) A0'%) A0'/') ",'() 'M*(A) $%)<br />
$<br />
$&&';A$J-')-*&'%(')C'')C,/)A0')'%A/$%A)A0$A)G*#'()*A);,(*A*#');/,C*A6)<br />
!<br />
-.//0!1EFG;5!1AB="!?@!A!B!C!(# ; !"!1!A!=HB%!.3456!78!5.8457!I74J!4J.!73
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
HANS W. GOTTINGER<br />
!"##$%&'()*+,*-./012,*3'#4"(5*67"%"8$'9*:"(*;.=H#?!(9! * !<br />
!<br />
IE##+F-!#*-D#"D,!#"+5!B*F'+#+*F!A+#"!#"4#!8*,!F*!JK*B.!=#?!(9 * "!@T!<br />
&!" . /"!!"#/#567#[8#9#:+# #7&5#9:$-#1"# ; #]#<br />
!<br />
!<br />
!" . !+5!'DB,D45+F-!+F!!!"##*4#!"#/#ƒ" ; !+8!<br />
.!=7#?!(9 * !<br />
!<br />
["E5!8*,!7@&P!=!3# ?;@# 7@&P=
HANS W. GOTTINGER<br />
2.2 NETWORK ECONOMICS FOR OPEN SOURCE TECHNOLOGIES<br />
!"#$%&'()*%)+,&*$-)+&*'%&'().'('$/&0)1,2/%$-)3!++.14)<br />
)<br />
5,-678)9((2'):)!;/*-?:)<br />
ƒ! . "#+*%&')!! $ )*()&,%&$#'8)A0'),;A*H$-)-*&'%(')C'')C,/)A0')*%&2HJ'%A)H2(A)J')G*#'%)JB)A0'),%')<br />
A0$A)H$K*H*L'()!! $ "#)M$--)A0*()-*&'%(')ƒ! + "#+'AA*%G)!!! $ /")%!# N)>)C/,H)3!?>4)O')G'A)<br />
)))))))))))))<br />
ƒ9 + )N)[PQ)")I!)")3:)R)S=#4)K9 , ])T?I))))))))))))))))))))3!??4)<br />
)<br />
U*#'%),2/)"'C*%*A*,%),C#ƒ! . &#%!# #ƒ! . )*(#,;A*H$-)C,/)'()V)>6))<br />
)<br />
W02()*C))#*#+,)VP-.#*+$/#'! 0 8)A0'%)A0')*%&2HJ'%A)&0$/G'(#ƒ! $ &#A0')'%A/$%A)&0$/G'()%(#G*#'%)JB)*A()<br />
/'$&A*,%)C2%&A*,%8)$%")J,A0)C*/H();/,"2&');,(*A*#')$H,2%A(6)<br />
)<br />
*%""+,"+,!"%"--.%$,/,<br />
X/,H) Y'HH$() ,+#$%") !S) O') *%C'/) A0$A) C,/) ,#*#)1+) $-.#*+$/# '!# 0 8) A0') *%&2HJ'%A) &0$/G'()*A()<br />
,;A*H$-)H,%,;,-B)-*&'%(')C''8)ƒ! 2 "##Z'%&')A0')(A$A'H'%A),C)A0')&,/,--$/B)*()A/2')*%)A0*()&$('6)X/,H)<br />
Y'HH$)!78)O0'%)=!TP)R))13#%#-.#*+$/#'!#%4,#*#)1+"#A0'),;A*H$-)%!#56#ƒ! . # "#[2A)*%)A0*()(&'%$/*,)<br />
H,%,;,-B);/,C*A()$/')*%&/'$(*%G)*%)A0')*%A'/#$-)[#%! 7 &#%! 8 ]#$()O')($O)*%)Y'HH$),9#6)W02()*A)H2(A)<br />
J')A0')&$(')A0$A)%!# 2 >ƒ! . )N)%! & ##"#X*%$--B8)O0'%)))#*#+,)VP-.#*+$/#'! 0 )&#A0'),;A*H$-)-*&'%(')C'')C,/)<br />
A0')*%&2HJ'%A)*()G*#'%)JB)ƒ! + )*%)'\2$A*,%)3!??46)<br />
)<br />
ƒ9 +) ")ƒ9 ] )N))")[=!)R)Q)R)?S)3?)R)=#4)K9 , ])T?I))6)<br />
W02()O0'%))#*#+,#:3-.#*+$/#'! , 8)ƒ9) ] >ƒ9 + )N)ƒ9 % )<br />
,<br />
*%""+,"+,0(1"%12,3,<br />
9A) &$%) J') &0'&Q'") A0$A) *%) A0') &$('() O0'%) A0'/') *() 2%/'(A/*&A'") '%A/B) $%") O0'%) A0'/') *()<br />
/'(A/*&A'") '%A/B) O*A0,2A) H,%,;,-B) ,2A;2A8) %!#
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'()*+!,$%-,""!<br />
./0123456!781285!95:5;05*&!GHHI+!","E-F%)J!(K81285*! !<br />
6$%#'2&)*7(38$#'#'("*&"0*9,#/(,%2'":*'"*&*;,88)5CM=!N5:48717CM=!O:787;2:3=!HNPQNOR!R783/1>28C=!S/82:4=!T55:U:VML477B:7;!<br />
4)%5+*-%*6+$%".*+51-3%<br />
WL83!XB!T7>>28C52>278!L86!K/>37/L8:53! 28! ;L8/[L:>/2783! 7[!<br />
17:]-283!:56!0M!2887\L>2\5!3/??125551! L86! S2:! L3! 3/??1257!<br />
653]>7?! ;L8/[L:>/278-0L356! 17:]-283B! G86556=! >453! L8! 5\71\28C!<br />
?7A5/3!;5;052>278! L3!<br />
L665C2:L11M!0M!.755Z>!<br />
23! 58\23LC56! 28! A42:4=! L>! L8M! C2\58! ?728>! 28! >2;5! 7[! >45!<br />
278342?=! 07>4! >45! 0/M545! 3/??125278! 32;/1>L857/31MB! X5! 4L>! >45!<br />
?278! [7;58>3! 28! 2887\L>278! 0M!<br />
;5;0545!3/??1M!:4L28!23!>7!28:4522L1!<br />
72\5! :4L8851! ?7A545! 3/??1M! :4L28B! Q137! 28!L!<br />
32>/L>278!A4545!0/M545!0/M52\51M! ?/45! :278! 7[! L! 3/03>2>/>5! >5:48717CM!<br />
0M! 28\53>28C! 28! 2887\L>278B! N45! ?278! [745!<br />
0/M5423!:L35!A7/16!05!>7!512;28L>5!>45!>5:48717CM!17:]-<br />
28!L86!05:7;5!2865?58658>B!<br />
X5! :L8! 177]! L>! 47A! >42:L1!28>5C278+!<br />
$*!H>L?153!28!>45!7[[2:5!?3!286/3>23>5?=!<br />
42C4-:73>! \L1/5! :4L28B! N45M! 58\2327856! L! [/>/7;556! 0M! 3475828C! >45! :4L28! >7! :L?>/423! \L1/5! ?278! A2>4! L! ;7651! >4L>!<br />
3/??756! 2>3! L21! 3>74! 2>3! 7A8! :58>278! :58>557! 512;28L>5!<br />
;26615;58!L86!28:
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/3!>4L>!>45!:L!L3!L8!5ZL;?15!<br />
7[! i>45! 65:7;?732>278! 7[! >45! \52:L11M! 28>5C56! 0/328533!<br />
;7651j![45!785!>4L>!/356!>7!05!\55C56!L3!?L!7[!<br />
L!342[>!28!L!\L1/5!85>B!G8!3/:4!:2L8:53=![2545!<br />
:5CM! 7/C4>! >7! 05!28! ?1L:5! >7!<br />
65[586! >45! L0212>M! >7! L??5! \L1/5! 0M!<br />
583/4L>! >45! 3/??12545! ! >45! [2! >7! 7A8! L! L015! >7! ?/>! >45;351\53! 28! L!<br />
?732>278!>7!15\57!>45![2=! ?4L>! 2>! 37/45! ;2:3>L8628C!3/::533[/1!?!6532C8!0/>!L!623L3>51!28!>45!<br />
3/??1M! :4L28! [756! 0/>! :4L1158C56! 0M!<br />
Q6\L8:56!S2:2783!>4L>!<br />
7::/423! 286/3>58>! 65[2856! 0M! >423!<br />
/?3>7!<br />
05AL45!bG8>51!G83265c!3M8658628C! >423! L! >7! >45! /?3>2783B!9/45! $d)#3=! >45! ?2:5! 7[! 35:786! 37/! :L8! L:>! L3! L! 35:786! 37/4L>!37;5!3715!<br />
3/??125756c!>4527;545M! 4L6! 17:]56! 28! >452! 6532C83! >7! >4L>!7[!>45!<br />
3/??125/45! 286/3>45!5\71/>278!7[!;L!3>/>2\5! 3>5CM! [757! ?1M! A4215! 7[[52:L1! >7! >4L>! 7[! >45! 2887\L>28C! [2423!<br />
4/!:7/16!/35!2>3!C7!<br />
3/??7!L!1L4!L!\25A!>7!65\517?28C! 2>3!<br />
7A8!85Z>!C585278!?3B!^756!<br />
L3!L!35:786!37/54L8!4L1[!L!67e58!:7;?L8253!A578=!"##$*B!<br />
Q8! 5ZL;?15! :L8! 05! [7/86! 28! >45! 3/??1M! :4L28! 7[!<br />
L/>7;702153! (R1L7=! $dd$=! R4L?3B%=! )*B! Q/-<br />
>7;70215! ?278! 05C283! A2>4! 6532C8=! A42:4! :78323>3! 7[!<br />
>43B! N45![2!515;58>!23! >45!:78:5?>!2>351[=!<br />
>45545!6532C8!7[!>45! \542:15!:L8!05! /35[/11M! 62\2656!<br />
28>7!>45!;L:45!65\517?;58>!7[!>45!0L32:!:4L3323=!<br />
3/0-L335;01M! L86! :7;?7858>! 3?5:2[2:L>278*! L86! >45! ;2:45! 65\517?;58>=! >7! LC2783=! 7[! >45!<br />
\542:15l3!:783>2>/58>!:7;?7858>3*B!N45![2!>A7!515;58>3!7[!<br />
6532C8! >586! >7! 05! /865L]58! 0M! >45! :L2:/1L45!L335;015L]53!:4L!7278B!<br />
G8! >45! [L:5! 7[! 28>5835! :7;?5>2>278=! 47A5\545! :73>3!<br />
L337:2L>56! A2>4! 65\517?28C! 85A! \542:15! ?7>M?53! 4L\5!<br />
28:3! (>45! ;2:7!<br />
C170L1! 3>L86LL]5! [/11! M! [745! 6532C8! 7[! 3/0-<br />
L335;01253!L86![745!:77278!7[!>45!35:786!L86!>4225!;L8/[L:>/!:78>5!>7!>45!?B!<br />
N45! 3>278! 7[! >4535! 3/0-L335;015278B! I7! 328C15!<br />
3/??125M?53! 7[! 3/0-L335;01MB! ^745! A77;7>2\5! 5_/2?;58>!<br />
;L8/[L:>/3! 5[[73! 78! 3>L55;3=! 3?L28C!L86!A2863:278!23!>4L>!>45!3/??1M!7[!<br />
?L2:/1L5;3! 4L3! 05:7;5! :78:58>56!<br />
L;78C3>!D/3>!L!4L86[/1!7[!;L8/[L:>/5:48717C2:L1!:7;?5>2>278!23!/356!>7!5ZL;285!>45!:7;?5>2>278!<br />
05>A558!>45!0/M545!3/??1252CL>278!7[!>423!>M?5!<br />
7[! 0/M5278! A7/16! 15L6! >7! L! 05>>5L8628C! 7[! >45! 6M8L;2:3! 7[! :711L07278! L;78C!<br />
3/??1M!:4L28!?L85528C! L3?5:>! 7[! >45! ?45! [L:>! >4L>! >45!<br />
3/??125!>L]5!28>7!L::7/8>!>45!2845!28:58>2\53![745!<br />
0/M57! 65\517?! L! b0L:]3>7?c! >5:48717CM! A42:4! :L8! 05!<br />
3/03>2>/>56![745!3/??1256!:7;?7858>B!N45!3/??1254!>45!<br />
]87A156C5! 7[! >423! 28>58>! 7[! >45! 0/M53! 3/:4! >4L>! >45!<br />
?3! L278! 7[! 3/03>2>/>5!<br />
>5:48717CM! 0M! >45! 0/M52;5! A458! 3/:4! 2887\L>278!<br />
;L>5!:L8!05!L[[5:>56!0M!Pk9!5[[73B!<br />
N45! ;7651! [7278! 23! 56! >7! L! :L35! 28! >45!<br />
5:787;2:3! 7[! 5Z4L/3>2015!
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,$F! WL83!XB!T7>>28C52>278!L86!K/>37/4L>!23!355]28C!>7!<br />
28\58>! L! 3/03>2>/>5! >5:48717CMB! N45M! L66278! 28! >45! 5:787;2:3! 7[! 5Z4L/3>2015! 5:48717C253! >7! 3/03>2>/>5! [75! 8L>/56! >7! 623:7\57?! >5:48717C253! 87>! 781M! >7!<br />
7\545!?4L>!!L137!<br />
>7! 5:48717C2:L1!<br />
:7;?5>2>278! 23! ?56! 28! H5:>278! "B! G8! H5:>278! ,! >45!<br />
62[[52L1! CL;5! 23! L8L1Me56! L86! >45! SL1"*+)#%<br />
N45! 87>L>2783! L86! 3M;0713! /356! 28! >45! ;7651! L![745!0/M55!>2;5!475C253B!<br />
>+!Q8!283>L8>!7[!>2;5!28!>45!6M8L;2:!CL;5!35>/?B!<br />
/ a (>*+!a/M5278!5[[7!(Pk9!5Z?5862>/2;5*B!<br />
/ H (>*+! H/??125/!>7!05!28:/!>7!05!28:/!28:/4L>! A7/16! 05! L! ?5!<br />
3/03>2>/>5![745!3/??125278B!<br />
!+!L8>!7[!>2;5!L>!A42:4!>45!15\51!7[!2887\L>278!<br />
(04278*!23!*+!>45!?M!>4L>!>45!0/M28C![245!2887\L>278!?!0M!>2;5!>=!>4L>!23!.*!p!^(>*B!<br />
] #$ +![/8:>278!45!3A2>:428C!5B!<br />
Z(>*+!3>L>5!6M8L;2:3!5Z>!7[!0/M54L>!23!0L356!78!<br />
2887\L>278! L86! _/L12>M! ?L5B! Q! 785-! >7-! 785!<br />
:745!0/M57!>45![28L1!:783/;5!15L3>!785!:7;?7858>!23!<br />
! :7;?5>2>278! 28! >45!<br />
[28L1!:783/;5B!<br />
G>! 23! L33556! >4L>! L! 785! q! >7! -! 785! :745!L8L1M323!28!>A7!623>28:>! ALM3B!^2=!<br />
>423!L117A3!5Z?12:2>!5ZL;28L>278!7[!65?58658:5!05>A558!>45!<br />
>A7! :711L0728C! ?L8545! 5\71/>278! 7[! ?7A5/278B! H5:786=! >423! 15863! L8L1M>2:L1!<br />
32;?12[2:L>278! [745! ?! [745!<br />
:73>!7[!?278!23!: H (/ H *!m!#!?5B!H/??1255!7[!<br />
?278! 23! L! [/8:>278! 7[! ?278! 5[[7! / H !L86!23!<br />
L33/;56!>7!05!"/ H B!^72:L1!32;?12:2>M!"!23!L33/;56!>7!<br />
05!5_/L1!>7!$B!N45!5!7[!?278!23!65[2856!L3=!/ H +!(#=#*!<br />
$! [#=! / H ]! N45! 45! 3>L>5! \L45! :78>5Z>! 7[! >45! ;7651! 3?5:2[2:L>278! L86! >45!<br />
/8652783=!>45!2:L1!<br />
[745!0/M545!3/??125278!7[!>45!3/??12545!5!7[!:4L8C5!7[!*!p!-!"!/ H !(Z(>**!p!%!/ H !(Z(>**&!Z(#*!p!Z 7B!<br />
Q3! >45! ?278! 5! 28:45! 278! 5! 7[! >45!<br />
3/??1257!:5!L!3/03>2>/>5!>5:48717CM!<br />
A42:4! 23! L33/;56! >7! 58L015! ?278! 7[! >45! ?!<br />
:7;?7858>!L>!L!/82>!:73>!: a !r!: H B!G8!>45!5\58>!7[!L!3/::533[/1!<br />
:278!7[!3/:4!L!>5:48717CM!>45B! N45! 28\53>;58>! ?712:M! 28! 2887\L>278!<br />
:47358! 0M! a! 23! 6587>56! 0M! / a +! (#=! #*! $! s#=! #*B! N423!<br />
28\53>;58>! 45! 0/M5278! 7[! >45! *B! aM! 28\53>28C! 28!<br />
2887\L>278=!>45!0/M545!>2;5!?L>4![745!0/M5*!23!C2\58!0M=!<br />
] • (>*!p!/ a !(Z(>**=!](#*!p!#B!<br />
N45!7/>:7;5!7[!2887\L>278!28\53>;58>3!23!:7832657!05!<br />
/8:5L28B! N45! >45! Pk9! 28\53>;58>3!<br />
678l>!?545!6L>5!7[!3/::533!7[!2887\L>278B!Q!<br />
8L>/!ALM!>7!278!<br />
23!>7!3/??735!>4L>!>45!283>L8>L857/3!?M!7[!>45!0/M528C=! :7862>278L1! 78! 87>! 4L\28C! 37! [L278! 7[! 2>3! :/!23!L33/;56!>7!<br />
>L]5! L8! 5Z?7858>2L1! [7! 7[! >45!<br />
L::/;/1L>56!]87A156C5!15\51=!](>*B!<br />
^(>*!@![$!q!^(>*]!p!/ a !(Z(>**!!!!!!!!!!!!!!!!!!!!!!!($*!<br />
A45*!23!>45!?M!>4L>!>45!0/M545!2887\L>278!?!0M!>2;5!>&!>4L>!23!.*!p!^(>*B!<br />
`5>! u(/ H *! p! s?(/ H *! q: H (/ H *v! w/ H ! 05! >45! 3/??125L8>L857/3! ?! [/8:>278B! N423! 283>L8>L857/3! ?!<br />
[/8:>278! L:>3! L3! >45! 70D5:>2\5! [/8:>278L1! [745! 3/??12545!?5Z>=!>45!0/M5!:7;?5>2>278B!N45!<br />
0/M53! 28! 2887\L>278! >7! 05:7;5! 2865?58658>! 7[! >45!<br />
171
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/! 7[! ?278! 5[[7! : H (/ H *! 23! :78>28/7/31M!<br />
62[[52L015!78!s#=( / H *+!: H (#*!p!#!78!s#=( / H *!L86!:l H (M*!23!L!<br />
:783>L8>B! N45! :73>! 7[! 2887\L>278! 5[[7! : n (/ a *! 23! >A2:5!<br />
:78>28/7/31M!62[[52L015!78!s#=!#*&!: n !(#*!p!:) n !(#*!!p!#&!<br />
:+ n !m!#!78!s#=!#*&!L86!:) n ! (M*! $! #! L3!M!$! #B! N423!<br />
L33/;?>278! 3/CC53>3! >4L>! :73>3! 28:!L137!3/CC53>3!>4L>!e52;L1!L86![745!3>5CM!/ a! A5!:L8!:78[285!7/5!<br />
>7!28>5423!;L!35>>28C=!2>!23!L33/;56!>4L>!>45!?!58D7M3!<br />
L! \5278!<br />
5! 7[! >45! 3/??12545! _/L8>2>M! 65;L8656! ?5!<br />
>2;5!28!>45!;LB!N45!28\5278!74545!?278!7[!/82>3!?2;5B!<br />
?(/ H (Z(>***pL%0"/ H (>*pLq0/ H (>*&!Lm: H m#=!0m#!LL8>3!("*!<br />
N45!\L!7[!?278!23!L33/;56!>7!05!1285L278!7[!?278!5[[7!/ H B!<br />
: H !=!: $ ![/ H]=!: $ !m!#!23!L!?M!:783>L8>!!!!!!!!!!!(,*!<br />
!"#$%&'()*7("(8()9*:%(0,2#'("*&"0*:%'2$6*<br />
N45! 3>5C253! / H (B*! L86! / a (B*! :783>2>/>5! L8! 5_/212045!<br />
?278!5[[7!/ H (B*!;LZ2;2e53!>45!3/??12545! 2887\L>278! 5[[7! / a (B*! ;LZ2;2e53! >45! 0/M55C253B!N45![7117A28C!L33/;?>2783!L45! ?!<br />
[/8:>278!L86!>45!:73>!7[!5[[7![/8:>278B!<br />
(2*! N45! 28\5278! ?(/ H *! 23! >A2:5!<br />
:78>28/7/31M!62[[52L015!78!s#=( / H ]+!<br />
A45!7[!5[[7![/8:>278![745!0/M57!05!<br />
3>1M!:78\5Z!L86!28:278!5[[7!/ a B!<br />
: n !=!: " ![/ a] " = ! : " !m!#!23!L!?M!:783>L8>!!!!!!!!!!('*!<br />
I7>5!>4L>!>423!?45!3L;5!3>L>5!6M8L;2:3![745!<br />
0/M545!3/??1252\5![/8:>278L1!7[!3/??12545!5Z?5:>56!?!/8>21!>45!L8>!7[!>2;5!!=!<br />
f H (Z*!=!O /a(B*! # ! !5 - !s?(/ H (Z(>***!q: H (/ H *vw/ H !Z(>**]6>!<br />
p!O /a(B*! # ! !5 - ![L!-!0/ H (Z(>**!%!: $ (/ H *(Z(>**/ H (!Z(>**]!6>!!!(%* !<br />
I7>5!>4L>!>45!3/??125![7!m!!!(328:5!<br />
65;L86!\L823453*B!N45!70D5:>2\5![/8:>278L1!7[![245!<br />
5Z?5:>56!CL28![278!5[[73=!C2\58!0M!<br />
f a (Z=!/ a *!=!O /a(B* !! # ! !5 - s&(/ H (Z(>***!q!: n (/ a *(Z(>***]6>!+!! ! # !5 - !' a !(/ a *6>!<br />
p!O /a(B*! # ! !5 - !s&(/ H (Z(>***!q: " (/ a *(Z(>***] " 6>!+!! ! # !5 - !' a !(/ a *6>!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!()*!<br />
&(/ H (Z(>***!=!! # /H !s?(M*6M!q/ H! (Z(>**?(/ H! *!=!! # /H ![L!-!0M]6M!%/ H (Z(>**![L!-!0/ H (Z(>**]!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(F*!<br />
23!>45!3/
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,$d! WL83!XB!T7>>28C52>278!L86!K/>37/***!q: H (/ H *v/ H Z(>**]6>!<br />
=!O /a(B*! # ! !5 - ![L!-!0/ H (Z(>**!%!: $ (/ H *(Z(>**]/ H (!Z(>**6>!<br />
3/0D5:>!>7!Z • (>*!p!%!/ H !(Z(>**&!Z(#*!p!Z 7 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(E*!<br />
f a (Z=/ a *!=!O /a(B*! # ! !5 - !s&(/ H (Z(>***q: n (/ a *(Z(>***]6>!+!! ! # !5 - !' a !(/ a *6>!<br />
N45!62[[52L1!CL;5!65[2856!0M!(E*!L86!(d*!05178C3!>7!>45!<br />
:1L33! 7[! ?25:5A235! 65>52:! CL;53! (97:]85! L1B=!<br />
"###*+! N45! CL;5! 4L3! >A7! ;7653! 781M+! ;765! #! 23! L:>2\5!<br />
05[7278!L86!;765!$!05:7;53!<br />
L:>2\5!2[![245!3/03>2>/>5!>5:48717CMB!<br />
N4/3=! >45! ;73>! 785! 3A2>:4! 7[! ;765B! N45!<br />
3A2>:428C!>2;5!23!>45!278!7[!!!23!^=!L3!C2\58!28!5_/L>278!($*B!<br />
C'%7#"1:$+$%)A%*6-%D)2-1%<br />
N45!;7651!23!L8!5Z>583278!7[!L!?>28C5***] " 6>!+!! ! # !5 - !' a !(/ a *6>!<br />
3/0D5:>!>7!Z • (>*!p!%!/ H !(Z(>**&!Z(#*!p!Z 7 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!(d*!<br />
3>L>278L278L! n H! (4=!Z*!23!L!<br />
;L??28C!n H +!S!,(#=!#*!$!s#=!( / H v=!A4545!<br />
35>! 7[! ;7653B! a/M5L>278L278! L??2783! 7[! >45! 3/??1252\51M![745!65[282>278!L86!>4577! ?25:5A235! 65>52:! 62[[52L1! CL;53!<br />
?45!Q??5862Z=!>4535!\L1/5![/8:>2783!L7!05!07/8656!L86!:78>28/7/31M!62[[52L015!L86![74L>! 3L>23[M! >45! WL;21>78-fL:702-a511;L8! (Wfa*!<br />
5_/L>2783B!<br />
3! 7[! / H-! s#=! ( / H v!L86!/ a-! s#=#*! >4L>!<br />
;LZ2;2e5!>45!2\5!\L1/5![/8:>2783!7[!>45!3/??12545!0/M5278!($#*!L86!($"*!L56!L3!0 H (Z*!L86!<br />
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/45! :73>! 7[! 5[[7! [/8:>278! [2783!("*!L86!5_/L>278!('*=!2\51MB!<br />
T H (e*!=!;LZ!{L!%!0/ H !%!: $ !/ H !]!/ H !%!/ H !e.#!1!/ H!1!( / H}!!($d*!<br />
T a (e*!=!;LZ!{e/ a !%!: " (/ a * " .#!1!/ a !}!!!!!!!!!!!!!("#*!<br />
n H (#=Z**!=!L5!>45!?732>2\5!5_/212045!3/??125458!!<br />
! "# $<br />
%&"'&<br />
"&#()*+$' , - . *%/ 0 $)' . -1&1% . +' !!!!!!!!!!!!!!!!!!!("'*!<br />
! 2# $ 134 & 51$1%6 7 111&1 "&#()*+$' ,- . *%/ 0 $)' . -1&1% . +<br />
!!!!!!!!("%*!<br />
"&*+$' , -<br />
;5$(%$3*=6%G[!n H 2 !L86!n a 2 !6587>5!>45!?732>2\5!5_/21202783!7[!>45!3/??125458!<br />
8 9 # $ :;! < 9 *=>?-1<br />
3*@$4 A -1<br />
8 B # *=> ?- $ 6C 1&16 C *=>?-<br />
$ 1:<br />
3*@$4 A - !!!!!!!!!!!!!!!!!!!!!!!!!!!(")*!<br />
34 3<br />
$ "D5*@E4 A -3 F6 C E54 3 G;: 3@<br />
3"3*@$4 A -"4 3<br />
!!!!!!!!!!("F*!<br />
;5$(%$3*>6!Q>!5_/212045!!>2;5!>!<br />
:L8!05!5Z?$$+)#%<br />
?@/$%A&#'("* ! 5_/212045! 3/??125278!5B!<br />
N45! 7035278! :L8! 05! 5Z?1L2856! 0M! 5Z?1745!<br />
;787?71M!?278!5!7[!>45!3/??12545!C2\58!35>>28CB!<br />
N45!;787?71M!?!7[!>45!3/??12545![2!77!>45!;787?71M!?278!5!C2\58!0M!L!/ H; !=!L/!"(0+: $ *B!<br />
N45!3/??1257!05!L!;787?71MB!G8!>45!C2\58! 35>/?! >45! 3/??125278!5!>7!CL28!L3!<br />
;/:4!7[!;787?71M!?3!L3!?7332015!05[7!7[!>45!;LB!Q8!211/3>278!7[!3/:4!L!;787?71M!?278!23!<br />
?
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
HANS W. GOTTINGER<br />
,"$! WL83!XB!T7>>28C52>278!L86!K/>37/45!3/??1253!L!85CL>2\5!5[[5:>!78!>45!<br />
0/M5278!5B!<br />
N45! 7035278! 3>L>56! 28! "(2*! 23! 28>/2>2\5! L86! :L8! 05!<br />
703545! 5Z?5! 0M! >45! 0/M556B! X2>4! 37;5! L1C502783! >45!<br />
5_/2120278!5!7[!>45!0/M545! L07\5! 5Z?! :L8! 05! 87>56! >4L>! >45!<br />
3/??1252\5!5[[5:>!<br />
78! >45! 0/M5278! 5[[73B! K035278! $! L86! "!<br />
:4L545! 8L>/45!C2\58!35>>28C!>453!87!<br />
;7>2\L>278![745!>A7!?1LM57!:711L075!L86!5L:4!L:>3!28!<br />
2>3!053>!28>5B!N45!3/??12545!;787?71M!<br />
?7A545!0/M528/53!28\53>;58>3!28!2887\L>278!>7!<br />
:5!3/03>2>/>5!>5:48717CMB!<br />
W7A5\52783B! I7>5=! >4L>! 28! 5_/L>278! ("d*! A5! 4L>!A2>4!L!67A8AL278B!N45!28\5
HANS W. GOTTINGER<br />
2.3 VERTICAL COMPETITION AND OUTSOURCING<br />
! G8>5278L1!f7/7!<br />
Q86!<br />
( 63 %<br />
7 8 "9 8<br />
":#; $ 8 +%? @ $>28C52>278j=!<br />
!"#$%"&#'("&)* +(,%"&)* (-* .,/'"$//* &"0! 12("(3'2/* 4$/$&%25!<br />
'("*=!)F-F$B!<br />
sdv! T7>>28C5
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
!"#$%"&#'("&)*+(,%"&)*(-*.,/'"$//*&"0*12("(3'2/*4$/$&%25*<br />
"#$%&!'(")*!+,-,$!<br />
./0123456!781285!9:;!"#$%!(4??@*AABBBC3
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
+G! V:83!WC!S7??28D5;*! ! I/@@1M-R4:28!R77@5?2?278! !<br />
!<br />
:8:1M_53! ?45! /`!:=!28Y!{?!>!#*!g? /I>/`!=!M!}CN45!Y2;3?!42??28D!?2E5!?7!?45!<br />
@728?!M!/865;!?45!3@5/`!C)!<br />
@(_)! *b@;2
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
! H8?5;8:?278:1!e7/;8:1!7Y!`/328533!:86!O@CG#]! :86! /?212_28D! ?45! ;53/1?3!<br />
Y;7E!(')!:86!(%)!D2Z53 !<br />
6!g? /I>/`!=!g? /I>/`[E(/ I (?)>!(/`(?))6?!+!/ I (?)# I !6W I (?)!!!/`(?)#`!<br />
6W`(?)]!!!!!!!!!!!!(+)!<br />
B45;5!?45!Y/8
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
,#! V:83!WC!S7??28D5;*! ! I/@@1M-R4:28!R77@5?2?278! !<br />
!<br />
c / (_)!=!3/@ /I !28Y /`!O _ !(5 !,' 0 /I>/`)!=!28Y /`!3/@ /I !O _ !(5 !,' 0 /I>/`)!>!Y7;!<br />
_!! !f![%`"!($!!!- " )!±!.p!]!/!"![%`"!($!o!- " !!!"!&!-)!o!"!,!($!!!& " )]!($")!<br />
:86!?45!:337!<br />
/`/(_)!=![%`/#`]![($!!&-!)0 + !!!$]!/!(!($!!!& " )(0 + ) " !!!$)!]!($=)!<br />
(")!R:35!`!<br />
N457;5E!"C!HY!?45!65D;55!7Y!E:;[5?!:6Z:8?:D5!7Y!:!3/@@125;!<br />
7Z5;!:!0/M5;!-!3:?23Y253!?45!!-!>!- + !B45;5!- ! !>- + !<br />
:;5!:3!65Y2856!28!(a)>!?458!?45!Z:1/5!7Y!?45!623$$+)#%<br />
W5!
2.4 SUPPLY-CHAIN COOPETITION<br />
HANS W. GOTTINGER<br />
! H8?5;8:?278:1!e7/;8:1!7Y!`/328533!:86!O!:!3/@@125;!B7/16!D:28!0M!E:[28D!<br />
28Z53?E58?3! 28! 2887Z:?278! ?4:?! :;5! 2865@58658?! 7Y! ?45!<br />
28Z53?E58?3!E:65!0M!?45!0/M5;C!c/;?45;E7;5>!:!3/@@125;!B7/16!<br />
:137! D:28! 0M! E:[28D! 28Z53?E58?3! 28! 2887Z:?2783! ?4:?! :;5!<br />
@7332015!3/03?2?/?53!7Y!?45!0/M5;r3!@;76/
HANS W. GOTTINGER<br />
2.4 SUPPLY-CHAIN COOPETITION<br />
182
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
A Tale <strong>on</strong> Chinese Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong>: A Case Study<br />
of the Chinese Telecommunicati<strong>on</strong>s Equipment Industry<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich Germany<br />
www. stratec-c<strong>on</strong>.net<br />
Extended Abstract.<br />
We examine the four largest Chinese suppliers of stored program c<strong>on</strong>trol<br />
(SPC) switches: Datang Telecom <strong>Technology</strong> (DTT), the Great Drag<strong>on</strong><br />
Informati<strong>on</strong> <strong>Technology</strong> (GDT), Huawei Technologies (Huawei), <strong>and</strong><br />
Zh<strong>on</strong>xing Telecommunicati<strong>on</strong> Equipment (ZTE). The tale shows how<br />
these four Chinese suppliers competed against the two largest SPC switch<br />
manufacturers that had foreign joint venture partners :Shanghai Bell an<br />
Alcatel joint venture, <strong>and</strong> Beijing Internati<strong>on</strong>al Switching System (BISC),<br />
a Siemens joint venture , <strong>and</strong> further how they competed globally against<br />
well established suppliers of telecommunicati<strong>on</strong>s equipment (TE)such as<br />
Alcatel, Cisco, Ericss<strong>on</strong>, Fuijitsu, NEC, Nokia, Nortel, <strong>and</strong> Siemens.<br />
First, the fast growth in the period after 1985 of two suppliers, Huawei<br />
<strong>and</strong> ZTE, requires appropriate attenti<strong>on</strong>. Their success threatens well<br />
established incumbents’ plans to dominate the global <strong>and</strong> China markets<br />
for TE. ZTE started operati<strong>on</strong>s in 1985, <strong>and</strong> Huawei in 1988. In slightly<br />
over ten years, Huawei became the number <strong>on</strong>e supplier of TE in China.<br />
In 1998. Huawei’s annual revenues exceeded those of the top two TE<br />
suppliers that had foreign joint venture partners: Shanghai Bell <strong>and</strong> BISC,<br />
<strong>and</strong> since then had the gap growing. Still, in the mid eighties, the<br />
possibility that any of the four Chinese suppliers could pose a serious<br />
threat to established global suppliers seemed very improbable. Today,<br />
Chinese suppliers compete aggressively against these well established<br />
global suppliers <strong>and</strong> their joint ventures in China.<br />
For example, reputable market share tracking firms ranked Huawei the<br />
number <strong>on</strong>e supplier in the global market for new extended switching<br />
equipment in 2003, the number <strong>on</strong>e supplier in the global market for new<br />
generati<strong>on</strong> networks in 2004, the number two supplier in the global<br />
market for digital subscriber line (DSL)access multiplexers in 2003, the<br />
number three supplier in the global market for l<strong>on</strong>g distance wavelength<br />
divisi<strong>on</strong> multiplexers, <strong>and</strong> the number four supplier in the global market<br />
for optical transmissi<strong>on</strong>. Huawei <strong>and</strong> ZTE have been ranked the number<br />
183
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
three <strong>and</strong> eight suppliers in the global market for integrated access<br />
networks. As the Wall Street Journal reported repeatedly in 2003 to 2005,<br />
executives of North American <strong>and</strong> European vendors of TE have become<br />
increasingly c<strong>on</strong>cerned about the head-to-head competiti<strong>on</strong> from Chinese<br />
suppliers.<br />
On the basis of previous models of competitive racing (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006,<br />
2009) we trace empirically the various stages of competitive strength in<br />
this strategic industry.<br />
In the first stage we observe that the four startups targeted the basic need<br />
for infrastructural development in telecommunicati<strong>on</strong>s in Western <strong>and</strong><br />
rural China in supplying low cost telecommunicati<strong>on</strong>s gear to those areas<br />
which were less lucrative for foreign vendors <strong>and</strong> joint ventures <strong>and</strong> were<br />
heavily encouraged by the Chinese nati<strong>on</strong>al government in a sort of<br />
infant industry protecti<strong>on</strong>. This kind of asymmetic competiti<strong>on</strong> separated<br />
the startups from the established players in the Chinese market, call it the<br />
separati<strong>on</strong> stage.<br />
In the course of this stage the’four horsemen’ underwent technological<br />
learning either through indigenous innovati<strong>on</strong> or imitati<strong>on</strong> of some sort,<br />
even industrial espi<strong>on</strong>age in China,therefore gaining competitive strength<br />
<strong>and</strong> competing against foreigners <strong>on</strong> large scale projects in the Chinese<br />
market. This is termed the c<strong>on</strong>vergence stage. When asymmetric<br />
competiti<strong>on</strong> turns symmetric we observe competitive c<strong>on</strong>vergence , in<br />
which each technology’s development is directed at exp<strong>and</strong>ing its appeal<br />
not <strong>on</strong>ly in its own home market but in its rival’s as well.<br />
While the Chinese companies with the implicit support of the Chinese<br />
government c<strong>on</strong>tinued to gain market share against foreign competitors<br />
<strong>and</strong> as their technological learning adavanced product quality at lower<br />
cost they exp<strong>and</strong>ed in actively seeking to bid successfully for<br />
telecommunicati<strong>on</strong>s projects in developing <strong>and</strong> emerging ec<strong>on</strong>omies<br />
where they gained further strength by competing <strong>on</strong> given product quality<br />
<strong>and</strong> lower prices. This is the globalizati<strong>on</strong> stage.<br />
However, whether Chinese companies keep <strong>on</strong> growing outside their<br />
home market will largely depend <strong>on</strong> whether they turn into genuine<br />
sustainable innovati<strong>on</strong> leaders rather than followers. So far they have not<br />
gained a notable footage in advanced markets for smart networks.<br />
184
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
Disruptive Innovati<strong>on</strong><br />
If we c<strong>on</strong>sider the competitive positi<strong>on</strong>ing of high technology firms as a<br />
technology race in which falling behind, getting ahead <strong>and</strong> catching-up in<br />
industry leadership is the name of the game. We may come across<br />
specific situati<strong>on</strong>s that would be c<strong>on</strong>nected with ’disruptive innovati<strong>on</strong>s’<br />
that could lead to a dramatic paradigm shift of that race. The management<br />
of disruptive technologies originated with the substantial works by C.<br />
Christensen et al.(2004) <strong>and</strong> C. Christensen (2005).<br />
Disruptive innovati<strong>on</strong>s introduce a new kind of product or service that is<br />
actually worse initially, as judged by the performance metrics that main<br />
stream customers value. In terms of the Chinese market where firms<br />
came up with simpler appropriate performance level of network gear at a<br />
much lower price for use in rural China those could be identified as<br />
disruptive. They also c<strong>on</strong>tributed to a market separati<strong>on</strong> that<br />
distinguished two different tiers of market segmentati<strong>on</strong> applying to rural<br />
<strong>and</strong> urbanized areas. In an industrial organizati<strong>on</strong> c<strong>on</strong>text incumbents<br />
have a high probability of beating entrant attackers when the competiti<strong>on</strong><br />
is about incremental innovati<strong>on</strong>s, but almost always lose to attackers<br />
armed with disruptive innovati<strong>on</strong>s (Christensen et al, 2004).. At the<br />
bottom line a cumulati<strong>on</strong> of incremental innovati<strong>on</strong>s <strong>on</strong> a particular<br />
design may end up in a disruptive innovati<strong>on</strong> as will <strong>on</strong> the upper end<br />
technologically radical innovati<strong>on</strong>s . This may lead to market situati<strong>on</strong>s<br />
<strong>and</strong> revenue streams of ’winner-take-all’ or ’winner-take-most’.<br />
We may categorize two aspects of disruptive innovati<strong>on</strong>s that relate to<br />
(1) new markets <strong>and</strong> (2) quality characteristics.<br />
(1) Products offered are too expensive, too complicated, too difficult too<br />
maintain.<br />
(2) Product quality characteristics are appropriate,’ good enough’ with<br />
significantly lower pricing. New entrants compete profitably , learn <strong>and</strong><br />
invest in R&D while pricing at deep discounts.<br />
Competitive Phases<br />
In adapting Christensen’s theory of innovati<strong>on</strong> to the Chinese telecom<br />
industry we may distinguish clearly competitive phases such as<br />
separati<strong>on</strong>( ’isolati<strong>on</strong>’ as Christensen’s term), c<strong>on</strong>vergence <strong>and</strong><br />
disrupti<strong>on</strong>. These phases seem to be more technologically motivated<br />
through innovati<strong>on</strong> but may be supported by strategic directi<strong>on</strong> or other<br />
factors.<br />
In a competitive separati<strong>on</strong> (isolati<strong>on</strong>) regime technologies do not interact<br />
(in terms of being replaceable) in the course of their development.<br />
185
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
SUMMARIZING REMARKS<br />
This case study is of interest as a general attempt in identifying Chinese<br />
efforts <strong>on</strong> innovati<strong>on</strong> at the interface of culture, politics, technology <strong>and</strong><br />
industry. The subject is so complex that it is unlikely to find a c<strong>on</strong>sistent<br />
interrelated interpretati<strong>on</strong> of innovati<strong>on</strong> with 'Chinese Characteristics'.<br />
(The well known Japanese growth ec<strong>on</strong>omist Michio Morishima in the<br />
1980s did this for Japan (CUP) <strong>and</strong> ended up with a tax<strong>on</strong>omy which yet<br />
failed to provide a c<strong>on</strong>sistent interpretati<strong>on</strong> of Japanese innovati<strong>on</strong>) In my<br />
brief comments I limit myself <strong>on</strong> a case study based Chinese high<br />
technology industry<br />
To substantiate some of the claims <strong>on</strong> industrial innovati<strong>on</strong> <strong>on</strong>e needs a<br />
set of industry specific case studies such as the <strong>on</strong>e submitted (A Tale of<br />
Innovati<strong>on</strong> <strong>and</strong> Competiti<strong>on</strong> in the Chinese Telecom Industry) . A few<br />
remarks are in order.<br />
Chinese advanced technology <strong>and</strong> science based industries tend to be<br />
biased toward" product innovati<strong>on</strong> through commercializati<strong>on</strong>" --- quick<br />
market access <strong>and</strong> then incrementally improving product performance<br />
(quality) in subsequent releases.Indigeneous innovati<strong>on</strong> could also<br />
involve 'rapid prototyping' in intenti<strong>on</strong>ally or inadvertently bypassing<br />
IPRs in favour of quick market introducti<strong>on</strong>. Indigeneous innovati<strong>on</strong><br />
could also be a camouflage word for nurturing Chinese startups (where<br />
state directed subsidizati<strong>on</strong> acts as the equivalence of private sector<br />
venture capital or angel investment), <strong>and</strong> shielding them from foreign<br />
competiti<strong>on</strong> (at least in the domestic market for a while) in support of<br />
'infant industry' protecti<strong>on</strong> <strong>and</strong> expansi<strong>on</strong>. This is coupled with state<br />
sp<strong>on</strong>sored private entrepreneurship with encouragement of<br />
Schumpeterian entrepreneurship as examples in the Chinese network<br />
equipment industry clearly show. Competiti<strong>on</strong> in the<br />
Chinese domestic market am<strong>on</strong>g indigeneous firms could be fierce <strong>and</strong><br />
adopts characteristics of a 'technology race' but since the domestic<br />
market is so large <strong>and</strong> with rising incomes<br />
dem<strong>and</strong> growing in leaps there sometimes is a lack of incentives to go for<br />
global markets (with some notable excepti<strong>on</strong>s outlined by industrial <strong>and</strong><br />
development policies).<br />
Chinese innovati<strong>on</strong> resembles a 'learning-by-doing' (Arrow) effect as the<br />
Japanese have d<strong>on</strong>e over several decades whereas the Chinese have<br />
accelerated this process by choice <strong>and</strong> through competiti<strong>on</strong>. Indigeneous<br />
innovati<strong>on</strong> through domestic competiti<strong>on</strong> may limit market risk <strong>and</strong><br />
therefore protect the company against total financial failure (bankruptcy).<br />
By industrial policy targeted companies can experiment <strong>on</strong> product<br />
development without severe downside risk. You d<strong>on</strong>'t bet the company if<br />
the product fails in the marketplace, therefore you can come up with<br />
186
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
incremental changes, whether true innovati<strong>on</strong> or not, <strong>and</strong> let the market<br />
decide in the Chinese seller's market without fearing drastic financial<br />
penalties. No Western companies can take these (financial) risks in their<br />
domestic markets besides being blocked by Chinese entry<br />
barriers(reinforced by Chinese regulati<strong>on</strong>)<br />
Christensen,C., Anth<strong>on</strong>y,S.D. <strong>and</strong> Roth,E.A. (2004),’Seeing what’s next:<br />
using the theories of innovati<strong>on</strong> to predict industry change’,<br />
Bost<strong>on</strong>:Harvard Business School Press<br />
Christensen,C. (2005), The Innovator’s Dilemma, New York,<br />
HarperCollins<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W.(2006), Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>,<br />
Routledge, L<strong>on</strong>d<strong>on</strong><br />
---- (2009), Strategic Ec<strong>on</strong>omics of Network Industries, NovaScience,<br />
New York<br />
---------------------------------------------------<br />
187
HANS W. GOTTINGER<br />
2.5 A TALE ON CHINESE INNOVATION AND COMPETITION<br />
188
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!<br />
!"#$%"&#'("&)*+,-'"$--*&".*/0("(1'0-*2$-$&%03*<br />
!"#$%&!'!()&!!*$!& &<br />
+,-./0123&45./52&6,52&#"7&!"#$&'1889)::;;;>?4,9O!DP-/5>4*2Q
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !$&<br />
&<br />
@,.8/9.2\2?0& /5& !""$7& 812& X,@-2?& F1?22& 0,99./2?& /5& 812& U4?2/>5& =4@928/84?0& M53& M0& 812/?& 82=154.4>/=M.& .2M?5/5>&<br />
>.4-M.& @M?Z28& U4?& .45>& 3/08M5=2& ;MQ2.25>81& 3/Q/0/45& M3QM5=23& 9?43,=8& V,M./8N& M8& .4;2?& =408& 812N& 2\9M5323& /5&<br />
@,.8/9.2\2?07& M53& 812& X,@-2?& b4,?& 0,99./2?& /5& 812& >.4-M.& M=8/Q2.N&022Z/5>&84&-/3&0,==200U,..N&U4?&82.2=4@@,5/=M8/450&<br />
@M?Z28&U4?&498/=M.&8?M50@/00/45&'I=454@/087&E298/5>&2=454@/20&;12?2&812N&<br />
`FI[0& 32Q2.49@258& 4Q2?M..& ;M0& ?2.M8/Q2.N& @4?2& @43208& >M/523& U,?812?& 08?25>81& -N& =4@928/5>& 45& >/Q25& 9?43,=8&<br />
'I=454@/087&d=8&<br />
I/>18& 0,99./2?0& /5& 812& >.4-M.& @M?Z28& U4?& /582>?M823& M==200& 0,08M/5M-.N&4,80/32&812/?& 14@2& @M?Z28& ;/..&.M?>2.N&329253&<br />
528;4?Z025,/52&/554QM8/45&.2M32?0&?M812?&<br />
!"##&R,M;2/&;M0&./0823&X,@-2?&d52&M53&`FI&M@45>&812& 81M5&U4..4;2?0M/523&M&548M-.2&U448M>2&<br />
849&F25& 812/?& @M?Z28&<br />
9M?8/=,.M?7& 1M0& -225& 0/5>.23& 4,8& M0& 812& 3M?Z& 14?02& /5& 812& 940/8/45&/5&M&09M8/M.&i-4M?3&>M@2[N[&7&/8&=4,.3&-2&<br />
'W2?N& ?2=258.N7& -2=M,02& 4U& M..2>23& 02=,?/8N& =45=2?50& 7& @M8=123& M53& 8?M=23& 84& E=488& ^44?@M5& '#YBY(& 45& @/./8M?N&<br />
R,M;2/& 1M0& -225& -M??23& U?4@& -/33/5>& 45& 9,-./=& 08?M82>N& M53& /5Q/820& M& @M8=1/5>& /5& 82=154.4>N& ?M=20& M53&<br />
/5U?M08?,=8,?M.& 9?4A2=80& /5& H,08?M./M& M53& U?4@& =4@9M5N& -,0/5200&;M?0&'L488/5>2?7&!""B7&!""Y(N[&'XMZM/&M53&FM5MZM7&<br />
#YY!%&L488/5>2?7&!""B(& &<br />
!"#"(& -,8& @4?2& 092=/U/=M..N& 812N& -481& M?2& 2@-23323& /5&<br />
T5&Q/2;&4U&32Q2.49@258&4U&3N5M@/=&=4@928/8/45&;2&8?M=2& S2/[=1/20&4U&=4@928/8/Q2&08?25>81&/5&81/0&<br />
08?M82>/=&/53,08?N2& ;2& 4-02?Q2& 81M8& 812& 81?22& 08M?8,90&<br />
?'%9-:%@)#5-A*$%)B%@)0A-*+*+)#%B).%<br />
8M?>2823& 812& -M0/=& 5223& U4?& /5U?M08?,=8,?M.& 32Q2.49@258& /5& C-*;).D%=#2>$*.+-$%<br />
82.2=4@@,5/=M8/450&/5&S2082?5&M53&?,?M.&J1/5M&/5&0,99.N/5>&<br />
.4;& =408& 82.2=4@@,5/=M8/450& >2M?& 84& 81402& M?2M0& ;1/=1&<br />
d52&4U&4U&812&@408&/@94?8M58&?20,.80&U?4@&812&/53,08?/M.&<br />
?2V,/?23& .200& 0491/08/=M823& 2V,/9@258& M53& ;2?2& .200&<br />
4?>M5/]M8/45& ./82?M8,?2& /0& 812& 0,>>208/45& 4U& M5& M5M.N8/=M.&<br />
.,=?M8/Q2&U4?&U4?2/>5&Q2534?0&M53&A4/58&Q258,?20&&<br />
3/08/5=8/45& -28;225& 9?43,=8& @M?Z28& =4@928/8/45& M53&<br />
?2>/45M.& 3/09M?/8N7& 258?N& /5& 81402& @M?Z280& ;2?2& 12MQ/.N&<br />
=4@928/8/45&/5&/554QM8/45M?3& ;2& U4..4;& @MA4?&<br />
25=4,?M>23& -N& 812& J1/5202& 5M8/45M.& >4Q2?5@258& M.04& /5&<br />
=45=2980&/5&WM?/M5&M53&E1M9/?4&'#YYD(&M53&6&9?43,=8/45&9?4=20020&M53&=4080(254,0& /554QM8/45& 81?4,>1& Gg_7812?2U4?2& >M/5/5>&<br />
M?2& =450/32?23& ?2.M8/Q2.N& [08M8/=[7& 9?43,=8& @M?Z28&<br />
=4@928/8/Q2& 08?25>81& M53& /5=?2M0/5>.N& =4@928/5>& M>M/508&<br />
=4@928/8/45& /0& 812& @M/5& =1M552.& 81?4,>1& ;1/=1& 812&<br />
U4?2/>52?0& 45& .M?>2& 0=M.2& 9?4A2=80& /5& 812& 258/?2& J1/5202&<br />
=4@928/8/Q2&9?4=200&8MZ20&9.M=2225=27& /5& ;1/=1& 2M=1& 82=154.4>N[0&<br />
9?43,=8/45& 9?4=20020%& 81/0& =4@928/8/45& /0& 4U825& M004=/M823&<br />
32Q2.49@258&/0&3/?2=823&M8&2\9M53/5>&/80&M992M.&548&45.N&/5&<br />
;/81&812&/32M0&4U&M&=4@928/8/Q2&81?2M8&U?4@&/554QM8/45&<br />
S1/.2&812&J1/5202&=4@9M5/20&;/81&812&8M=/8&0,994?8&4U&812&<br />
',5=44?3/5M823(& /554QM8/Q2& M=8/Q/8N& 81M8& =M5& -2& /3258/U/23&<br />
J1/5202&>4Q2?5@258&=458/5,23&84&>M/5&@M?Z28&01M?2&M>M/508&<br />
;/81&812&<br />
! "#!<br />
190
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!C&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
9?4092=8&4U&/58?43,=8/45&4U&9?43,=80&4?&02?Q/=20&81M8&;/..&<br />
=4@9282& /5& 812& U,8,?21& M& 9?4=200& 4U&<br />
.2M?5/5>*-N*34/5>& /5&<br />
/554QM8/457&;2&2\92=8&812/?&32=/0/450&?2>M?3/5>&/554QM8/Q2&<br />
08?M82>N&84&-2&/5U.,25=23&<br />
0/>5/U/=M58.N& -N& 812& /554QM8/Q2& 08?M82>/20& 2@9.4N23& -N&<br />
?/QM.&U/?@0&j&UM/.,?2&84&/554QM82&0,==200U,..N&;/..&.2M3&548&<br />
A,08& 84& .400& 4U& 948258/M.& 9?4U/8& -,8& M.04& ?/0Z& UM../5>& -21/53&<br />
/554QM8/Q2&?/QM.0N&<br />
4U&[08M8/=[&M53&[3N5M@/=[&=4@928/8/45&812&0M@2&./520&<br />
M0& M-4Q27& 812& 08M8/=& 3/@250/45& 4U& =4@928/8/45& /0& ,0,M..N&<br />
?2.M823&84&9?/=2&=4@928/8/457&;1/=1&8MZ20&9.M=2&>/Q25&812&028&<br />
4U&9?43,=80&4?&02?Q/=20&81M8&M?2&@M?Z2823&-N&U/?@0,/3M5=2&U4?&=4@928/8/45&94./=N&9?M=8/=2&/554QM8/45&M53&@M?Z28&3N5M@/=0&/584&<br />
M==4,58& U4?& 812& 9,?94020& 4U& =4@928/8/45& 94./=N1& M5M.N0/0& 4U& =4@928/8/45& ;/81/5& M& ?2.2QM58& @M?Z282&/5&812&=45=2980&4U&=4@928/8/45&<br />
,023&U4?&M00200@258,/3M5=2&32Q2.4923&<br />
/5& 0,-02V,258& 02=8/450& /0& 84& ?2.M82& 81202& 3/@250/450& 4U&<br />
=4@928/8/45&84&9?M=8/=M.&=45=2980&,023&/5&=4@928/8/45&94./=N&<br />
M00200@2587&/5&9M?8/=,.M?&84&@M?Z28&32U/5/8/45&M5M.N0/0M5/]M8/45&@432.07&M53&812&?2Q/2;&4U&08,3/20&<br />
4U&528;4?Z&/53,08?/207&0,>>208&81M8&=4@928/8/Q2&/582?M=8/45&/5&<br />
/554QM8/Q2&M=8/Q/8/20&@MN&8MZ2&3/UU2?258&U4?@0&M==4?3/5>&84&<br />
812&08?,=8,?2&4U&9MN4UU0&4U&/554QM8/Q2&M=8/Q/8/20&84&[;/552?0[&<br />
M53&[.402?0[M5/]M8/45&@432.07&M53&812&?2Q/2;&4U&08,3/20&4U&<br />
528;4?Z& /53,08?/207& 0,>>208& 81M8& =4@928/8/Q2& /582?M=8/45& /5&<br />
/554QM8/Q2&M=8/Q/8/20&@MN&8MZ2&3/UU2?258&U4?@0&M==4?3/5>&84&<br />
812&08?,=8,?2&4U&9MN4UU0&4U&/554QM8/Q2&M=8/Q/8/20&84&[;/552?0[&<br />
M53&[.402?0[&-252U/80&84&812&@M?Z280&<br />
U4?&0,99.N&4U&=4@9M8/-.2&04U8;M?27&-N&2\9M53/5>&812&@M?Z28&<br />
0,=1&81M8&M&>?2M82?&QM?/28N&4U&04U8;M?2&/0&4UU2?23&4?&812&,5/8&<br />
9?/=2&32=?2M020&'0/5=2&U/\23&=4080&4U&04U8;M?2&32Q2.49@258&<br />
=M5&-2&09?2M3&4Q2?&M&>?2M82?&5,@-2?&4U&948258/M.&-,N2?0(&0N082@&@4?2&QM.,M-.27&M53&81,0&/5=?2M02&32@M53&<br />
U4?&/8/=&<br />
?2094502& 4U& 04U8;M?2& 0,99./2?0& 84& 812& /5=?2M02& /5& 812&<br />
=450,@98/45&4U&492?M8/5>&0N082@0
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !O&<br />
&<br />
-2882?&81M5&2\/08/5>&9?43,=80&45.N&-N&04@2&,02?0&'4?&U4?&04@2&<br />
,020(&M0/32&9?/=2&=450/32?M8/4507&/0&<br />
=450/32?23&-2882?&81M5&2\/08/5>&9?43,=80&-N&M..&,02?0&'U4?&M..&<br />
,020(%&125=2&812&/32M&81M8&9?43,=80&=M5&-2&?M5Z23&M==4?3/5>&<br />
84&M&lV,M./8N&.M332?m?4,90&81M8&<br />
=M5&=42\/08&;/81&4.32?&9?43,=8&>?4,90252?M..N&<br />
M004=/M823&;/81&M&9?4=200&;12?2-N&52;&M53&-2882?&9?43,=80&<br />
3/09.M=2&4.32?&4-04.282&9?43,=80&U?4@&812&@M?Z28&<br />
lV,M./8N&;/534;m&2\/0807&4,80/32&;1/=1&U/?@0&=M5548&0,?Q/Q2&<br />
/5& 812& @M?Z28(,/32& 45& 812&<br />
5M8,?2& 4U& @M?Z28& 3N5M@/=0& 81M8& @MN& MUU2=8& M& 9M?8/=,.M?&<br />
/53,08?N>208& 81M8& /5=,@-258& U/?@0& @MN&<br />
1MQ2&3/UU2?258&/5=258/Q20&84&/554QM82&81M5&258?M580/Q25&812&3/UU/=,.8/20&81M8&/5=,@-258&.2M3/5>&U/?@0&@MN&1MQ2&<br />
84&32M.&;/81&3?M08/=&=1M5>20&o&M&U4?@&4U&4U&/58?M*U/?@&<br />
0N082@/=&/52?8/M&o&812&4994?8,5/8/20&U4?&0829*;/02&/554QM8/45&<br />
@MN&?2.M8/Q2.N&UMQ4?&/5=,@-258&U/?@0252?M.&8124?N252?/=& .2Q2.7&<br />
>/Q25& 812& =,??258& .2Q2.& 4U& ,532?08M53/5>& 4U& 81202& 2UU2=80&812&./5Z&-28;225&9?43,=8&@M?Z28&<br />
=4@928/8/45&M53&/554QM8/45,/3M5=2&U?4@&812&./82?M8,?2&45&81/0&/00,27&;1/=1&<br />
/0&92?1M90&-208&M33?20023&45&M&=M02*-N*=M02&-M0/01& eKFE& ;2?2& 0Z/@@23& 4UU& U?4@&<br />
M58/=/9M8/5>&@M?Z28&9M?8/=/9M5807&45.N&84&Z54;&.M82?&81M8&?/QM.&<br />
08M53M?30& -M023& 45& 52;& 82=154.4>/20& 'S/*b/(& -N9M00/5>&<br />
eKFE7& ;4,.3& =?2M82& =4@9.282.N& 52;& @M?Z280& 548& -2/5>&<br />
M58/=/9M823
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!B&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
@M?Z280&812&?M5>2&<br />
4U& 9400/-.2& 0,-08/8,8202& @MN&<br />
MUU2=8&812&08?,=8,?2&4U&812&@M?Z28&M0&32U/5237&U4?&2\M@9.27&-N&<br />
@MZ/5>& 2=454@/20& 4U& 0=M.2& @4?2& 4?& .200& /@94?8M5820&@MN&-2&3,2&84&/554QM8/45&,532?8MZ25&-N&812&U/?@0&/5&<br />
812& @M?Z287& 4?& 3,2& 84& UM=84?0& 4,80/32& 812& @M?Z28& 0,=1& M0&<br />
2\82?5M.&82=154.4>/=M.&32Q2.49@25807&4?&01/U80&/5&=450,@2?&<br />
8M0820&M==4,58&4U&81202&UM=84?0&/0&548&0/@9.N&M&V,208/45&4U&<br />
2\M@/5/5>& 258?N& 9400/-/./8/2020&<br />
4Q2?&8/@22?7&812&8;4&@2?>/5>&U/?@0&@MN&<br />
=,??258.N& -2& 492?M8/5>& /5& =4@9.282.N& 029M?M82& ?2.2QM58&<br />
@M?Z2807& 81,0& ?M/0/5>& 54& =,??258& =4@928/8/45& =45=2?52&/5&812&U,8,?27&0,=1&81M8&81202&U/?@0&M?2&2\92=823&84&<br />
-2&=4@928/84?0&/5&812&U,8,?22?&<br />
014,.3& -2& -.4=Z23%& 0/@9.N& 81M8& =4@928/8/45& M5M.N0/0& @MN&<br />
5223&84&-2&U4?;M?3*.44Z/5>&M0&84&14;&=,??258& @M?Z280&M?2&<br />
./Z2.N&84&2Q4.Q2&/5&U,8,?21./>1823&M-4Q27&M&U/?@&@MN&548&-2&M=8/Q2&45&M5N&=,??258&<br />
@M?Z280&-,8&@MN&-2&32Q2.49/5>&9?43,=80&81M8&;4,.3&=4@9282&<br />
/5&812&U,8,?2M8/45& 45& 812& -M0/0& 4U& 04@2& 9?2*32U/523& /53/=/M>208&812&-252U/80&4U&812&<br />
U,?812?&M5M.N0/0&9?494023&/5&81/0&08,3N20& /5& ?2.M8/Q2&<br />
9?/=20&%&81202&U2M8,?20&M?2&3/0=,0023&-?/2U.N&-2.4;&4U&9M82580&/0&329253258&45&04@2&U4?@&4U&<br />
54Q2.8N7& /554QM8/45& /0& 52=200M?N& 'M.814,>1& 812& ?2.M8/Q2&<br />
/@94?8M5=2&4U&/554QM8/45&/5&812&=4@928/8/Q2&9?4=200&/0&548&<br />
U,..N&208M-./0123(/207&=1M?M=82?/08/=0&M53&9?/=20
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !q&<br />
&<br />
7"C"$HA+63B(.$35$)& &<br />
!! /58?43,=8/45& 4U& 9?43,=80& 81M8& 2@-43N& M& 52;&<br />
82=154.4>N%& &<br />
!! 014?8&9?43,=8&./U2*=N=.20%& &<br />
!! 1/>1&9?494?8/45&4U&@M?Z28&8,?54Q2?&M==4,5823&U4?&-N&<br />
52;& 9?43,=80& '/$*.:%"#2%<br />
G".D-*%F:#"0+5$% %<br />
M53&R,M;2/&81402&7&/5&812&=4,?02&4U&32Q2.49@2587&<br />
?2M..N& 45.N& 81?22& _FF7& `FI& M53& R,M;2/& 1MQ2& -2=4@2&<br />
/582?5M8/45M..N&4U&=45=2?57&M53&812N&M?2&812&34@/5M58&U4=,0&<br />
4U&81/0&08,3N& M53& T>2.& '!""#(& 08,3/23& 81202& U4,?& J1/5202&<br />
0,99./2?0& ;/81& 8;4& A4/58& Q258,?207& ^2/A/5>& T582?5M8/45M.&<br />
E;/8=1/5>& EN082@& '^TEJ(7& M& E/2@250& A4/58& Q258,?2& M53&<br />
E1M5>1M/&^2..7&M53&H.=M82.*a,=258&A4/58&Q258,?24Q2?5@258& M00/08M5=2& ;2?2&<br />
/@94?8M58&84&J1/5202&0,99./2?0[&08?M82>/20&;1/.2&U4?2/>5&A4/58&<br />
Q258,?20&?2./23&45&U4?2/>5&=4@9M5/204Q2?5@258(4Q2?5@258&1M0&8?/23&84&32=?2M02&812&/@94?8&<br />
4U&E+J&0;/8=120&23&82.291452&492?M84?0&84&,02&<br />
9?43,=80&U?4@&5M8/45M.&@M5,UM=8,?2?0&84&`1M5>&M53&T>2.&'!""#(&7&/5&812&2M?.N&#YY"0&<br />
J1/5202*4;523& U/?@0& 12.3& 45.N& M-4,8& #& 9=& 4U& 812& @M?Z28&<br />
01M?2&4U&=258?M.&4UU/=2&2\=1M5>20M/523&<br />
@4?2& 81M5& 1M.U& 4U& 812& 5M8/45M.& @M?Z28& 0/5=2& #YYq&<br />
08?45>& =1M..25>20& U?4@& 34@208/=& U/?@0& M53& .4;2?& 9?4U/80&<br />
04@2&.2M3/5>&6M9M5202&U/?@07&0,=1&M0&b,A/80,&M53&XIJ7&1MQ2&<br />
;/813?M;5& U?4@& J1/5M[0& E+J& @M?Z287& M53& 04@2& .2M3/5>&<br />
S2082?5& E+J& 9?43,=2?0& 1MQ2& -225& =450/32?/5>& .2MQ/5>& M0&<br />
;2..M/508&A4/58&Q258,?20&M53&U4?2/>5&=4@928/8/45&/5&<br />
?,?M.& M53& ?2@482& J1/5202& @M?Z280& ;1/=1& ;2& =M..& 812&<br />
029M?M8/45&91M02&4U&=4@928/8/45&812N&1MQ2&-225&.2M?5/5>&45[&<br />
=4,?027& =45Q/=8/45& M53& =M9M-/./8N[& 'T^K& ^,0/5200& T508/8,82&<br />
45&J4@9M5N&L?4;81(&81?4,>1&=M8=1,9&3?/Q25&/554QM8/45&84&<br />
812& 2\8258& 81M8& 812N& =4,.3& =4@928282& 12M3& 45& ;/81& @MA4?&<br />
U4?2/>5&0,99./2?0&4?&812/?&A4/58&Q258,?20&/5&@MA4?&,?-M5&M?2M0&<br />
M.45>& 812& =4M08M.& =?20=258& 4U& J1/5M7& 08?28=1/5>& U?4@& Rf7&<br />
E125]257& E1M5>1M/& M53& ^2/A/5>25=2&91M02& 4U& 1/>1&<br />
82=154.4>N& U/?@0& M0& M& 82=154.4>N& ?M=2& /5& ;1/=1& UM../5>&<br />
-21/537&>288/5>&M12M3&M53&=M8=1/5>*,9&84&/53,08?N&.2M32?01/9&<br />
/0&812&5M@2&4U&812&>M@2&'L488/5>2?7!""B7J1M9.2& U/\23& ./52& =M??/23& M& O7"""& G2@/@-/&<br />
'GK^(&=408&7&M53&4U825&M54812?&O7"""&GK^&;M0&522323&M0&<br />
M5&2\923/8/5>&U22&'_2S40Z/57!""#(&/5U?M08?,=8,?M.&?2U4?@&M53&>?4;81&/5&<br />
J1/5M&/8&;M0&32=/323&84&,02&82.2=4@@,5/=M8/450&2V,/9@258&<br />
0,99./2?0&84&9?4@482&812&32Q2.49@258&4U&545*A4/58&Q258,?2&<br />
=M9M-/./8N& /5& J1/5M& M0& 2M?.N& M0& @M?Z28& 8?M50/8/45/5>& ;M0&<br />
8M?>2823&'X4,>18457!""q7&J1M9&528;4?Z&2V,/9@258&@M?Z28&4?/>/5M..N&<br />
U4,?& =4@9M5/20& ;2?2& /3258/U/23& M0& Z2N& 2\9452580& 4U& 81/0&<br />
08?M82>N)& 6,.45>& 'L_F(7& _M8M5>& '_FF(7& `145>\/5>& '`FI(&<br />
52;& 9?43,=8& 32Q2.49@258& /5[M99?49/M82[& 82=154.4>N& 81M8&<br />
! "'!<br />
194
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
!D&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
=4,.3&.2M3&84&M&9M?M3/>@&01/U8&4U&81M8&?M=22524,0&U/?@0&=M@2&,9&<br />
;/81&0/@9.2?7&0,UU/=/258&92?U4?@M5=2&.2Q2.&4U&528;4?Z&>2M?&M8&<br />
M&@,=1&.4;2?&9?/=2&U4?&,02&/5&?,?M.&J1/5M& ;1/=1&=4,.3&-2&<br />
/3258/U/23& M0& M99?49?/M82& 82=154.4>N@258M8/45&M99.N/5>&84&?,?M.&M53&,?-M5:@28?494./8M5&M?2M0.N7& 812& 0M@2& 844.0& 4U& @M?Z28& 92528?M8/45& ;2?2&<br />
M99./23&-N&J1/5202&U/?@0&.M82?&45&45&/5&4812?&2@2?>/5>&M53&<br />
8?M50/8/45M.&2=454@/20& &<br />
Q/?8,M..N& M..& M?4,53& 812& ;4?.32823& M8& ?,?M.& M53& ?2@482& M?2M0& 4U& J1/5M7& =.2M?.N&/5&M&<br />
32Q2.49/5>&2=454@N&7&=4,.3&-2&82?@23&[?2Q2?02&/554QM8/45[&<br />
/5&812& @2M5/5>&4U&L4Q/53M?MAM5&M53&F?/@-.2&'!"#!(&0/5=2&<br />
812N& 4?/>/5M823& /5& 812& 52230& M53& ?2V,/?2@2580& 4U& 81/0&<br />
/5U?M08?,=8,?M.& 02=84?& 81M8& ;M0& 029M?M82& 84& 812& M3QM5=23&<br />
,?-M5&M?2M0&/5&J1/5M4?/]2&8;4&M092=80&4U&81202&/554QM8/450&81M8&<br />
?2.M82&84&812&3/=1484@N&4U&812&82.2=4@&@M?Z28&/5&J1/5M&/5&812&<br />
2M?.N&#YY"0)&<br />
'#(&52;&@M?Z280&M53&'!(&V,M./8N&=1M?M=82?/08/=0443&<br />
254,>1[&;/81&<br />
0/>5/U/=M58.N& .4;2?& 9?/=/5>& M8& 3229&<br />
3/0=4,580"#"$E3B@(+&+&=($GI,.(.$ $<br />
@M?Z280&0/>5M..23&812&08M?8&4U&812&=4@928/8/Q2&=45Q2?>25=2&<br />
9M815&0,99./2?0&=4@92823&12M3*84*12M3&<br />
U4?8&12&,?-M5&M53&?/?M.&J1/5202&@M?Z28&M53&8125&/5&812&>.4-M.&<br />
@M?Z28?2M82?&<br />
5,@-2?0& 4U& 3/>/8M.& 0,-0=?/-2?& ./520& '_Ea(& /5& J1/5M& 81M5&<br />
U4?2/>5& 0,99./2?0& /5=.,3/5>& 812/?& A4/58& Q258,?20&<br />
3/3'SE67!""C%_/88-2?52?7!""C(&=45=2?523&M-4,8&812&<br />
U/2?=2& =4@928/8/45& U?4@& J1/5202& 0,99./2?0& /5& 812& >.4-M.&<br />
@M?Z2805M..23&M.?2M3N&M&8?M50/8/45&U?4@&812&=4@928/8/Q2&<br />
=45Q2?>25=2& 9M81& 84& 812& =4@928/8Q2& >.4-M./]M8/45& 9M81& U4?&<br />
J1/5202& 0,99./2?025=2& 9M81&<br />
82=154.4>/20& 2Q4.Q2& 84& =4@9282& 12M3*45& U4?& 812& 0M@2&<br />
=,084@2?05&<br />
0,99./2?0& 04.3& 1/>1& =M9M=/8N& E+J& 0;/8=120& 84& 0M8/0UN& 812&<br />
52230&4U&.M?>2&7&UM08&>?4;/5>&=/8/202823& 812& /5U?M08?,=8,?M.7&<br />
545*=450,@2?&@M?*Z28&M53&M3323&QM.,2&84&UM=/./8M82&>?4;81&<br />
9?4092=80&/5&,532?32Q2.4923&J1/5202&?2>/450&=2582?0'.2Q2.O(1M/&<br />
^2..[0& 9?43,=80& ;2?2& 04.3& 84& .M?>2& 3/08?/=8& =2582?& '.2Q2.&<br />
#(79?4Q/5=/M.&0;/8=1/5>&=2582?0&'.2Q2.&!(&M53&=/8N&0;/8=1/5>&<br />
=2582?0&'.2Q2.&$(4&U?4@&.M?>2&84&0@M..&<br />
>24>?M91/=&,5/805& 0,99./2?0&<br />
M53&812/?&A4/58&Q25,?20&M==4,5823&U4?&qD9=&'KTT7&!""C(& & &<br />
H54812?& ?2M045& U4?& 812& 029M?M8/45& -28;225& J1/5202&<br />
0,99./2?0& M53& 208M-./0123& 0,99./2?0& ;M0& 81M8& ?2>,.M84?N&<br />
?2V,/?2@2580&-N&812&+408&M53&F2.2=4@@,5/=M8/45&^,?2M,0&<br />
'+F^0(& /5& ?,?M.& M?2M0& ;2?2& V,/82& 3/UU2?258& 81M5& 812&<br />
?2V,/?2@2580& 4U& 812& 9F^0& /5& 812& =/8/20&<br />
F1,07&/5&2\9.M5/5>&812&91M020&4U&@M?Z28&32Q2.49@258&U4?&<br />
J1/5202&U/?@07&<br />
/5&82?@0&4U&1/084?N&;2&@MN&/3258/UN&812&029M?M8/45&91M02&<br />
?4,>1.N& ;/81& 812& 92?/43& 4U& @/3& #YD"0& 84& @/3& #YY"07& 812&<br />
=45Q2?>25=2& 91M02& U?4@& @/3& #YY"0& 84& 2M?.N& !""0& M53&812&<br />
>.4-M./]M8/45&91M02&U4?&812&?208&,58/.&54;/20&'./Z2&E+J& 0;/8=120(&<br />
-,8& M.04& 84& -?4M3& 9?43,=8& >?4,90& M0& 528;4?Z& ?4,82?0& M53&<br />
02?Q2?0& M0& ;2..& M0& 84& =450,@2?& 253& 9?43,=80& 0,=1& M0&<br />
0@M?8914520&M53&8M-.28&32Q/=20&814,>1&81402&@M?Z280&1MQ2&<br />
2@2?>23& 45.N& ?2=258.N/=M..N& @48/QM823& 81?4,>1& /554QM8/45& -,8& @MN& -2&<br />
0,994?823& -N& 08?M82>/=& 3/?2=8/45& 4?& 4812?& UM=84?0& 0,=1& M0&<br />
>4Q2?5@258&3/?2=823&/53,08?/M.&94./=N/20&34&<br />
548&/582?M=8&'/5&82?@0&4U&-2/5>&?29.M=2M-.2(&/5&812&=4,?02&4U&<br />
812/?&32Q2.49@25825=2&91M02&7&82=154.4>/20&<br />
2Q4.Q2& 84& =4@9282& 12M3*45& U4?& 812& 0M@2& =,084@2?0.4-M./]M8/45&9M81&=4@928/8/45&2Q4.Q20&81?4,>1&<br />
82=154.4>N& 92?U4?@M5=2& M53& .4=M.& @M?Z28& =1M?M=82?/08/=0&<br />
0,=1& M0& ?2>,.M84?N& ?2V,/?2@2580& 4U& 32Q2.49/5>& 2=454@/207&<br />
U4?2/>5& >4Q2?5@258& 0,994?8& M53& 0,-0/3/]M8/45& /=.,3/5>&<br />
948258/M.&=4??,98/45&0=12@20&/5&82532?/5>1&=M9M=/8N&E+J&<br />
0;/8=120&/5&#YYO&U4?&0M.2&/584&812&J1/5202&,?-M5&M53&?,?M.&<br />
9.M8U4?@0'I3V,/087&!""$(
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& !Y&<br />
&<br />
>"!"$G)3:A6+$F(=(/3@B('+$J$0,'A5,6+A)&'K$E,@,L&/&+D$,':$<br />
7&!""!(&M53&a/,7&!""$(2& 9M?80& 4U& 812/?&<br />
?2Q25,2&/5&Gg_&?2V,/?23&84&32Q2.49&1/>1&=M9M=/8N&0;/8=120252?M823& 4Q2?& 812&<br />
9?2Q/4,0& U4,?& N2M?0&84&32Q2.49&/80&RJ& 0;/8=10&84&32Q2.49&/80&RJ&0;/8=11& =M9M=/8N& 0;/8=120& ;M0& =450/32?23& M5& /@94?8M58&<br />
-?2MZ*81?4,>1& 84& 8?M50/8& 84& 812& =45Q2?>25=2& 91M02>2?&528;4?Z&M3QM58M>2&/0&./Z2.N&84&M88?M=8&@4?2&=,084@2?0&<br />
/5&812&M-025=2&4U&/554QM8/451&9?4-M-/./8N7&8125&MU82?&<br />
?2M./]M8/45& 81/0& U/?@& ;/..& >28& M12M3& 4U& /80& ?/QM.0& M53& >M/5&<br />
@M?Z28&01M?2&45&/80&-?M53&/5329253258&4U&9?/=/5>
2.6 CHINESE INNOVATION AND COMPETITION<br />
HANS W. GOTTINGER<br />
$"&<br />
RM50&S2?)&J1/5202&T554QM8/45&M53&J4@928/8/45)H5&T53,08?N&JM02&4U&812&L.4-M.& &<br />
F2.2=4@@,5/=M8/450&IV,/9@258&T53,08?N&<br />
.2M32?01/9&940/8/45& &<br />
812N&1MQ2&548&45.N&84&=M8=1&,9&-,8&M.04&84&.2M9U?4>&812/?&<br />
?/QM.0N&U?458/2?&/8&/0&M&@,=1&@4?2&3/UU/=,.8&81/5>&84&342M?7&<br />
J/0=4& EN082@07& =450/32?0& R,M;2/& M0& /80& 84,>1208& ?/QM.&<br />
'SE67!"#!3(/=& 940/8/45/5>& 4U& 812&<br />
/53/>2524,0& J1/5202& U/?@01& 812/?& =4@928/8/Q2& 91M020&<br />
U4..4;23& 812& ?,.20& 4U& L4& 4?& S2/[=1/7& M& 82??/84?/M.& >M@2&<br />
320/>523& 84& 0,-3,2& N4,?& ?/QM.& -N& 09M8/M..N& 25=/?=.2& M53&<br />
=4?52?&1/@&M53&=M98,?2&1/0&940/8/45&-N&34@/5M8/5>&1/@&4Q2?&<br />
812&-4M?3M?37&812&J1/5202&@M?Z28&.44Z0&./Z2&M&L4&<br />
-4M?3& ;1/=1& ;125& 2\9M5323& 84& 2@2?>/5>& M53& 8?M50/8/45M.&<br />
2=454@/20& 258M/.0& 812& >.4-M.& @M?Z28M@2&/0&548&4Q2?&N28N&/0&=4@92../5>&-,8&52230&<br />
U,?812?&?202M?=1&812&849&#""&J1/5202&2.2=8?45/=0:TF&=4@9M5/20/5>&<br />
M53&8?M50/8/45M.&2=454@/20&-,8&M.04&02Q2?M.&dIJ_&=4,58?/202&=M9M=/8N&E+J&2\=1M5>2&M53&M==200&<br />
528;4?Z&2V,/9@258&/5&#YY$&M53&#YYC7&?2092=8/Q2.N/=&9?/4?/8/20)&'#(&_2Q2.49&U?4@&-2/5>&M&<br />
0/5>.2& 0;/8=1/5>& 9?43,=8& =4@9M5N& 84& M& @,.8/*9?43,=8&<br />
=4@9M5N7&'!(&M33?200&812&52230&4U&812&,?-M5&@M?Z28&M0&;2..&<br />
M0& 81402& 4U& 812& ?,?M.& @M?Z28& M53& '$(& M33?200& 812& 52230& 4U&<br />
/582?5M8/45M.& &<br />
KM?Z280&M53&M==200&9?43,=80M>23& /5& ?202M?=1& 9M?852?01/90& 'G6W0(&<br />
;/81&.M?>27&;2..*208M-./0123&U/?@0&0,=1&M0&F2\M0&T508?,@25807&<br />
K484?4.M&M53&H>2?2&EN082@0&'`FI[0&M55,M.&?294?8&4U&!""$(N[7&X2;&h4?Z)&<br />
d\U4?3&e5/Q*^,0/5200& S22Z'!"#!(7[`FI& M/@0& 84& 9M00& ?/QM.&<br />
R,M;2/&/5&0@M?891452&0M.20&/5&J1/5M[7&6M57_
HANS W. GOTTINGER<br />
2.6 CHINESE INNOVATION AND COMPETITION<br />
& T582?5M8/45M.&^,0/5200&M53&I=454@/=0&G202M?=1&!"#$%&!'!()&!!*$!& $#&<br />
&<br />
^2/A/5>)J4582@94?M?N&J1/5M&+?200& &<br />
sBt! J1/5M& F49& #""& I.2=8?45/=0& J4@9M5N& a/08'!""$(7&<br />
;;;M?& &<br />
s##t! _2S40Z/576/8M.& 0;/8=1/5>& 08M8,0& M53&<br />
U4?2=M08[7&;;;N& M53&<br />
RN92?=4@928/8/457&a45345)&G4,8.23>2& &<br />
s#Ot! L488/5>2?7R2)&KTF&+?200& &<br />
s!qt! +4?82?7K1& F2=154.4>N&<br />
J4@928/8/457&JM@-?/3>2)&RM?QM?3&e5/Q?200&<br />
=4,.3&0V,22]2&-/>&9.MN2?0[7&KM?=1&q& &<br />
s$Ct! F12& SM..& E8?228& 64,?5M.& '!"#!-(7[R,M;2/& M/@0& U4?& 1/>12?&<br />
9?4U/.2&/5&0@M?8914520[7&KM?=1&D& &<br />
s$Ot! F12& SM..& E8?228& 64,?5M.& '!"#!=(7[R,M;2/& M/@0& 84& -2=4@2&<br />
@4?2&@M8,?2[7&H0/M&F2=154.4>N7KM?=1&#D& &<br />
s$Bt! F12& SM..& E8?228& 64,?5M.'!"#!3(7[J/0=4& JId& 0220& J1/5M[0&<br />
R,M;2/&M0&/80&84,>1208&?/QM.[7&H9?/.&B& &<br />
s$qt! F12& SM..& E8?228& 64,?5M.& '!"#!2(7[J1/5M& SM?50& R,M;2/&<br />
G294?8&=4,.3&1M?@&e
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
!"#$%&$%'()("*(%+,",-(.("#/%012$%3/%41$%3/%3333% 1<br />
!"#$%&'&'(%)$"*'&'"+'+,)&-."/,-)*&.0&%,'1)022'0+1%)<br />
3".#0&'"+4).%('%5)0+6)*7+&-%*'*)<br />
!"#$%&'(#'()*+,,-#.'()<br />
Institute of Ec<strong>on</strong>omic Analysis,<br />
University of Osaka (KGU),<br />
Osaka, Japan<br />
E-mail: hansgottinger@aol.com<br />
!"#$%&'$( This paper explores competitive positi<strong>on</strong>ing through network<br />
competiti<strong>on</strong> <strong>on</strong> the basis of alliance formati<strong>on</strong> (strategic alliances, joint<br />
ventures). From a strategic perspective, technological competiti<strong>on</strong> will be<br />
refined <strong>and</strong> exp<strong>and</strong>ed into new markets, or new markets will be created through<br />
alliance formati<strong>on</strong>. Alliance formati<strong>on</strong> could speed up competitive positi<strong>on</strong>ing<br />
<strong>and</strong> technological leadership in strategically important, though geographically<br />
diverse, markets. Revenue management tools can help in linking alliance<br />
formati<strong>on</strong> to better competitive outcomes, thereby improving strategic<br />
directi<strong>on</strong>s.<br />
)*+,-%.#( competiti<strong>on</strong>; corporate governance; managerial ec<strong>on</strong>omics;<br />
networks; revenue management; strategic alliance.<br />
/*0*%*1'* to this paper should be made as follows: <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H-W. (xxxx)<br />
‘Competitive positi<strong>on</strong>ing through strategic alliance formati<strong>on</strong>: review <strong>and</strong><br />
synthesis’, !"#$%&$%'()("*(%+,",-(.("#, Vol. x, No. x, pp.xx–xx.<br />
23"43-5%&673'8 1-$*#( Dr. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> is a Professor of Ec<strong>on</strong>omics at the<br />
University of Osaka (KGU), <strong>and</strong> the Director of the Institute of Management<br />
Science, Maastricht, NL. He has internati<strong>on</strong>ally taught <strong>and</strong> widely published in<br />
the areas of industrial, energy <strong>and</strong> envir<strong>on</strong>mental ec<strong>on</strong>omics. In these areas he<br />
has also advised internati<strong>on</strong>al organisati<strong>on</strong>s. His most recent forthcoming book<br />
titles, !""1),#51"/%6(78"121-9%,":%;9#=*7#*=(, illuminated the impact of the rise of the<br />
railway <strong>and</strong> related industries during the 19th century <strong>on</strong> the development of the modem<br />
capitalist corporati<strong>on</strong>. During the Industrial Revoluti<strong>on</strong>, fixed assets produced value<br />
through the creati<strong>on</strong> <strong>and</strong> distributi<strong>on</strong> of hard goods, often across l<strong>on</strong>g distances. The wide<br />
geographic, temporal <strong>and</strong> financial requirements of managing <strong>and</strong> operating a railway<br />
compelled the inventi<strong>on</strong> <strong>and</strong> evoluti<strong>on</strong> of substantially more sophisticated <strong>and</strong> structured<br />
corporate organisati<strong>on</strong>s. Earlier organisati<strong>on</strong>al forms <strong>and</strong> business practices were<br />
incapable of effectively managing in this new envir<strong>on</strong>ment. As such, the railway<br />
compelled the evoluti<strong>on</strong> of the modern global corporati<strong>on</strong>. Later, the breadth <strong>and</strong> depth of<br />
modern corporati<strong>on</strong>s encouraged the evoluti<strong>on</strong> of ever more complex functi<strong>on</strong>al<br />
Copyright © 200x Inderscience Enterprises Ltd.<br />
199
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
2% ;?@$%A1##5"-(=%<br />
hierarchies to coordinate disparate resources within expansive, multi-divisi<strong>on</strong>al firms<br />
(Ch<strong>and</strong>ler, 1962, 1990).<br />
We witness a similar phenomen<strong>on</strong> as we enter the 21st century. In c<strong>on</strong>trast to the<br />
Industrial Age, the emerging ec<strong>on</strong>omy primarily generates value through the creati<strong>on</strong>,<br />
disseminati<strong>on</strong> <strong>and</strong> applicati<strong>on</strong> of knowledge. Since the 1980s, networking technologies<br />
have created a dynamic similar to that of the railway in the 19th century, influencing the<br />
opti<strong>on</strong>s for corporate structures, relati<strong>on</strong>ships <strong>and</strong> competiti<strong>on</strong>. The most important <strong>and</strong><br />
far-reaching impacts occur as a result of how people <strong>and</strong> firms use these tools to create<br />
value. The c<strong>on</strong>flict between proprietary ownership as a necessary means for profit <strong>and</strong> the<br />
social nature of knowledge comprises the fundamental dynamic compelling the<br />
transformati<strong>on</strong> of organisati<strong>on</strong>al forms during the present period.<br />
In partial resp<strong>on</strong>se to this challenge, building <strong>and</strong> maintaining corporate alliances has<br />
become an increasingly important capability for the pursuit of both operati<strong>on</strong>al<br />
efficiencies <strong>and</strong> competitive advantage. Alliances have been transforming competiti<strong>on</strong><br />
<strong>and</strong>, as such, corporate strategy. The primary objective of this paper is to develop <strong>and</strong><br />
apply a new c<strong>on</strong>ceptual framework providing insight into the impact of alliances <strong>on</strong><br />
strategy (Secti<strong>on</strong> 5).<br />
Before we will deal with some pertinent issues of strategic envir<strong>on</strong>ments that<br />
facilitate or c<strong>on</strong>strain alliance formati<strong>on</strong>,<br />
Secti<strong>on</strong> 2 shows how managerial strategy drives competiti<strong>on</strong> but in turn is driven by<br />
competiti<strong>on</strong>. Secti<strong>on</strong> 3 shows how Transacti<strong>on</strong> Cost Ec<strong>on</strong>omics (TCE) <strong>on</strong> <strong>on</strong>e side limits<br />
the strategic manoeuvrability but, <strong>on</strong> the other h<strong>and</strong>, provides new opportunities to get<br />
ahead of rivals.<br />
The role of corporate governance in view of achieving superior competitive<br />
performance of the firm is discussed in Secti<strong>on</strong> 4. Secti<strong>on</strong> 5 builds the case for alliance<br />
formati<strong>on</strong> to strengthen competitive performance. Secti<strong>on</strong> 6 pinpoints the dimensi<strong>on</strong>s of<br />
network ec<strong>on</strong>omics in achieving leading market positi<strong>on</strong>s. Secti<strong>on</strong> 7 adds the time<br />
dimensi<strong>on</strong> as the facilitator of change management. Secti<strong>on</strong> 8 outlines the scope of<br />
applicati<strong>on</strong>s of network formati<strong>on</strong> to various high-tech industries.<br />
>&8-08#$%&$*5+(8'->>3$>*1$?8;1'*%$&31$+8&1.8'7&15*8<br />
C<strong>on</strong>structing <strong>and</strong> executing strategy, proactively or passively, requires predicti<strong>on</strong>s, yet<br />
the future always presents uncertainty. Herein lies the dilemma of the strategist –<br />
committing resources to an uncertain future. As market <strong>and</strong> technology change<br />
accelerates, uncertainty becomes of increasing c<strong>on</strong>cern. In a most basic challenge to<br />
strategic planning, in the view of Porter (1980), how can firms effectively positi<strong>on</strong><br />
themselves <strong>and</strong> their offerings if marketplace positi<strong>on</strong>s keep changing? Alternatively,<br />
executing <strong>on</strong> strategy in the real world requires firms to commit to directi<strong>on</strong>s they believe<br />
will provide sustainable profits. Commitment theory, as advanced by P. Ghemawat,<br />
argues that strategy requires resource investments that might be difficult or even<br />
impossible to recover from in the future, if too many strategic decisi<strong>on</strong>s turn out to have<br />
been wr<strong>on</strong>g (Ghemawat, 1991). Strategically phrased, you cannot arrive anywhere in<br />
particular if you do not commit to a directi<strong>on</strong> – <strong>and</strong> you cannot know in advance if you<br />
have chosen a good destinati<strong>on</strong>. Developing advantageous competency <strong>and</strong> resource<br />
combinati<strong>on</strong>s require substantial time <strong>and</strong> effort, so re-directi<strong>on</strong> can be costly.<br />
200
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
4% ;?@$%A1##5"-(=%<br />
substantially decreased the uncertainty regarding the marketplace availability of inputs or<br />
capabilities to accomplish ec<strong>on</strong>omic objectives. If <strong>on</strong>e needs something made or<br />
accomplished, <strong>on</strong>e can probably (out) source it.<br />
Certainly, the c<strong>on</strong>trol over resources provided by vertical integrati<strong>on</strong> still provides a<br />
compelling rati<strong>on</strong>ale for the formati<strong>on</strong> <strong>and</strong> boundaries of firms; however, the uncertainty<br />
of another sort increasingly dominates as we move to the future. Rapid technology <strong>and</strong><br />
marketplace change threatens firms by challenging the relevance of every product,<br />
service <strong>and</strong> business model. This represents the uncertainty regarding what to execute or,<br />
by extensi<strong>on</strong>, what, exactly, to integrate. Which competencies <strong>and</strong> other resources will<br />
provide a competitive advantage over the l<strong>on</strong>g term, most effectively integrated into the<br />
firm? While this issue presents the most significant challenge in dynamic markets, even<br />
mature industries can transform as a result of many factors, such as changing regulatory<br />
regimes, new entrants, technologies or business models, or a significant re-directi<strong>on</strong> by<br />
an established player. Moreover, smaller firms typically do not have the capability or<br />
resources necessary to achieve a broadly integrated firm, even if they have the<br />
philosophical wherewithal.<br />
The greater the uncertainty regarding the future, the more firms must c<strong>on</strong>struct a<br />
portfolio of strategic opti<strong>on</strong>s, without compromising the focus necessary to successfully<br />
compete. So effective strategy requires commitments towards an uncertain future, while<br />
marketplace change threatens to undermine or even destroy a firm’s value creating<br />
capabilities.<br />
<strong>Markets</strong> characterised by rapid technology change exhibit high levels of alliance<br />
formati<strong>on</strong> compared with mature industries, which tend to exhibit c<strong>on</strong>solidati<strong>on</strong> <strong>and</strong> even<br />
decline (Hagedoorn, 1993). The high frequency of alliance formati<strong>on</strong> in emerging<br />
industries will become evident when we look at the biotechnology <strong>and</strong> pharmaceutical<br />
industries, which evidenced a ninefold increase in operative alliances between 1993 <strong>and</strong><br />
2000 (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2004). More mature industries encountering re-organisati<strong>on</strong> or<br />
transformati<strong>on</strong> also experience increases in alliances, as occurred when the global<br />
chemical industry experienced severe overcapacity in the 1980s. Alliances <strong>and</strong> mergers<br />
allowed the industry to rati<strong>on</strong>alise producti<strong>on</strong> in a manner that competing firms could not<br />
have accomplished individually. These alliances represented an industry-wide alternative<br />
to c<strong>on</strong>solidati<strong>on</strong> through Mergers <strong>and</strong> Acquisiti<strong>on</strong>s (M&As), which occurred as well<br />
during this period.<br />
So marketplace actors apply network strategies in the periods of significant change<br />
<strong>and</strong> uncertainty. This seems to c<strong>on</strong>tradict the noti<strong>on</strong> that uncertainty engenders vertical<br />
integrati<strong>on</strong>. We are dealing with differing types of uncertainty, but the resp<strong>on</strong>ses are<br />
more similar than might appear. Vertical integrati<strong>on</strong> <strong>and</strong> alliances both integrate activities<br />
more closely than in the open market. To an extent, alliances represent integrati<strong>on</strong> by<br />
other means. N<strong>on</strong>etheless, alliances present significant strategic <strong>and</strong> managerial c<strong>on</strong>trasts<br />
to integrated firms, not the least of which is the opportunity for firms to interact with a<br />
broader, more diverse <strong>and</strong> potentially more productive universe of knowledge creating<br />
opportunities than is possible within a single firm. Let us first examine its precedents in<br />
the established organisati<strong>on</strong>al ec<strong>on</strong>omics <strong>and</strong> strategy literature.<br />
202
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.64*D In the early 1990s, Eastman Kodak Corporati<strong>on</strong> became a pi<strong>on</strong>eer in<br />
outsourcing Informati<strong>on</strong> <strong>Technology</strong> (IT) services. While many corporati<strong>on</strong>s outsourced<br />
various aspects of their IT operati<strong>on</strong>s, Kodak was <strong>on</strong>e of the first large, multinati<strong>on</strong>al<br />
corporati<strong>on</strong>s to outsource the majority of its corporate IT functi<strong>on</strong>. After a l<strong>on</strong>g period of<br />
planning, Kodak outsourced to three firms: Digital Equipment Corporati<strong>on</strong> (DEC),<br />
BusinessL<strong>and</strong> <strong>and</strong> IBM. Although a Harvard Business School case <strong>on</strong> the outsourcing<br />
strategy (Applegate <strong>and</strong> M<strong>on</strong>tealegre, 1995) reports overall success of the outsourcing<br />
venture, the process of building the integrati<strong>on</strong> <strong>and</strong> management processes between these<br />
four firms proved more challenging than Kodak’s CIO originally imagined. Kodak did<br />
achieve cost savings as a result; however, the overall cost savings were lower than<br />
anticipated. Kodak found that by outsourcing they saved the costs from shifting IT<br />
pers<strong>on</strong>nel <strong>and</strong> infrastructure to the firms whose core competencies were IT services.<br />
N<strong>on</strong>etheless, outsourcing introduced new costs involving developing staff <strong>and</strong> processes<br />
203
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
6% ;?@$%A1##5"-(=%<br />
to manage, or rather, coordinate the inter-firm efforts. C<strong>on</strong>trol costs were swapped for<br />
coordinati<strong>on</strong> costs.<br />
The TCE perspective biases the discussi<strong>on</strong> of this example towards the costs of<br />
achieving business results. Other motivati<strong>on</strong>s for outsourcing should include higher<br />
quality <strong>and</strong> improved corporate focus <strong>on</strong> the part of all firms in outsourcing relati<strong>on</strong>ships.<br />
Effective outsourcing should provide equal or higher-quality service than would be<br />
possible in-house, or at least a satisfactory level of service, at a similar or lower cost.<br />
Thus, outsourcing should represent the net value creati<strong>on</strong> between the c<strong>on</strong>tracting<br />
entities, allowing each to focus <strong>on</strong> its core competencies. While the TCE perspective<br />
approaches a successful outsourcing arrangement as a more efficient governance<br />
c<strong>on</strong>figurati<strong>on</strong>, focusing <strong>on</strong> costs can neglect or de-emphasise the potential benefits of<br />
greater value creati<strong>on</strong>.<br />
The transacti<strong>on</strong> cost TOF presents transacti<strong>on</strong> costs reducti<strong>on</strong> as the organising principle<br />
of firms; as such, TCE should prescribe any organisati<strong>on</strong>al form that provides the lowest<br />
transacti<strong>on</strong> costs for the ec<strong>on</strong>omic value created. It is partly a result of the efficiency <strong>and</strong><br />
effectiveness of the modem corporate form that a transacti<strong>on</strong> cost approach has provided<br />
theoretical justificati<strong>on</strong> for the existence of firms. It is also partly historical. Firms, in the<br />
traditi<strong>on</strong>al definiti<strong>on</strong>, exist, so they therefore must provide transacti<strong>on</strong> cost advantages<br />
over market governance, given TCE’s foundati<strong>on</strong> in bounded rati<strong>on</strong>ality (Sim<strong>on</strong>, 1947;<br />
Williams<strong>on</strong>, 1985) <strong>and</strong> the path-dependent nature of organisati<strong>on</strong>al <strong>and</strong> instituti<strong>on</strong>al<br />
development (Nels<strong>on</strong> <strong>and</strong> Winter, 1982; Hannan <strong>and</strong> Freeman, 1989). Moreover, the<br />
original c<strong>on</strong>text within which Coase developed his transacti<strong>on</strong> cost insights to<br />
underst<strong>and</strong>ing firm existence <strong>and</strong> boundaries likely biased the foundati<strong>on</strong> of the<br />
approach. Coase extrapolated from his experience working at the Ford Motor Company<br />
in the 1930s, recognising the company’s vertical integrati<strong>on</strong> strategy.<br />
He postulated that vertical integrati<strong>on</strong> was justified to decrease the transacti<strong>on</strong> costs<br />
that the firm would have otherwise incurred purchasing inputs in the open market. In the<br />
1930s, the industrial producti<strong>on</strong> model defined by Ford strove almost exclusively for<br />
operati<strong>on</strong>al efficiency. TCE is fundamentally an efficiency-based approach to<br />
underst<strong>and</strong>ing firms. Industries <strong>and</strong> strategies defined by operati<strong>on</strong>al efficiencies seem<br />
best suited to TCE. A number of researchers have asserted over the past two decades that<br />
inter-organisati<strong>on</strong>al relati<strong>on</strong>ships are often not well explained by TCE, particularly when<br />
efficiency issues do not represent the primary decisi<strong>on</strong> factors (Mariti <strong>and</strong> Smiley, 1983;<br />
Zajac <strong>and</strong> Olsen, 1993; Hamel <strong>and</strong> Prahalad, 1994). Decisi<strong>on</strong>s to pursue new firm<br />
competencies, or to form l<strong>on</strong>g-term R&D alliances are unlikely to be atomisable into<br />
discrete transacti<strong>on</strong>s (Argyres <strong>and</strong> Silverman, 2004).<br />
E8<br />
F-%6-%&$*85-G*%1&1'*8<br />
Alliances can be a critical strategic mechanism for the l<strong>on</strong>g-term success of ventures;<br />
however, alliances include a broad range of types, characterised by a number of factors<br />
such as integrati<strong>on</strong>, co-ownership, trust <strong>and</strong> l<strong>on</strong>gevity (Inkpen, 2001). Varied forms of<br />
alliance c<strong>on</strong>stellati<strong>on</strong>s also create varied c<strong>on</strong>straints for participants, from the l<strong>on</strong>g-term<br />
commitments into which they enter to other potential partner opportunities which<br />
participants forego to build <strong>and</strong> maintain membership. For instance, membership in<br />
certain strategic blocks can preclude involvement with competing blocks, the clearest<br />
204
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
8% ;?@$%A1##5"-(=%<br />
So when do network arrangements foster a competitive advantage, relative to firm or<br />
market organisati<strong>on</strong>? When are the assets or resources worth more as part of an alliance<br />
than within an individual firm or the open market? How do networks assist in the<br />
creati<strong>on</strong>, appropriati<strong>on</strong> <strong>and</strong> sustainability of value?<br />
H8<br />
=%3G31580-%'*#8"*731.8&443&1'*80-%>&$3-18<br />
After all, why do firms form alliances?<br />
The implicati<strong>on</strong>s of network strategies vary substantially, depending <strong>on</strong> the purposes<br />
for which a particular network of firms forms, as well as the purposes for which the<br />
network actually operates (which are not always identical). The creati<strong>on</strong> <strong>and</strong> management<br />
of alliance c<strong>on</strong>stellati<strong>on</strong>s must be understood in light of motivati<strong>on</strong>s for their creati<strong>on</strong>.<br />
While TCE presents a compelling approach to underst<strong>and</strong>ing firm boundaries<br />
(Milgrom <strong>and</strong> Roberts, 1992, Ch. 2), managers clearly engage in M&A, firm growth <strong>and</strong><br />
divestiture, re-organisati<strong>on</strong> <strong>and</strong> other boundary-shifting initiatives for motivati<strong>on</strong>s quite<br />
distinct from transacti<strong>on</strong> cost minimisati<strong>on</strong>. An extensive survey of decisi<strong>on</strong> makers<br />
involved in alliances in the early 1980s found that n<strong>on</strong>e of them cited decreasing<br />
transacti<strong>on</strong> costs as a primary rati<strong>on</strong>ale for their decisi<strong>on</strong>s (Mariti <strong>and</strong> Smiley, 1983).<br />
Marketplace competiti<strong>on</strong> encourages firms to minimise costs, but it also compels firms in<br />
other directi<strong>on</strong>s: to acquire firms to pre-empt or resp<strong>on</strong>d to competitors, to adjust<br />
strategic visi<strong>on</strong> in the face of disruptive technologies, to innovate to create value over the<br />
l<strong>on</strong>g term, to name a few. Innovati<strong>on</strong> by its nature requires investment in the creati<strong>on</strong> of<br />
new knowledge <strong>and</strong> capabilities, whether these capabilities are new to a particular firm or<br />
new to the marketplace. In the case of knowledge capital creati<strong>on</strong>, such as in R&D<br />
partnerships, ‘transacti<strong>on</strong>’ costs in the form of inter-firm coordinati<strong>on</strong> can in many cases<br />
be higher than an internally c<strong>on</strong>trolled effort; however, the overall value created by the<br />
combinati<strong>on</strong> of capabilities between firms can outweigh the increased costs of<br />
coordinati<strong>on</strong>. Zajac <strong>and</strong> Olsen’s noti<strong>on</strong> of ‘transacti<strong>on</strong>al value’ attempts to reflect the<br />
importance of value calculati<strong>on</strong>s in underst<strong>and</strong>ing inter-organisati<strong>on</strong>al arrangements<br />
(Zajac <strong>and</strong> Olsen, 1993).<br />
While value creati<strong>on</strong> <strong>and</strong> cost reducti<strong>on</strong> reflect the complementary noti<strong>on</strong>s for the<br />
factors that encourage firms to change their boundaries through M&A or interorganisati<strong>on</strong>al<br />
arrangements, these approaches do not assist much in elucidating the<br />
complexity of motivati<strong>on</strong>s influencing organisati<strong>on</strong>al decisi<strong>on</strong>s. In broad terms of the<br />
field of industrial organisati<strong>on</strong>, cost <strong>and</strong> value creati<strong>on</strong> provide appropriate<br />
generalisati<strong>on</strong>s. For the purposes of applied corporate strategy, cost <strong>and</strong> value paradigms<br />
are by themselves woefully limiting, particularly given an ever-uncertain future.<br />
Ec<strong>on</strong>omists attempt to reflect the behaviour of market actors based <strong>on</strong> preferences,<br />
available data <strong>and</strong> other decisi<strong>on</strong> factors, as well as provide prescripti<strong>on</strong>s for the most<br />
effective ec<strong>on</strong>omic acti<strong>on</strong>s as a result of the insights. The study of corporate strategy<br />
attempts to decipher why firms organise <strong>and</strong> act the way they do, how perhaps they<br />
should act regarding the development <strong>and</strong> executi<strong>on</strong> of effective strategies (but often do<br />
not) <strong>and</strong> the outcomes associated with various scenarios (Grant, 2001). The fact that<br />
much of the strategy literature seeks grounding in ec<strong>on</strong>omics reflects their<br />
complementary nature. Implicit in both of these disciplines is the noti<strong>on</strong> of motivati<strong>on</strong><br />
(Milgrom <strong>and</strong> Roberts, 1992, Ch. 16). Ec<strong>on</strong>omic behaviour, while observable in the<br />
marketplace, can <strong>on</strong>ly be modelled in the classical sense of optimising functi<strong>on</strong>s (even<br />
206
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.&$3-18<br />
What are the key building essentials the firms might need for network formati<strong>on</strong>? There<br />
are potentially limitless reas<strong>on</strong>s, but all rati<strong>on</strong>al, optimising motivati<strong>on</strong>s for forming<br />
alliances from a firm’s perspective’ can be categorised into three types, aside from purely<br />
financial motivati<strong>on</strong>s.<br />
1 4(#F1=G%(71"1.57C. A firm is attempting to compete in some manner under the<br />
c<strong>on</strong>diti<strong>on</strong>s influenced by network ec<strong>on</strong>omies. Network ec<strong>on</strong>omies refer to the<br />
situati<strong>on</strong>s in which network ec<strong>on</strong>omic phenomena significantly influence the<br />
dynamics of a particular industry or market (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2003).<br />
2 B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
10% ;?@$%A1##5"-(=%<br />
envir<strong>on</strong>mental, instituti<strong>on</strong>al <strong>and</strong> competitive factors present in a particular industry,<br />
market <strong>and</strong>/or ec<strong>on</strong>omy. The following list presents a summary of the examples<br />
collected:<br />
Factors favouring alliances<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
Knowledge sharing <strong>and</strong> creati<strong>on</strong><br />
Distributi<strong>on</strong><br />
Risk sharing<br />
R&D outsourcing<br />
Network effects<br />
Leveraging complementary competencies<br />
Manufacturing outsourcing<br />
Increasing focus <strong>on</strong> specific competencies<br />
Valuati<strong>on</strong><br />
News<br />
Entering a new product line or line of business<br />
Entering a new market, geographic or offering br<strong>and</strong> leverage<br />
Capital access investment<br />
Pre-emptive move<br />
Market spanning<br />
While such a list might never be complete, most factors for alliances can be dominantly<br />
<strong>and</strong> usefully categorised into <strong>on</strong>e, two or three of the types presented. The tax<strong>on</strong>omy need<br />
not be exhaustive to be effective. The point is that these three categories – Network<br />
Ec<strong>on</strong>omics, Competencies <strong>and</strong> Market Structure (NCM) – provide the dimensi<strong>on</strong>s that<br />
most powerfully present implicati<strong>on</strong>s for the creati<strong>on</strong> <strong>and</strong> prosecuti<strong>on</strong> of network<br />
strategy. Competencies refer to the internal decisi<strong>on</strong>s a firm makes regarding the<br />
competencies it develops in-house, which competencies it chooses to source through<br />
partners, <strong>and</strong> which competencies it might leave for the market to provide. Market<br />
structure refers to the envir<strong>on</strong>mental, instituti<strong>on</strong>al <strong>and</strong> competitive factors of a particular<br />
industry, market <strong>and</strong>/or ec<strong>on</strong>omy. Network ec<strong>on</strong>omics refers to whether network<br />
ec<strong>on</strong>omic phenomena significantly influence the dynamics of a particular industry or<br />
market under c<strong>on</strong>siderati<strong>on</strong>. Network ec<strong>on</strong>omics includes all of the issues traditi<strong>on</strong>ally<br />
associated with this field of ec<strong>on</strong>omics, such as dem<strong>and</strong>-side ec<strong>on</strong>omies of scale, network<br />
effects <strong>and</strong> positive <strong>and</strong> negative feedback loops (Shapiro <strong>and</strong> Varian, 1999; Varian,<br />
Farrell <strong>and</strong> Shapiro, 2004). In such envir<strong>on</strong>ments where fixed costs are high, variable<br />
costs low <strong>and</strong> decreasing <strong>on</strong> scale, dem<strong>and</strong> is volatile <strong>and</strong> competiti<strong>on</strong> is keen, profitable<br />
growth can be achieved <strong>on</strong>ly with sophisticated price discriminating strategies designed<br />
to help you provide the right product, at the right time, for the right customer, at the right<br />
price, that is <strong>on</strong> ‘mass customisati<strong>on</strong>’, as a major focus of revenue management.<br />
208
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
12% ;?@$%A1##5"-(=%<br />
the firm’s effective m<strong>on</strong>opoly. Any government m<strong>and</strong>ated m<strong>on</strong>opoly falls into this<br />
category as well, but can also overlap with other categories, as in the case of nati<strong>on</strong>al<br />
telecommunicati<strong>on</strong>s m<strong>on</strong>opolies (network ec<strong>on</strong>omics) or nati<strong>on</strong>al c<strong>on</strong>trol of raw<br />
materials mining <strong>and</strong> export. In the case of an oligopoly, existing players might, officially<br />
or unofficially, ally to maintain the status quo, as in the case of a cartel. The opposite can<br />
also be true. In highly disaggregated markets, firms might decide to ally in order to create<br />
ec<strong>on</strong>omies of scale, lowering costs <strong>and</strong> developing bargaining power against suppliers.<br />
Ec<strong>on</strong>omists use metrics such as the Herfindahl Index or a top four-firm c<strong>on</strong>centrati<strong>on</strong><br />
ratio to represent market c<strong>on</strong>centrati<strong>on</strong>.<br />
A literature exists <strong>on</strong> the implicati<strong>on</strong>s of the level of market c<strong>on</strong>centrati<strong>on</strong> in the<br />
industrial organisati<strong>on</strong> literature. The purpose here is not to develop <strong>and</strong> test many<br />
possible hypotheses related to market c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> alliance formati<strong>on</strong>, but rather to<br />
underscore that market structure plays a role in motivating the creati<strong>on</strong> of networks of<br />
firms, which holds the implicati<strong>on</strong>s for network strategy.<br />
Market structure dimensi<strong>on</strong>s are not limited to m<strong>on</strong>opolistic or oligopolistic<br />
c<strong>on</strong>diti<strong>on</strong>s. As evident from the examples presented earlier, the ‘market structure’<br />
category includes instituti<strong>on</strong>al influences such as government regulati<strong>on</strong> <strong>and</strong> anti-trust<br />
issues that might motivate inter-organisati<strong>on</strong>al arrangements over acquisiti<strong>on</strong> or market<br />
relati<strong>on</strong>ships. Here we define ‘instituti<strong>on</strong>al’ in the terms developed by North in the<br />
ec<strong>on</strong>omic field of instituti<strong>on</strong>al analysis (North, 1990). These instituti<strong>on</strong>al issues relate to<br />
the market structure category of the tax<strong>on</strong>omy, given the fact that they influence the<br />
organisati<strong>on</strong>al decisi<strong>on</strong>s between firms (i.e. to integrate or ally) by impacting competitive<br />
market c<strong>on</strong>diti<strong>on</strong>s, providing rules of the game. In this sense, it is not necessary to<br />
separate market structure <strong>and</strong> instituti<strong>on</strong>al issues into separate categories. The insights<br />
turn out <strong>on</strong> close inspecti<strong>on</strong> to be similar.<br />
H$M% J=5)(=C%D1=%,225,"7(CK%71.
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.6-%&48.3>*1#3-1(8%*6%*#*1$3158'7&15*8-G*%8$3>*8<br />
Firm c<strong>on</strong>stellati<strong>on</strong>s change over time. Sometimes, this occurs as a natural process of<br />
growth, or a deepening of c<strong>on</strong>tact between firms. Other times, participant firm objectives<br />
diverge. By definiti<strong>on</strong>, the NCM tax<strong>on</strong>omy underscores the influences that motivate<br />
firms to form alliances. As those influences change <strong>on</strong> a given c<strong>on</strong>stellati<strong>on</strong>, so too will<br />
their nature <strong>and</strong> c<strong>on</strong>structi<strong>on</strong>. The ec<strong>on</strong>omics literature does address firm life cycles when<br />
211
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
14% ;?@$%A1##5"-(=%<br />
c<strong>on</strong>sidering l<strong>on</strong>g-term decisi<strong>on</strong>s. In these cases, ec<strong>on</strong>omists c<strong>on</strong>sider ‘entry <strong>and</strong> exit’" of<br />
firms over the course of the term in questi<strong>on</strong>.<br />
N<strong>on</strong>etheless, the assumpti<strong>on</strong>s of firm perpetuity, both by researchers <strong>and</strong><br />
practiti<strong>on</strong>ers, can be misleading. Rarely do firms last more than a few generati<strong>on</strong>s. Those<br />
that do, usually undergo significant change over time. While firms like Ford Motor<br />
Company c<strong>on</strong>tinue to create value based <strong>on</strong> the original purpose for the company<br />
(automobile producti<strong>on</strong>), other venerable firms compete in completely different markets<br />
to their predecessors. Nokia’s current dominance of the wireless h<strong>and</strong>set <strong>and</strong><br />
telecommunicati<strong>on</strong>s industries bears very little relati<strong>on</strong>ship with its past, which has<br />
included l<strong>on</strong>g periods in forest products, paper <strong>and</strong> pulp, <strong>and</strong> as a c<strong>on</strong>glomerate with<br />
many unrelated lines of business. Can the Nokia of 1865 , the year of its founding, be<br />
c<strong>on</strong>sidered the same firm as today? So, despite any legal c<strong>on</strong>tinuity (<strong>and</strong> a company’s<br />
legal status can change), firms change markedly over time. Given that firms are by no<br />
means perpetual, researchers should treat them as entities finite in time <strong>and</strong> seek to<br />
underst<strong>and</strong> firm lifespans <strong>and</strong> growth cycles (Jovanovic, 1982; Greiner, 1972).<br />
Similarly, hybrid forms should be treated as arrangements in flux, developed in<br />
resp<strong>on</strong>se to, <strong>and</strong> as a result of, participant firms’ motivati<strong>on</strong>s <strong>and</strong> changing marketplace<br />
c<strong>on</strong>diti<strong>on</strong>s. The tax<strong>on</strong>omy provides varied insights into the life cycles of firm<br />
c<strong>on</strong>stellati<strong>on</strong>s. C<strong>on</strong>sortia created to motivate st<strong>and</strong>ards development in an emerging<br />
industry, at the interface of network ec<strong>on</strong>omics <strong>and</strong> competencies dimensi<strong>on</strong>s, would be<br />
likely to migrate further to the network ec<strong>on</strong>omics after the st<strong>and</strong>ard has been developed<br />
<strong>and</strong> disseminated. Once the st<strong>and</strong>ards race has been w<strong>on</strong> (not necessarily by the<br />
c<strong>on</strong>sortia), the motivati<strong>on</strong> for the alliance recedes, <strong>and</strong> the c<strong>on</strong>sortia is likely to be<br />
disb<strong>and</strong>ed unless alternative benefits have been developed between member firms. By<br />
c<strong>on</strong>trast, alliances in the competencies dimensi<strong>on</strong> often take <strong>on</strong>e of the two predominant<br />
directi<strong>on</strong>s, assuming the competencies that motivated the alliances c<strong>on</strong>tinue to be<br />
perceived as useful by the participants. Either firms in a competency-based alliance build<br />
integrated, mutually dependent competencies, which may encourage alliance l<strong>on</strong>gevity,<br />
or <strong>on</strong>e or more of the participant firms decides to disengage from the alliance, perhaps<br />
having changed strategic priorities or having acquired the competency in house. Often,<br />
firms form alliances to develop a competency in-house, with the assistance of a partner.<br />
Sometimes firms do so avowedly, while in other cases knowledge appropriati<strong>on</strong> occurs in<br />
a more surreptitious manner. Firms can participate in an alliance until they have acquired<br />
necessary knowledge from their partner, <strong>and</strong> then defect from cooperati<strong>on</strong>. Microsoft has<br />
become notorious for this, particularly with small firms. Andy Groves’ famous<br />
‘paranoia’” (Groves, 1996) has been part of the reas<strong>on</strong> Intel’s partnership with Microsoft<br />
has endured successfully for so l<strong>on</strong>g. Intel has maintained a healthy suspici<strong>on</strong> regarding<br />
its alliance with Microsoft. If <strong>on</strong>e or more of the alliance members’ strategic interest, in<br />
particular competencies changes, competency alliances are most likely to be affected.<br />
Market structure provides the most obvious example of how the factors in each<br />
motivati<strong>on</strong> space can impact alliance evoluti<strong>on</strong>. Government regulati<strong>on</strong>s can very quickly<br />
upset the competitive balances between c<strong>on</strong>stellati<strong>on</strong>s of firms, as has occurred in most<br />
instances of de-regulati<strong>on</strong> or privatisati<strong>on</strong>. The massive, global airline alliances (such as<br />
One World <strong>and</strong> Star Alliance), fit in the dimensi<strong>on</strong>s Network Ec<strong>on</strong>omics <strong>and</strong> Market<br />
Structure. Airline de-regulati<strong>on</strong> in the USA <strong>and</strong> elsewhere exerted a significant impact<br />
during the 1980s <strong>and</strong> 1990s, leading to a proliferati<strong>on</strong> of airlines. Moreover, network<br />
ec<strong>on</strong>omics heavily impacts the airline industry, characterised by communicating<br />
networks of routes, hub-<strong>and</strong>-spoke methods though anti-trust regulati<strong>on</strong>s c<strong>on</strong>tinue to exist<br />
212
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.*1#3-18$-8G&%3-;#831.;#$%3*#8<br />
The network formati<strong>on</strong> dimensi<strong>on</strong> can be applied to high-tech industries to assist in<br />
underst<strong>and</strong>ing the evoluti<strong>on</strong> of c<strong>on</strong>stellati<strong>on</strong>s of firms <strong>and</strong> network strategy under varied<br />
c<strong>on</strong>diti<strong>on</strong>s. As an example, the dimensi<strong>on</strong> applies to underst<strong>and</strong> the evoluti<strong>on</strong> of alliances<br />
within the biotechnology <strong>and</strong> pharmaceutical industries. The analysis will provide a proof<br />
of principle of the efficacy of the dimensi<strong>on</strong>, as well as an example of its applicati<strong>on</strong>.<br />
A cursory examinati<strong>on</strong> of industry network strategy might suggest that specific<br />
dimensi<strong>on</strong> types dominate in given industries. Most clearly, c<strong>on</strong>stellati<strong>on</strong>s of firms in<br />
industries dominated by network ec<strong>on</strong>omics, such as telecommunicati<strong>on</strong>s equipment or<br />
the airlines industries, would likely coalesce within the network ec<strong>on</strong>omics space.<br />
Industries dominated by the creati<strong>on</strong> of new knowledge, such as those substantially<br />
involved in commercialising emerging technologies (e.g. biotechnology, optical<br />
computing), would favour the competencies dimensi<strong>on</strong> space.<br />
While these distincti<strong>on</strong>s make intuitive sense, <strong>and</strong> further validate the applicati<strong>on</strong> of<br />
the tax<strong>on</strong>omy, any purely static applicati<strong>on</strong> of the tax<strong>on</strong>omy neglects a good deal of its<br />
explanatory power. Networks of firms, <strong>and</strong> network strategy in particular, must be<br />
understood dynamically. Strategy aims to change or maintain market relati<strong>on</strong>ships to the<br />
advantage of a particular firm or group of firms, relative to competitors (<strong>and</strong> in many<br />
cases relative to partners as well). Transforming market relati<strong>on</strong>ships for competitive<br />
advantage, as well as attempting to sustain an advantage, requires an underst<strong>and</strong>ing of<br />
how <strong>and</strong> why the relati<strong>on</strong>ships between firms change over time. While a particular<br />
category of the tax<strong>on</strong>omy might dominate an industry during a given period, examinati<strong>on</strong><br />
over time will undoubtedly uncover shifting influences. A network strategist must<br />
become adept at m<strong>on</strong>itoring industry developments critical to alliance formati<strong>on</strong>,<br />
recognising these changing c<strong>on</strong>diti<strong>on</strong>s early <strong>on</strong>, <strong>and</strong> quickly developing resp<strong>on</strong>ses to<br />
maintain as advantageous a positi<strong>on</strong> as possible. An effective network strategy should<br />
maintain a vigilant eye for events substantially impacting any of the dimensi<strong>on</strong> spaces<br />
within or between industries, or throughout a sector or ec<strong>on</strong>omy-wide. Such events can<br />
transform the relati<strong>on</strong>ships between firms, <strong>and</strong> can provide alert firms with the<br />
opportunity to drive transformati<strong>on</strong>s to their own advantage, or at least to successfully<br />
resp<strong>on</strong>d.<br />
213
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
16% ;?@$%A1##5"-(=%<br />
Both biotechnology <strong>and</strong> large pharmaceutical firms compete in an industry<br />
characterised by rapid technology change; in particular, these firms depend <strong>on</strong> the<br />
creati<strong>on</strong> of new knowledge. Alliance competencies should be prevalent in any market<br />
characterised by fast-changing, intangible assets, given the difficulties inherent in trading<br />
intangibles; moreover, in industries with very high rates of technology change,<br />
technologies can be introduced that create new market segments, obsolesce existing<br />
product lines <strong>and</strong> create substantial competitors from previously little-known firms.<br />
Under such c<strong>on</strong>diti<strong>on</strong>s, few firms can afford to c<strong>on</strong>duct research in enough directi<strong>on</strong>s to<br />
build sufficient R&D opti<strong>on</strong>s. Alliances offer opportunities for firms, in essence, to<br />
outsource R&D efforts, creating opti<strong>on</strong>s <strong>on</strong> knowledge development, without requiring<br />
merger or acquisiti<strong>on</strong>. Biotechnology provides a critical example of such intangible asses<br />
that can be difficult to trade, <strong>and</strong> are, as well, not necessarily appropriate c<strong>and</strong>idates for<br />
acquisiti<strong>on</strong>. Additi<strong>on</strong>ally, under c<strong>on</strong>diti<strong>on</strong>s of fast change <strong>and</strong> high uncertainty, the<br />
network forms of governance provide preferred access to informati<strong>on</strong>, decreasing<br />
informati<strong>on</strong> asymmetries <strong>and</strong> allowing firms involved in a network to scan a broader<br />
envir<strong>on</strong>ment.<br />
/*0*%*1'*#8<br />
Applegate, L.M. <strong>and</strong> M<strong>on</strong>tealegre, R. (1995) N,C#.,"%O1:,G%B1$K%+,",-5"-%!"D1=.,#51"%>9C#(.C%<br />
#8=1*-8%>#=,#(-57%P225,"7(C. Bost<strong>on</strong>: Harvard Business School Press.<br />
Argyres, N.S. <strong>and</strong> Silverman, B.S. (2004) ‘R&D, organizati<strong>on</strong>al structure <strong>and</strong> the development of<br />
corporate technological knowledge’, >#=,#(-57%+,",-(.("#%&1*=",2, Vol. 25, pp.929–958.<br />
Besanko, D., Dranove, D. <strong>and</strong> Shanley, M. (2000) N71"1.57C% 1D% >#=,#(-9 (2nd ed.). New York:<br />
Wiley.<br />
Ch<strong>and</strong>ler, A.D. (1962) >#=,#(-9%,":%>#=*7#*=(. Cambridge, MA: MIT Press.<br />
Ch<strong>and</strong>ler, A.D. (1990) >7,2(% ,":% >71#=,#(-9. New York: Free Press.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2003) N71"1.5(C%1D%4(#F1=G%!":*C#=5(C. L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H.W. (2004) ‘Biotech <strong>and</strong> pharma industries: applicati<strong>on</strong> of network ec<strong>on</strong>omies’,<br />
R8,=.,S(*#5C78(%!":*C#=5(, Vol. 66, pp.1475–1478.<br />
Grant, R.M. (2001) B1"#(.#=,#(-9%P",29C5C. Oxford: Blackwell.<br />
Greiner, L. (1972) ‘Evoluti<strong>on</strong> <strong>and</strong> revoluti<strong>on</strong> as organizati<strong>on</strong>s grow’, ;,=),=:% T*C5"(CC% '()5(F,<br />
Vol. 50, pp.37–46.<br />
Groves, A. (1996) U"29%#8(%R,=,"15:%>*=)5)(. New York: Nort<strong>on</strong>.<br />
Gulati, R. (1999) ‘Network locati<strong>on</strong> <strong>and</strong> learning. The influence of network resources <strong>and</strong> firm<br />
capabilities <strong>on</strong> alliance formati<strong>on</strong>’, >#=,#(-57%+,",-(.("#%&1*=",2, Vol. 20, pp.397–420.<br />
214
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
HANS W. GOTTINGER<br />
% B1.#=,#(-57%<br />
+,",-(.("#%&1*=",2, Vol. 12, pp.105–124.<br />
North, D.C. (1990) ‘!"C#5#*#51"C/% !"C#5#*#51",2% B8,"-(% ,":% N71"1.57% R(=D1=."7(. Cambridge:<br />
Cambridge Univ. Press.<br />
Oxley, J.E. <strong>and</strong> Samps<strong>on</strong>, R.C. (2004) ‘The scope of governance of internati<strong>on</strong>al R&D alliances’,<br />
>#=,#(-57%+,",-(.("#%&1*=",2, Vol. 25, pp.723–749.<br />
Porter, M.E. (1980) B1.#=,#(-9. New York: Free Press.<br />
Shapiro, C. <strong>and</strong> Varian, H.R. (1999) !"D1=.,#51"% '*2(C. Bost<strong>on</strong>, MA: Harvard Business School<br />
Press.<br />
Sim<strong>on</strong>, H. (1947) P:.5"C#=,#5)(%T(8,)51=. New York: Macmillan.<br />
Stinchcombe, A.L. (1990) !"D1=.,#51"%,":%U=-,"5S,#51"C$ Berkeley, CA: University of California<br />
Press.<br />
Varian, H., Farrell, J. <strong>and</strong> Shapiro, C. (2004) 68(%N71"1.57C%1D%!"D1=.,#51" 6(78"121-9. Raffaele<br />
Mattioli Lectures, Cambridge: Cambridge University Press.<br />
Williams<strong>on</strong>, O.E. (1975) +,=G(#C%,":%;5(=,785(C/%P",29C5C%1D%P"#5#=*C#%!.
HANS W. GOTTINGER<br />
2.7 COMPETITIVE POSITIONING THROUGH STRATEGIC ALLIANCES<br />
216
3.<br />
GROWTH<br />
AND<br />
DEVELOPMENT
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Supplementing: Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up . H.W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> M. Goosen,<br />
eds., NovaScience, New York, 2012<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, STRATEC Germany<br />
www.stratec-c<strong>on</strong>.net<br />
A Comprehensive Review of the Microec<strong>on</strong>omic Foundati<strong>on</strong>s of Ec<strong>on</strong>omic Growth<br />
This review provides a summary of the main findings of the literature review (which, as<br />
supporting document for the text “Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catchup”, is attached<br />
separately ). The secti<strong>on</strong> then provides a brief discussi<strong>on</strong> of key theories <strong>and</strong> c<strong>on</strong>cepts that we drew<br />
from the literature<br />
Summary Findings from the Literature<br />
Technological innovati<strong>on</strong> has an important role to play both as part of the competitive process <strong>and</strong><br />
as a driver of ec<strong>on</strong>omic growth <strong>and</strong> development. However, perhaps surprisingly, for a l<strong>on</strong>g time<br />
innovati<strong>on</strong> was c<strong>on</strong>sidered a field where ec<strong>on</strong>omists, as opposed to other scientists <strong>and</strong><br />
engineers, <strong>on</strong>ly viewed this as a ‘black box’.. The forces underlying the process of technological<br />
change were believed to be substantially independent of ec<strong>on</strong>omic incentives <strong>and</strong> mostly affected by<br />
the exogenous evoluti<strong>on</strong> of scientific knowledge <strong>and</strong> its technological applicati<strong>on</strong>s.<br />
It has been <strong>on</strong>ly recently, mainly over the last two decades, that the ec<strong>on</strong>omic forces behind<br />
technological innovati<strong>on</strong> have started to be investigated in more detail within mainstream<br />
ec<strong>on</strong>omics. As a result, a wide variety of theories <strong>and</strong> models, sometimes very diverse in spirit,<br />
describing the ec<strong>on</strong>omics of innovati<strong>on</strong> are now available. All these theories share the comm<strong>on</strong><br />
aim of providing a c<strong>on</strong>ceptual foundati<strong>on</strong> for underst<strong>and</strong>ing how innovati<strong>on</strong> affects the ec<strong>on</strong>omy,<br />
how ec<strong>on</strong>omic forces affect the emergence of technological changes, <strong>and</strong> the decisi<strong>on</strong>-making<br />
processes through which technological innovati<strong>on</strong> occurs.<br />
We have made no attempt to provide the most representative <strong>and</strong> comprehensive summaries of<br />
each str<strong>and</strong> of literature; this is not our goal. Rather it is to give a sufficient summary to enable<br />
subsequent identificati<strong>on</strong> of how the study can be taken forward to develop analytical tools for<br />
assessing market dynamics in competiti<strong>on</strong> policy investigati<strong>on</strong>s.<br />
This secti<strong>on</strong> provides a brief summary of some key elements of the different str<strong>and</strong>s of literature<br />
overviewed in much of this work.<br />
Industrial organizati<strong>on</strong><br />
Studies of dynamic competiti<strong>on</strong> within ec<strong>on</strong>omics have developed from <strong>on</strong>ce<br />
inherently static neoclassical ec<strong>on</strong>omic theories to modern analyses of innovati<strong>on</strong> <strong>and</strong> dynamic<br />
competiti<strong>on</strong>.<br />
N<strong>on</strong>etheless, a str<strong>on</strong>g relati<strong>on</strong>ship with neoclassical ec<strong>on</strong>omics remains in the general<br />
methodological approach adopted in these theories, which are characterized by the analysis of<br />
equilibrium models where fully rati<strong>on</strong>al ec<strong>on</strong>omic agents make optimizing choices.<br />
Generally, traditi<strong>on</strong>al models of innovati<strong>on</strong> focus <strong>on</strong> the study of firms’ incentives to invest<br />
resources in Research <strong>and</strong> <strong>Development</strong> (R&D) activities. Game-theoretical models developed<br />
in the industrial organizati<strong>on</strong> literature have investigated firms’ R&D decisi<strong>on</strong>s in strategic<br />
1<br />
219
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
envir<strong>on</strong>ments. (Endogenous growth models, discussed in the next sub-secti<strong>on</strong>, have developed<br />
the study of market dynamics in models that explain the relati<strong>on</strong>ship between firms’ investments,<br />
innovati<strong>on</strong> <strong>and</strong> ec<strong>on</strong>omic growth.)<br />
Game-theoretical models suggest that there are two main forces that underlie firms’ investment in<br />
R&D: the search for higher profits <strong>and</strong> the threat posed by falling behind potential innovating<br />
rivals. Game theoretical models study these forces in a variety of market situati<strong>on</strong>s <strong>and</strong> address<br />
issues such as the interplay between innovati<strong>on</strong> <strong>and</strong> market structure, the dynamics of<br />
competiti<strong>on</strong> <strong>and</strong> the nature of the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
These models provide a rich picture of what the possible strategies <strong>and</strong> industry equilibria in<br />
dynamic markets can be. However, general predicti<strong>on</strong>s, that can be c<strong>on</strong>sidered appropriate<br />
across all situati<strong>on</strong>s <strong>and</strong> industries, are scarce. On <strong>on</strong>e h<strong>and</strong>, this is a result that seems to stress<br />
the lack of predictive power of these models, i.e. (almost) everything can be rati<strong>on</strong>alized; <strong>on</strong> the<br />
other h<strong>and</strong>, however, the variety of results seems well to fit with the variety of observed<br />
behaviours <strong>and</strong> “equilibria”. There is no general model that can uncritically be applied to any<br />
case: the underst<strong>and</strong>ing of the specific characteristics of the single situati<strong>on</strong> needs to underpin<br />
any appropriate choice of a modelling framework.<br />
Despite the absence of general results, these models are certainly useful tools to underst<strong>and</strong><br />
firms’ incentives to invest in R&D activities in strategic envir<strong>on</strong>ments <strong>and</strong> to suggest what main<br />
factors may be central in shaping the nature of dynamic competiti<strong>on</strong>. For instance, these models<br />
suggest that:<br />
• In order to underst<strong>and</strong> R&D investments in strategic envir<strong>on</strong>ments it is necessary to<br />
underst<strong>and</strong> how innovati<strong>on</strong> may affect profits both of successful <strong>and</strong> n<strong>on</strong>-successful<br />
innovators. The first perspective captures the idea that firms want to innovate to increase<br />
their profits; the sec<strong>on</strong>d captures the idea that firms want to innovate to maintain<br />
competitiveness.<br />
•<br />
•<br />
The relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> of an industry <strong>and</strong> its rate of technological<br />
innovati<strong>on</strong> is certainly complex <strong>and</strong> in general not a causal <strong>on</strong>e: both should be thought of<br />
as the outcomes of the operati<strong>on</strong> of market forces <strong>and</strong> exogenous factors such as the<br />
nature of dem<strong>and</strong>, technological opportunity <strong>and</strong> the c<strong>on</strong>diti<strong>on</strong>s governing appropriability.<br />
Dynamic competiti<strong>on</strong> may be characterized by persistent dominance of the incumbent<br />
leader or by acti<strong>on</strong>-reacti<strong>on</strong> whereby incumbents are overtaken by a rival whose<br />
incumbency is itself then short-lived. The nature of market dynamics depends <strong>on</strong> a<br />
number of factors, such as the type of innovati<strong>on</strong>, i.e. drastic or n<strong>on</strong>-drastic, the<br />
uncertainties involved in R&D activities, the nature of patent protecti<strong>on</strong> <strong>and</strong> of knowledge<br />
spill-overs, the intensity of product market competiti<strong>on</strong>, etc.<br />
•<br />
When the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is investigated, it is necessary<br />
to be clear what the noti<strong>on</strong> of intensity of competiti<strong>on</strong> describes <strong>and</strong> how this relates (or<br />
does not relate) to market structure. Indeed a market where competiti<strong>on</strong> is tougher may<br />
be more c<strong>on</strong>centrated simply because inefficient firms cannot survive. There may be a<br />
2<br />
220
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
trade-off between the intensity of static competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. In general, the<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> need not be m<strong>on</strong>ot<strong>on</strong>ic at all.<br />
Endogenous growth models have recently developed al<strong>on</strong>g the earlier game theoretical literature <strong>on</strong><br />
innovati<strong>on</strong> in the c<strong>on</strong>text of studies that seek to explain the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
ec<strong>on</strong>omic growth. These models suggest that innovati<strong>on</strong>, resulting from intenti<strong>on</strong>al R&D<br />
investments by profit-maximizing firms or simply by unintenti<strong>on</strong>al learning-by-doing, is a<br />
fundamental driver of ec<strong>on</strong>omic growth in the l<strong>on</strong>g run.<br />
Early Schumpeterian endogenous growth models stressed the importance of ex-post rents for<br />
innovati<strong>on</strong>: competiti<strong>on</strong> would have a detrimental effect <strong>on</strong> innovati<strong>on</strong> by decreasing the rents that<br />
an innovator would be able to appropriate. More recent models have emphasized another<br />
mechanism by which competiti<strong>on</strong> affects innovati<strong>on</strong>: tougher competiti<strong>on</strong> may increase the<br />
incentives of firms to innovate in order to escape from fierce competiti<strong>on</strong>. These recent studies<br />
suggest that the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> may not well be m<strong>on</strong>ot<strong>on</strong>ic <strong>and</strong><br />
that instead, <strong>on</strong>e should expect an inverse-U shaped relati<strong>on</strong>ship: when competiti<strong>on</strong> is low, an<br />
increase in competiti<strong>on</strong> would foster innovati<strong>on</strong>; the reverse would happen when competiti<strong>on</strong> is<br />
fierce.<br />
The result that competiti<strong>on</strong> may be c<strong>on</strong>ducive to innovati<strong>on</strong> is also obtained in studies where the<br />
traditi<strong>on</strong>al behavioral assumpti<strong>on</strong> of profit-maximizing firms is relaxed. When principal-agent<br />
c<strong>on</strong>siderati<strong>on</strong>s are introduced to explain managers’ behaviors, another mechanism by which<br />
competiti<strong>on</strong> may favor innovati<strong>on</strong> is suggested: the speed of innovati<strong>on</strong> may be retarded by the<br />
slack of managers who tend to avoid private costs associated with innovati<strong>on</strong>. When competiti<strong>on</strong><br />
intensifies, the higher threat of bankruptcy may force managers to speed up the process at which<br />
new ideas are adopted. Hence, competiti<strong>on</strong> may be c<strong>on</strong>ducive to faster rates of innovati<strong>on</strong>.<br />
New ec<strong>on</strong>omic geography<br />
New ec<strong>on</strong>omic geography is a branch of ec<strong>on</strong>omics which is mainly c<strong>on</strong>cerned with spatial<br />
aspects of ec<strong>on</strong>omics. In particular, it seeks to explain why <strong>and</strong> how given ec<strong>on</strong>omic activities<br />
c<strong>on</strong>centrate geographically, either within individual countries (agglomerati<strong>on</strong>s) or between<br />
different countries (industrial clustering). Furthermore, it c<strong>on</strong>siders the inter-relati<strong>on</strong>ships between<br />
geographical c<strong>on</strong>centrati<strong>on</strong> of industry, internati<strong>on</strong>al trade <strong>and</strong> ec<strong>on</strong>omic development.<br />
At a broad level, several themes emerge from the literature.<br />
•<br />
•<br />
Geographical c<strong>on</strong>centrati<strong>on</strong> of ec<strong>on</strong>omic activity will occur where transport costs are low<br />
enough <strong>and</strong> the labor force is mobile.<br />
Key characteristics of a particular industry have also an important role to play. The higher<br />
the scope for vertical integrati<strong>on</strong>, the higher the share of intermediate goods in output <strong>and</strong><br />
the lower the labor intensity are, the higher will be the drive for c<strong>on</strong>centrati<strong>on</strong>.<br />
3<br />
221
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Industrial c<strong>on</strong>centrati<strong>on</strong> implies the possibility of geographical proximity of firms that lays<br />
the foundati<strong>on</strong> for technological cooperati<strong>on</strong> <strong>and</strong> thus lead to increased innovative<br />
activity.<br />
•<br />
A high degree of c<strong>on</strong>centrati<strong>on</strong> leads to very large markets — <strong>and</strong> the larger the size of<br />
the market, the higher the potential reward of successful innovati<strong>on</strong>s.<br />
The evoluti<strong>on</strong>ary ec<strong>on</strong>omics of innovati<strong>on</strong><br />
The evoluti<strong>on</strong>ary approach to the study of innovati<strong>on</strong> has been developed <strong>on</strong> very different<br />
methodological basis than those underlying traditi<strong>on</strong>al ec<strong>on</strong>omic models of innovati<strong>on</strong>. In<br />
particular, we observe the rejecti<strong>on</strong> of the modelling assumpti<strong>on</strong>s of rati<strong>on</strong>ality <strong>and</strong> equilibrium<br />
that are fundamental to the traditi<strong>on</strong>al approach.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics looks from the outset at dynamic processes. In particular, it is associated<br />
with the use of analogies from evoluti<strong>on</strong>ary biology to explain ec<strong>on</strong>omic growth <strong>and</strong> the process<br />
of competiti<strong>on</strong>. Thus the cornerst<strong>on</strong>es of an evoluti<strong>on</strong>ary analysis of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
are variety, selecti<strong>on</strong> <strong>and</strong> imitati<strong>on</strong>.<br />
At a basic level, using Darwinian analogies, we can begin to appreciate the role of the market in<br />
selecting the more fit firms (efficient <strong>and</strong> profitable), products <strong>and</strong> techniques at the expense of<br />
less fit firms, products or techniques. In additi<strong>on</strong> to this effect, we would expect to see imitati<strong>on</strong> of<br />
winning ideas by those whose survival is otherwise threatened (although this is limited by the tacit<br />
nature of knowledge).<br />
Inherent in this model of competiti<strong>on</strong> is the associati<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> experimentati<strong>on</strong><br />
<strong>and</strong> variety. A variety of experiments allows, through the process of selecti<strong>on</strong>, for greater<br />
ec<strong>on</strong>omic progress than would be available through uniform optimizati<strong>on</strong>.<br />
Thus ec<strong>on</strong>omic development <strong>and</strong> innovati<strong>on</strong> can be seen as a combined effect of selecti<strong>on</strong> (via<br />
competiti<strong>on</strong>) from a variety of competing routines <strong>and</strong> practices as well as the more endogenous<br />
process of agents seeking improved routines <strong>and</strong> practices. While the latter is certainly<br />
incorporated, if treated somewhat differently, under the mainstream neoclassical approach, the<br />
emphasis <strong>on</strong> selecti<strong>on</strong> from variety seems an important additi<strong>on</strong> to this approach.<br />
This suggests that models found in other branches of ec<strong>on</strong>omics might miss something important<br />
when they analyze dynamics with reference to homogenous profit-maximizing firms: namely the<br />
benefits of selecti<strong>on</strong> from heterogeneity in capabilities <strong>and</strong> innovative experimentati<strong>on</strong>.<br />
Systems of innovati<strong>on</strong><br />
At an aggregate level, the evoluti<strong>on</strong>ary approach to the study of ec<strong>on</strong>omic growth draws attenti<strong>on</strong><br />
to the importance of instituti<strong>on</strong>s in the process of ec<strong>on</strong>omic growth. The key findings of the<br />
literature are as follows:<br />
• Innovati<strong>on</strong> <strong>and</strong> its diffusi<strong>on</strong> take place within systems of interc<strong>on</strong>nected organizati<strong>on</strong>s <strong>and</strong><br />
instituti<strong>on</strong>s. Important c<strong>on</strong>stituent elements of such systems are organizati<strong>on</strong>s such as<br />
4<br />
222
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
firms, governments <strong>and</strong> universities. Instituti<strong>on</strong>s, with which these organizati<strong>on</strong>s interact,<br />
reflect laws <strong>and</strong> statutes (e.g. the instituti<strong>on</strong> of patent protecti<strong>on</strong>) as well as more abstract<br />
elements (e.g. cultural aspects of the ec<strong>on</strong>omy such as the spirit of entrepreneurial<br />
activity).<br />
•<br />
•<br />
Innovati<strong>on</strong> within the system will depend not <strong>on</strong>ly <strong>on</strong> single instituti<strong>on</strong>s but also <strong>on</strong> the<br />
nature <strong>and</strong> intensity of interacti<strong>on</strong>s between the various elements of the system.<br />
When c<strong>on</strong>sidering the interacti<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>, we must remain<br />
aware that the effects of such interacti<strong>on</strong> will depend <strong>on</strong> the evolving instituti<strong>on</strong>al<br />
background against which agents in the ec<strong>on</strong>omy operate.<br />
The overview of the cliometric literature has shown that cliometrics is a methodological framework<br />
for the study of ec<strong>on</strong>omic history. The existing body of cliometrics draws attenti<strong>on</strong> to the following<br />
factors relevant to “systems of innovati<strong>on</strong>”:<br />
• Ec<strong>on</strong>omic growth is a result of a complex interacti<strong>on</strong> between instituti<strong>on</strong>s <strong>and</strong> markets.<br />
Examples of instituti<strong>on</strong>s that are frequently important for innovati<strong>on</strong> are the protecti<strong>on</strong> of<br />
property rights (for example, against piracy) <strong>and</strong> patent policy.<br />
Innovati<strong>on</strong> plays a critical role in ec<strong>on</strong>omic progress <strong>and</strong> development. This involves not<br />
<strong>on</strong>ly technological innovati<strong>on</strong>s but also changes in instituti<strong>on</strong>s. Thus, instituti<strong>on</strong>s<br />
themselves evolve according to changing circumstances. These two last points<br />
complement the c<strong>on</strong>cepts of systems of innovati<strong>on</strong> found in the macroec<strong>on</strong>omic field of<br />
the evoluti<strong>on</strong>ary approach.<br />
5<br />
223
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
INTRODUCTION<br />
1<br />
1.1<br />
1.2<br />
1.3<br />
1.4<br />
2<br />
2.1<br />
2.2<br />
2.3<br />
3<br />
3.1<br />
3.2<br />
3.3<br />
4<br />
4.1<br />
4.2<br />
4.3<br />
4.4<br />
4.5<br />
4.6<br />
ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
Industrial Organizati<strong>on</strong><br />
Endogenous Growth Models<br />
New Ec<strong>on</strong>omic Geography<br />
THE EVOLUTIONARY ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
Microec<strong>on</strong>omics of Innovati<strong>on</strong><br />
Systems of Innovati<strong>on</strong> <strong>and</strong> the Role of Instituti<strong>on</strong>s<br />
STRATEGIES OF INNOVATIVE FIRMS<br />
Innovati<strong>on</strong> <strong>and</strong> Competitive Advantage<br />
Timing of Innovati<strong>on</strong> <strong>and</strong> Life Cycles<br />
Issues Relating to Network Industries<br />
BIBLIOGRAPHY<br />
Industrial Organizati<strong>on</strong><br />
Endogenous Growth<br />
New Ec<strong>on</strong>omic Geography<br />
Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics<br />
Cliometrics<br />
Strategic Management <strong>and</strong> Networks<br />
224
7<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
INTRODUCTION<br />
This is an overview, with references to the seminal literature <strong>on</strong> the spectrum of dynamic<br />
approaches in micro <strong>and</strong> meso ec<strong>on</strong>omics covering at least the last two decades – the<br />
overview should avoid a pure focus <strong>on</strong> mainstream approaches <strong>and</strong> draw benefits from<br />
overlaps with neighboring disciplines such as systems theory, cliometrics <strong>and</strong><br />
instituti<strong>on</strong>alism.<br />
This is a brief overview <strong>and</strong> classificati<strong>on</strong> based <strong>on</strong> the relevant literature of the most<br />
important ec<strong>on</strong>omic strategies applied by innovative companies to enter new markets or<br />
introduce new products.<br />
Secti<strong>on</strong> 1 presents the survey of mainstream theories <strong>and</strong> models of innovati<strong>on</strong> <strong>and</strong> Secti<strong>on</strong> 2<br />
discusses evoluti<strong>on</strong>ary ec<strong>on</strong>omics, systems theory, cliometrics <strong>and</strong> instituti<strong>on</strong>alism.<br />
Secti<strong>on</strong> 3 focuses instead <strong>on</strong> firms’ strategies related to innovative activities.<br />
225
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
1<br />
1.1<br />
ECONOMICS OF INNOVATION<br />
Introducti<strong>on</strong><br />
The neo-classical approach to innovati<strong>on</strong><br />
In neo-classical theories, producti<strong>on</strong> opportunities are described by specifying a producti<strong>on</strong><br />
functi<strong>on</strong>, which describes the technological relati<strong>on</strong>ship between inputs <strong>and</strong> output. Technological<br />
progress is captured by the ability of a firm, or the ec<strong>on</strong>omy, to produce, for each combinati<strong>on</strong> of<br />
inputs used, more output. However, such progress was for a l<strong>on</strong>g time essentially c<strong>on</strong>sidered<br />
exogenous, as if it were knowledge that falls, like manna, from heaven.<br />
Nevertheless, neo-classical ec<strong>on</strong>omists were c<strong>on</strong>scious that the c<strong>on</strong>siderati<strong>on</strong> of innovati<strong>on</strong> as an<br />
ec<strong>on</strong>omic phenomen<strong>on</strong> posed some challenges to established theories of perfect competiti<strong>on</strong><br />
<strong>and</strong> that a trade-off may exist between static <strong>and</strong> dynamic efficiency.<br />
Innovati<strong>on</strong> is essentially the producti<strong>on</strong> <strong>and</strong> commercial applicati<strong>on</strong> of new knowledge. Neoclassical<br />
ec<strong>on</strong>omists, following the pi<strong>on</strong>eering work of Arrow (1962), have usually treated<br />
knowledge as informati<strong>on</strong>, that is, as a pure n<strong>on</strong>-rivalrous <strong>and</strong> n<strong>on</strong>-excludable good. The<br />
c<strong>on</strong>cepts of n<strong>on</strong>-rivalry <strong>and</strong> n<strong>on</strong>-excludability are crucial in underst<strong>and</strong>ing the neo-classical<br />
ec<strong>on</strong>omic analysis of innovati<strong>on</strong> <strong>and</strong> can be explained as follows:<br />
(a) N<strong>on</strong>-rivalry implies that the same piece of knowledge can be c<strong>on</strong>sumed by different<br />
individuals, the marginal cost involved being essentially zero. Knowledge is<br />
c<strong>on</strong>sidered to be essentially disembodied (as opposed, for instance, to knowledge<br />
embodied in organizati<strong>on</strong>al routines — a central feature of the evoluti<strong>on</strong>ary approach<br />
overviewed in secti<strong>on</strong> 2) <strong>and</strong> available in a comm<strong>on</strong> stock or pool. In other words<br />
knowledge is essentially assimilated to informati<strong>on</strong>.<br />
(b) N<strong>on</strong>-excludability is the inability of the innovator to prevent others from accessing the<br />
knowledge produced. N<strong>on</strong>-excludability arises largely because it may not be feasible,<br />
or it is extremely costly, to m<strong>on</strong>itor <strong>and</strong> charge for the use <strong>and</strong> enjoyment of a piece of<br />
knowledge. In traditi<strong>on</strong>al models of innovati<strong>on</strong>, it is often related to the noti<strong>on</strong> of<br />
knowledge spillovers, which refers to the positive externality by which the knowledge<br />
generated by an agent enhances the technological capabilities of other agents or their<br />
ability to discover new knowledge. N<strong>on</strong>-excludability is not entirely an intrinsic quality<br />
of knowledge <strong>and</strong> depends largely <strong>on</strong> the legal regime in place <strong>and</strong>, in particular, <strong>on</strong><br />
the system of property rights (e.g. patents). Indeed, a patent might allow the innovator<br />
to appropriate the value of an (otherwise n<strong>on</strong>-excludable) innovati<strong>on</strong> by granting the<br />
exclusive right of its commercial use for a certain period of time.<br />
The neo-classical analysis of the ec<strong>on</strong>omic forces underlying the producti<strong>on</strong> of new technological<br />
knowledge follows from the general theoretical framework described above <strong>and</strong> is developed <strong>on</strong><br />
the assumpti<strong>on</strong> that ec<strong>on</strong>omic agents are rati<strong>on</strong>al <strong>and</strong> act to maximize their utility or profit.<br />
When n<strong>on</strong>-rivalry is coupled with n<strong>on</strong>-excludability, <strong>and</strong> both are c<strong>on</strong>sidered to be essential<br />
features of technological knowledge, innovati<strong>on</strong> acquires the characteristics of a pure public good<br />
<strong>and</strong>, accordingly, profit maximizing agents are c<strong>on</strong>sidered not to have incentives to invest in its<br />
producti<strong>on</strong>.<br />
8<br />
226
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In fact, if the process of imitati<strong>on</strong> of an innovati<strong>on</strong> were immediate, a firm would not have the<br />
incentive to invest resources in R&D activities since ex-post competiti<strong>on</strong> would not allow the<br />
innovator to earn profits to pay back its investment. However, although a single firm would not<br />
have incentives to discover new knowledge, innovati<strong>on</strong> may well be beneficial for society so that<br />
a role for public interventi<strong>on</strong> in the form of direct R&D investment <strong>and</strong>/or appropriate public policy<br />
is usually called for.<br />
These c<strong>on</strong>siderati<strong>on</strong>s suggest that a trade-off may exist between static <strong>and</strong> dynamic efficiency<br />
<strong>and</strong> provide the ec<strong>on</strong>omic rati<strong>on</strong>ale for some important public policies such as public funding of<br />
research <strong>and</strong> the definiti<strong>on</strong> of a patent system that protects the innovator from imitati<strong>on</strong>. The<br />
basic trade-off of a patent system is that between the static costs that derive from the innovator’s<br />
potential m<strong>on</strong>opoly power <strong>and</strong> the l<strong>on</strong>g-term benefits associated with the discovery of the new<br />
knowledge.<br />
Core models of dynamic competiti<strong>on</strong><br />
In the last two decades, the analysis of innovati<strong>on</strong> <strong>and</strong> technological change has attracted much<br />
attenti<strong>on</strong> within mainstream ec<strong>on</strong>omics, due to the development of two closely related disciplines<br />
— game-theoretical industrial organizati<strong>on</strong> <strong>and</strong> endogenous growth theories. As a c<strong>on</strong>sequence,<br />
many of the themes that were <strong>on</strong>ce a feature of the less c<strong>on</strong>venti<strong>on</strong>al theories of technological<br />
advance have now found their way into ec<strong>on</strong>omics.<br />
Studies of dynamic competiti<strong>on</strong> within mainstream ec<strong>on</strong>omics have developed <strong>on</strong>ce inherently<br />
static neoclassical ec<strong>on</strong>omic theories into modern analyses of innovati<strong>on</strong> <strong>and</strong> dynamic<br />
competiti<strong>on</strong>. N<strong>on</strong>etheless, a str<strong>on</strong>g relati<strong>on</strong>ship with neoclassical ec<strong>on</strong>omics remains in the<br />
general methodological approach adopted in these theories, which are characterized by the<br />
analysis of equilibrium models where fully rati<strong>on</strong>al ec<strong>on</strong>omic agents make optimizing choices.<br />
Core models of dynamic competiti<strong>on</strong> are highly stylized <strong>and</strong> this is also reflected in the<br />
definiti<strong>on</strong> of innovati<strong>on</strong> adopted. In these models, innovati<strong>on</strong> is usually defined as the reducti<strong>on</strong> of<br />
producti<strong>on</strong> costs or as the commercializati<strong>on</strong> of new <strong>and</strong>/or better products. Apparently this<br />
definiti<strong>on</strong> identifies innovati<strong>on</strong> with technological change <strong>and</strong> downsizes the importance of other<br />
forms of innovati<strong>on</strong>, such as managerial <strong>and</strong> organizati<strong>on</strong>al changes.<br />
However, notwithst<strong>and</strong>ing the emphasis placed by these models <strong>on</strong> innovati<strong>on</strong> as the commercial<br />
use of new or improved products <strong>and</strong> processes, this stylized definiti<strong>on</strong> of innovati<strong>on</strong> can also be<br />
c<strong>on</strong>sidered to encompass n<strong>on</strong>-technological changes inasmuch these ultimately matter because<br />
result in a reducti<strong>on</strong> of producti<strong>on</strong> costs or new <strong>and</strong> better products.<br />
What mainstream models suggest is that the particular form of innovati<strong>on</strong> (e.g. new technological<br />
knowledge rather than new organizati<strong>on</strong>al knowledge) is not central to the underst<strong>and</strong>ing of firms’<br />
incentives to exert effort to improve their ability to produce efficiently <strong>and</strong> to market new <strong>and</strong> or<br />
better products, which is the main questi<strong>on</strong> addressed in these theories.<br />
9<br />
227
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Similar to the definiti<strong>on</strong> of innovati<strong>on</strong> , also the treatment of the process by which new knowledge is<br />
generated is very stylized in these models. Knowledge is usually c<strong>on</strong>sidered as deriving from<br />
learning-by-doing or from intenti<strong>on</strong>al R&D investments, optimally chosen by profit-maximizing<br />
agents. The R&D process is captured by the definiti<strong>on</strong> of a deterministic or stochastic knowledge<br />
producti<strong>on</strong> functi<strong>on</strong> which related the input used in R&D — notably capital, labor, available<br />
knowledge – to the new knowledge produced <strong>and</strong>/or the timing of innovati<strong>on</strong>.<br />
Game-theoretical models study firms’ incentives to invest resources in R&D activities in strategic<br />
settings <strong>and</strong> address issues such as the interplay between innovati<strong>on</strong> <strong>and</strong> market structure, the<br />
dynamics of competiti<strong>on</strong> <strong>and</strong> the nature of the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>.<br />
In general, these models are structured <strong>on</strong> the fundamental distincti<strong>on</strong> between static <strong>and</strong><br />
dynamic competiti<strong>on</strong>. Static or product-market competiti<strong>on</strong> refers to the strategic interacti<strong>on</strong> that<br />
arises between firms, taking as given the capabilities of these firms in terms of the variety, quality<br />
<strong>and</strong> producti<strong>on</strong> costs of the products in their portfolios. Dynamic competiti<strong>on</strong> refers instead to the<br />
strategic interacti<strong>on</strong> am<strong>on</strong>g rival firms to change these capabilities.<br />
This distincti<strong>on</strong> is usually captured by the structure of game-theoretical models: firms are<br />
c<strong>on</strong>sidered to play a game where they first invest in innovative activities <strong>and</strong> then compete at the<br />
product-market level. The distincti<strong>on</strong> is a fictitious <strong>on</strong>e to a certain extent, since in reality both<br />
dimensi<strong>on</strong>s of competiti<strong>on</strong> are c<strong>on</strong>tinuously overlapping, but it is a c<strong>on</strong>vincing <strong>and</strong> powerful<br />
c<strong>on</strong>ceptual categorizati<strong>on</strong> since it allows distinguishing activities that occur over different timehoriz<strong>on</strong>s,<br />
i.e. short-term price competiti<strong>on</strong> versus. l<strong>on</strong>g-term dynamic competiti<strong>on</strong>. When firms<br />
compete in the short-term they have to take as given a number of c<strong>on</strong>straints in terms of their <strong>and</strong><br />
their rivals’ “capabilities” that derive from previous innovative activities; when competiti<strong>on</strong> is<br />
c<strong>on</strong>sidered over a l<strong>on</strong>ger-time horiz<strong>on</strong>, however, these “capabilities” become endogenous to the<br />
competitive process as well.<br />
Despite the absence of general results, game-theoretical models are certainly useful to<br />
underst<strong>and</strong> what fundamental strategic forces <strong>and</strong> incentives drive firms’ investments in R&D <strong>and</strong><br />
what factors may lead to different types of dynamic competiti<strong>on</strong> (i.e. persistence of m<strong>on</strong>opoly or<br />
acti<strong>on</strong>-reacti<strong>on</strong>).<br />
Endogenous (or new) growth models, pi<strong>on</strong>eered by Romer, Lucas, Aghi<strong>on</strong> <strong>and</strong> Howitt have tried<br />
to explain the ec<strong>on</strong>omic forces driving the technological progress underlying ec<strong>on</strong>omic growth. In<br />
these models ec<strong>on</strong>omic growth, in general, derives from the growth of the stock of knowledge<br />
available to society, which in turn is the outcome of ec<strong>on</strong>omic choices of profit-maximizing agents.<br />
Earlier models developed in the 1960s were instead mainly focused <strong>on</strong> capital accumulati<strong>on</strong> <strong>and</strong><br />
did not treat technological change as an ec<strong>on</strong>omic phenomen<strong>on</strong> amenable to ec<strong>on</strong>omic analysis.<br />
Although these models are based <strong>on</strong> the equilibrium methodology typical of the neo-classical<br />
approach they are inherently dynamic. The evoluti<strong>on</strong> of the stock of knowledge is explained as<br />
either being the product of unintenti<strong>on</strong>al learning-by-doing activities or the outcome of R&D<br />
investments undertaken by profit-maximizing agents <strong>and</strong> rewarded by the m<strong>on</strong>opolistic rents that<br />
the innovator is able to appropriate. Externalities involved in the producti<strong>on</strong> of new knowledge<br />
10<br />
228
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
usually play an important role in these models, <strong>and</strong> in particular knowledge spillovers by which<br />
the knowledge generated by an agent enhances the technological capabilities of other agents or<br />
their ability to discover new knowledge.<br />
This secti<strong>on</strong> c<strong>on</strong>siders first game-theoretical microec<strong>on</strong>omic models of innovati<strong>on</strong> developed in<br />
the industrial organizati<strong>on</strong> literature. It then c<strong>on</strong>siders endogenous growth theories <strong>and</strong> the new<br />
ec<strong>on</strong>omic geography theories.<br />
1.2 Industrial Organizati<strong>on</strong><br />
The game-theoretical literature <strong>on</strong> innovati<strong>on</strong>, which was developed mainly in the 1980s <strong>and</strong><br />
1990s, addresses the strategic interacti<strong>on</strong> that characterizes R&D investment decisi<strong>on</strong>s in<br />
oligopolistic industries. These models provide the fundamental building blocks of more recent<br />
endogenous growth models, which cannot be properly assessed without reference to this earlier<br />
literature.<br />
More recently, game-theoretical models have focused <strong>on</strong> the analysis of competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> in markets characterized by network effects, <strong>on</strong> the analysis of co-operati<strong>on</strong> in<br />
research <strong>and</strong> development activities <strong>and</strong> <strong>on</strong> the study of the relati<strong>on</strong>ship between intensity of<br />
competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
The starting point of this secti<strong>on</strong> is a categorizati<strong>on</strong> of game-theoretical models that should act as<br />
a guiding map to this literature. Underst<strong>and</strong>ing the nature of the models, <strong>and</strong> not <strong>on</strong>ly their<br />
results, is unavoidable to assess the adequacy of a particular model as theoretical reference in a<br />
practical case. The results of game theoretical models are generally very sensitive to the specific<br />
assumpti<strong>on</strong>s adopted <strong>and</strong> <strong>on</strong>ly by having clear the link between the former <strong>and</strong> the latter can<br />
these studies offer useful guidance for practical analyses of competiti<strong>on</strong> <strong>and</strong> market dynamics.<br />
The following secti<strong>on</strong>s focus <strong>on</strong> some of the issues that are addressed in the game-theoretical<br />
literature <strong>and</strong> that are most relevant to the analysis of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> for antitrust<br />
purposes:<br />
•<br />
•<br />
•<br />
•<br />
the relati<strong>on</strong>ship between market structure <strong>and</strong> innovati<strong>on</strong>;<br />
the issue of persistence of m<strong>on</strong>opoly in dynamic markets;<br />
the relati<strong>on</strong>ship between toughness of product market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>; <strong>and</strong><br />
R&D co-operati<strong>on</strong>.<br />
An overview of the game-theoretical literature <strong>on</strong> R&D<br />
Game-theoretical models of innovati<strong>on</strong> explain the strategic interacti<strong>on</strong> that underlies firms’<br />
investments in research <strong>and</strong> development activities. The focus <strong>on</strong> strategic behavior<br />
distinguishes this literature from earlier decisi<strong>on</strong> theoretic models, e.g. Kamien <strong>and</strong> Schwartz<br />
(1980).<br />
11<br />
229
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Game-theoretical models of innovati<strong>on</strong> can be categorized as bel<strong>on</strong>ging to <strong>on</strong>e of three broad<br />
paradigms:<br />
• N<strong>on</strong>-tournament models. These models capture the strategic interacti<strong>on</strong> that can arise in<br />
those cases where firms pursue different paths of technological advance so that a<br />
discovery made al<strong>on</strong>g <strong>on</strong>e path can be used irrespectively <strong>on</strong> whether a competitor<br />
innovates al<strong>on</strong>g another technological path. These models have been investigated by<br />
Dasgupta <strong>and</strong> Stiglitz (1980a), Spence (1984) <strong>and</strong> Judd (1985).<br />
•<br />
•<br />
Deterministic aucti<strong>on</strong> models. Aucti<strong>on</strong> models of innovati<strong>on</strong> describe R&D competiti<strong>on</strong> in<br />
terms of rival firms bidding for a given process or product innovati<strong>on</strong>. Important<br />
c<strong>on</strong>tributi<strong>on</strong>s include Gilbert <strong>and</strong> Newbery (1982), Katz <strong>and</strong> Shapiro (1987) <strong>and</strong> Vickers<br />
(1986).<br />
Stochastic tournament models (or “patent races”). These models capture the strategic<br />
interacti<strong>on</strong> that arises when firms compete to be the first to introduce an innovati<strong>on</strong> <strong>and</strong><br />
the issue of timing is crucial to firms’ strategic interacti<strong>on</strong>. Papers in this category include<br />
Loury (1979), Lee <strong>and</strong> Wilde (1980), Dasgupta <strong>and</strong> Stiglitz (1980b), Grossman <strong>and</strong><br />
Shapiro (1987), Harris <strong>and</strong> Vickers (1987), Reinganum (1981,1982).<br />
Each of these approaches describes R&D investments in a particular setting, which may be an<br />
appropriate depicti<strong>on</strong> of reality in some circumstances <strong>and</strong> a less appropriate theoretical<br />
reference in relati<strong>on</strong> to some other industries or to particular stages of a technology life cycle.<br />
Despite the differences, however, game-theoretical models of innovati<strong>on</strong> illustrate that there are<br />
two central ec<strong>on</strong>omic incentives at the basis of firms’ investments in R&D activities, whose<br />
relative importance varies according to the specific envir<strong>on</strong>ment c<strong>on</strong>sidered: the profit incentive<br />
<strong>and</strong> the competitive threat (or replacement effect).<br />
The profit incentive, or st<strong>and</strong>-al<strong>on</strong>e incentive, is related to the noti<strong>on</strong> that allocating resources<br />
to innovative research <strong>and</strong> development would, if successful, increase a firm’s profits. This<br />
incentive can be usually thought of as the incentive to invest in R&D of a firm that takes the<br />
innovati<strong>on</strong> decisi<strong>on</strong> in isolati<strong>on</strong>. In fact, it is equal to the difference between the profits that the<br />
firm would get if it innovates <strong>and</strong> those that it would get if it does not innovate, all else equal.<br />
The competitive threat, or pre-empti<strong>on</strong> incentive, arises when strategic interacti<strong>on</strong> in innovati<strong>on</strong><br />
activities is c<strong>on</strong>sidered: a firm not <strong>on</strong>ly has to take into account the benefits c<strong>on</strong>nected to<br />
innovati<strong>on</strong> but also the possible loss of competitiveness where it does not innovate <strong>and</strong> a<br />
competitor does. This incentive reflects the threat posed by the existence of an active rival <strong>and</strong><br />
can be thought of as the difference between the profits if the firm innovates <strong>and</strong> the profits if it is<br />
left to a rival firm to innovate.<br />
12<br />
230
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In general settings, both the profit incentive <strong>and</strong> the competitive threat may play a role in shaping<br />
the nature of dynamic competiti<strong>on</strong>. This would depend crucially <strong>on</strong> the specific characteristics of<br />
the market envir<strong>on</strong>ment c<strong>on</strong>sidered <strong>and</strong>, in the literature, <strong>on</strong> the specific assumpti<strong>on</strong>s adopted in<br />
each model.<br />
N<strong>on</strong>-tournament models<br />
N<strong>on</strong>-tournament models of innovati<strong>on</strong> investigate R&D investments in a setting where there can<br />
be multiple discoverers <strong>and</strong> firms do not compete to be the first to innovate. In these models<br />
symmetric competitors cannot prevent each other from getting equivalent improvements from<br />
spending equivalent amounts of resources in R&D.<br />
For this reas<strong>on</strong>, such models may explain R&D investments in those envir<strong>on</strong>ments where firms<br />
pursue different research paths so that an advance made <strong>on</strong> a single path can be appropriated<br />
irrespective of rivals’ advances <strong>on</strong> other research trajectories.<br />
A typical n<strong>on</strong>-tournament model, such as the <strong>on</strong>e by Dasgupta <strong>and</strong> Stiglitz (1980a), is structured<br />
in 3 stages: an entry stage, an R&D stage <strong>and</strong> a market competiti<strong>on</strong> stage. Firms first decide<br />
whether or not to enter an industry; active firms would then invest resources in R&D activities to<br />
decrease their marginal cost of producti<strong>on</strong> or increase the quality of their products <strong>and</strong> would<br />
finally compete in the marketplace.<br />
R&D activities are described in terms of a deterministic producti<strong>on</strong> functi<strong>on</strong> that specifies the size<br />
of innovati<strong>on</strong> according to the amount of cost reducti<strong>on</strong> or quality improvement per unit of R&D<br />
investment. Each firm can reduce its costs or improve the qualities of its products independently<br />
of parallel innovati<strong>on</strong> by rival firms, although it is usually assumed that a firm can indirectly benefit<br />
from a rival’s R&D activity for the existence of knowledge spillovers.<br />
Knowledge spillovers capture the noti<strong>on</strong> that innovative knowledge produced by R&D activities<br />
may, to some extent, be n<strong>on</strong>-excludable so that firms can, at least partially, appropriate the results<br />
of R&D activities of a rival. Cohen <strong>and</strong> Levinthal (1989) suggest that such learning may not be<br />
costless <strong>and</strong> that firms may have to invest in own R&D programs in order to access external<br />
knowledge <strong>and</strong> benefit from spillovers from innovating rivals.<br />
N<strong>on</strong>-tournament models have especially been employed to assess the extent of innovati<strong>on</strong><br />
achieved by the operati<strong>on</strong> of market forces <strong>and</strong> how it compares to the socially optimal <strong>on</strong>e, to<br />
discuss technological policy, <strong>and</strong> to investigate the relati<strong>on</strong>ship between market structure (i.e.<br />
c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong>.<br />
However, n<strong>on</strong>-tournament models assume that there are a large number of research paths <strong>and</strong><br />
the relati<strong>on</strong>ship between R&D investments <strong>and</strong> rewards from innovati<strong>on</strong> is essentially c<strong>on</strong>tinuous.<br />
For this reas<strong>on</strong> they are not well suited to describe market envir<strong>on</strong>ments where innovati<strong>on</strong> takes<br />
the form of a race <strong>and</strong> competiti<strong>on</strong> is for the market rather than in the market.<br />
Although this may be an appropriate descripti<strong>on</strong> for technological change in some markets, in<br />
many others innovati<strong>on</strong> is a more disruptive process that leads to winners <strong>and</strong> losers in very<br />
different competitive positi<strong>on</strong>s. In these circumstances, the rewards to R&D investments are<br />
13<br />
231
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
disc<strong>on</strong>tinuous, performance is rewarded <strong>on</strong> the basis of the rank within the set of realized<br />
performances, <strong>and</strong> competiti<strong>on</strong> takes the form of rival firms racing to be first to innovate<br />
(Dasgupta, 1986). An example of this type of race could be that characterizing competiti<strong>on</strong><br />
between Intel <strong>and</strong> AMD to introduce better PC processors.<br />
Aucti<strong>on</strong> models<br />
Aucti<strong>on</strong> models of innovati<strong>on</strong> describe R&D competiti<strong>on</strong> in terms of rival firms bidding for a given<br />
process or product innovati<strong>on</strong>. The innovati<strong>on</strong> is allocated to the firm which makes the highest bid<br />
<strong>and</strong> the issue is to investigate the identity of the firm which would have the larger willingness to<br />
pay for the innovati<strong>on</strong>.<br />
In general, the maximum bid that a firm would make is equal to the difference between the profits<br />
it would get if it were successful in the aucti<strong>on</strong> <strong>and</strong> those it would get were it not. For this reas<strong>on</strong>,<br />
aucti<strong>on</strong> models of innovati<strong>on</strong> emphasize the competitive threat as the incentive underlying<br />
competiti<strong>on</strong> in innovati<strong>on</strong>.<br />
Aucti<strong>on</strong> models have been widely used to investigate dynamic competiti<strong>on</strong> <strong>and</strong> whether market<br />
dynamics are characterized by persistent dominance by a technological leader or by “creative<br />
destructi<strong>on</strong>” whereby the incumbent is overtaken by some rival whose incumbency is itself shortlived.<br />
In a single aucti<strong>on</strong> model, the issue of dynamic competiti<strong>on</strong> is investigated by c<strong>on</strong>sidering a set of<br />
firms with different technological levels, i.e. marginal costs in the case of process innovati<strong>on</strong>, <strong>and</strong><br />
studying their incentives to purchase an innovati<strong>on</strong>. Market persistence is the result of such<br />
competiti<strong>on</strong> if the winner of the aucti<strong>on</strong> is the technological leader, otherwise dynamic competiti<strong>on</strong><br />
is characterized by acti<strong>on</strong>-reacti<strong>on</strong>.<br />
Assessing market dynamics in a single-innovati<strong>on</strong> aucti<strong>on</strong> model, however, may not help to<br />
underst<strong>and</strong> dynamic competiti<strong>on</strong> because in these models firms do not take into account how the<br />
outcome of the current bid may affect the outcome of future aucti<strong>on</strong>s. These inter-temporal links,<br />
however, may be an important element to explain firms’ incentives to invest in R&D in some<br />
c<strong>on</strong>texts where a sequence of innovative opportunities is in prospect. For this reas<strong>on</strong>, the<br />
questi<strong>on</strong> of dynamic competiti<strong>on</strong> may be better posed in the c<strong>on</strong>text of sequences of innovati<strong>on</strong><br />
rather than of a single innovati<strong>on</strong>.<br />
Aucti<strong>on</strong> models are intrinsically deterministic <strong>and</strong> this leads to the somewhat disappointing result<br />
that losers do not commit resources in R&D <strong>and</strong> that the process of technological advance is not<br />
really explained.<br />
14<br />
232
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Stochastic tournament models<br />
Tournament models of innovati<strong>on</strong> capture the particular interacti<strong>on</strong> that arises in those markets<br />
where dynamic competiti<strong>on</strong> takes the form of a race to be the first to make a discovery. As<br />
opposed to deterministic aucti<strong>on</strong> models, tournament models c<strong>on</strong>sider the case where there is<br />
technological uncertainty. This can be captured, for instance, by c<strong>on</strong>sidering a stochastic<br />
relati<strong>on</strong>ship between R&D effort <strong>and</strong> success.<br />
As such, these models emphasize timing in the c<strong>on</strong>text of R&D competiti<strong>on</strong> <strong>and</strong> are well suited to<br />
describe those envir<strong>on</strong>ments where competiti<strong>on</strong> is for the market rather than in the market. This<br />
nature of market dynamics can arise because technological change proceeds al<strong>on</strong>g a small<br />
number of research trajectories <strong>and</strong> there is competiti<strong>on</strong> to be the first to innovate because<br />
success by a firm would somehow prevent appropriability by subsequent innovators.<br />
The advantage of being the first to innovate is usually explained <strong>on</strong> the basis of the ability of the<br />
winner to appropriate the value of the innovati<strong>on</strong> by means of patent protecti<strong>on</strong>. More generally,<br />
however, appropriability may depend <strong>on</strong> factors other than patent protecti<strong>on</strong>, such as lead-time,<br />
access to complementary resources <strong>and</strong> secrecy. The relative importance of various<br />
appropriability factors is likely to be different across industries. (Levin et al., 1987)<br />
In aucti<strong>on</strong> models it is the competitive threat that characterizes firms’ incentives to invest in R&D<br />
activities. When r<strong>and</strong>omness is introduced, however, both the profit incentive <strong>and</strong> the competitive<br />
threat are relevant in firms’ strategic interacti<strong>on</strong>.<br />
In general, these forces need not be symmetric <strong>and</strong> have to be c<strong>on</strong>sidered for each firm involved:<br />
<strong>on</strong>e firm may face the greater competitive threat <strong>and</strong> another may have the greater profit<br />
incentive. Clearly, if a firm faces both the greater profit incentive <strong>and</strong> competitive threat it would<br />
undertake more R&D <strong>and</strong> hence be more likely to win the race. However, where there are<br />
asymmetric incentives, the nature of the outcome is not as straightforward.<br />
The literature usually focuses <strong>on</strong> the case where the competitive threat exceeds the profit<br />
incentive for all firms involved in the race. However, it could be the case that the profit incentive<br />
for both firms exceeds the competitive threat <strong>and</strong> that competiti<strong>on</strong> in R&D takes the form of a<br />
waiting game where firms tend to behave as free-riders <strong>and</strong> invest little resources in R&D. This<br />
type of situati<strong>on</strong> has been c<strong>on</strong>sidered, for instance, by Katz <strong>and</strong> Shapiro (1987) who explicitly<br />
account for the possibility of imitati<strong>on</strong> <strong>and</strong> licensing of innovati<strong>on</strong>.<br />
An important distincti<strong>on</strong> am<strong>on</strong>g tournament models is related to the assumpti<strong>on</strong> <strong>on</strong> the nature of<br />
R&D expenditures <strong>and</strong> whether these are c<strong>on</strong>tractual (e.g. Dasgupta <strong>and</strong> Stiglitz, 1980, Loury,<br />
1979) or n<strong>on</strong>-c<strong>on</strong>tractual (e.g. Lee <strong>and</strong> Wilde, 1980; <strong>and</strong> Reinganum, 1981,1982). In the first<br />
case R&D outlays take the form of a lump-sum cost incurred at the outset; in the sec<strong>on</strong>d case<br />
R&D can be c<strong>on</strong>sidered as a flow — a cost that it is incurred until the successful innovati<strong>on</strong> is<br />
achieved.<br />
Beath, Katsoulacos <strong>and</strong> Ulph (1994) discuss the implicati<strong>on</strong>s of these different assumpti<strong>on</strong>s.<br />
When R&D is of the c<strong>on</strong>tractual kind, equilibrium investments can be explained <strong>on</strong> a relatively<br />
15<br />
233
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
simple marginal benefit/marginal cost c<strong>on</strong>diti<strong>on</strong>. On the other h<strong>and</strong>, when R&D is of the n<strong>on</strong>c<strong>on</strong>tractual<br />
kind, it is the competitive threat that is the main determinant of a firm’s R&D.<br />
As with aucti<strong>on</strong> models, tournament models have also been developed to c<strong>on</strong>sider a sequence of<br />
tournaments in which firms complete a number of R&D stages before competing in the market.<br />
Examples of models that fall in this category are Fudenberg et al. (1983), Harris <strong>and</strong> Vickers<br />
(1987) <strong>and</strong> Grossman <strong>and</strong> Shapiro (1987).<br />
In these models innovati<strong>on</strong> is obtained <strong>on</strong>ly as the result of the completi<strong>on</strong> of a number of stages.<br />
A crucial issue that arises in such a setting is the nature of the innovative process <strong>and</strong> in particular<br />
whether each firm has to discover each technological step itself before it can move <strong>on</strong> to discover<br />
the next or whether instead the follower firm can compete directly with the leader for the new<br />
state-of-the-art technology.<br />
The latter case may arise if knowledge spillovers are relevant so that a laggard firm has access to<br />
(but may not have right to use) the same technological knowledge of the technological leader. On<br />
the other h<strong>and</strong>, knowledge may be excludable to some extent or in-house R&D is very important<br />
to assimilate external knowledge so that the technological fr<strong>on</strong>tier has to be achieved<br />
independently by each firm.<br />
The Grossman <strong>and</strong> Shapiro, Park <strong>and</strong> Harris <strong>and</strong> Vickers step-by-step models assume that each<br />
firm has to discover each technological step itself before it can move <strong>on</strong> to discover the next.<br />
Reinganum (1985) <strong>and</strong> Beath et al. (1987) are examples of leapfrog models where a follower is<br />
allowed at any time to compete directly with the leader for the new best-practice technology.<br />
In a step-by-step model firms can compete neck-to-neck or have different technological levels. In<br />
a leapfrog model firms are always asymmetric in terms of their technological ability. This<br />
distincti<strong>on</strong> turns out to be crucial in the analysis of the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>, as recent c<strong>on</strong>tributi<strong>on</strong>s in the endogenous growth literature surveyed below have<br />
shown.<br />
Endogenous growth theories have further developed these n<strong>on</strong>-tournament models by<br />
c<strong>on</strong>sidering sequences of tournaments where firms compete to innovate <strong>and</strong> compete in the<br />
product market before starting the following race to innovate.<br />
Market structure <strong>and</strong> innovati<strong>on</strong><br />
A l<strong>on</strong>g-st<strong>and</strong>ing questi<strong>on</strong> in the ec<strong>on</strong>omics of technological change has been the nature of the<br />
relati<strong>on</strong>ship between market structure <strong>and</strong> innovati<strong>on</strong>.<br />
The relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> of an industry <strong>and</strong> its rate of technological innovati<strong>on</strong> is<br />
complex. Market structure may have an impact <strong>on</strong> the rate of innovati<strong>on</strong>, but innovati<strong>on</strong> is also<br />
an important factor that shapes market structure. In fact, it is necessary to recognize that the<br />
16<br />
234
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is not a causal <strong>on</strong>e: both are the endogenous<br />
outcomes of the operati<strong>on</strong> of market forces <strong>and</strong> exogenous factors such as the nature of<br />
dem<strong>and</strong>, technological opportunity, the c<strong>on</strong>diti<strong>on</strong>s governing appropriability, <strong>and</strong> pure chance.<br />
The classical point of departure in the ec<strong>on</strong>omic analysis of the relati<strong>on</strong>ship between static market<br />
structure (i.e. c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong> is the work by Arrow (1962). Arrow c<strong>on</strong>siders the<br />
case where a cost-reducing innovati<strong>on</strong> is exogenously available <strong>and</strong> investigates firms’<br />
willingness to pay for the innovati<strong>on</strong> under different market structures. For a drastic innovati<strong>on</strong>,<br />
Arrow’s analysis shows that a firm that is already a m<strong>on</strong>opolist would have lower incentives to<br />
innovate than a firm that is currently in a perfectly competitive envir<strong>on</strong>ment, essentially because it<br />
would have the lower profit incentive.<br />
On the other h<strong>and</strong>, innovati<strong>on</strong> is certainly an important factor that affects market structure:<br />
innovati<strong>on</strong> is a means by which a firm tries to escape the c<strong>on</strong>straints imposed by competiti<strong>on</strong>.<br />
Studies in the Schumpeterian traditi<strong>on</strong> have emphasized the importance of ex-post market power<br />
for firms’ incentives to innovate. Some degree of market power is necessary for a firm to cover its<br />
R&D outlays: dynamic <strong>and</strong> static efficiency are somehow c<strong>on</strong>flicting. This is a theme that has<br />
been well developed in the recent literature <strong>on</strong> endogenous growth.<br />
Dasgupta <strong>and</strong> Stiglitz (1980a) discuss the relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> in a<br />
n<strong>on</strong>-tournament model. If there are exogenous entry barriers, an increase in the number of firms<br />
causes each firm to spend less <strong>on</strong> R&D in equilibrium, however total R&D expenditure increases<br />
with the number of firms. When entry is c<strong>on</strong>sidered to be endogenous, <strong>on</strong>e would observe more<br />
innovati<strong>on</strong> in those industries that are characterized by a higher degree of m<strong>on</strong>opoly power,<br />
although no causality should be imputed to this relati<strong>on</strong>ship.<br />
Sutt<strong>on</strong> (1998) adopts an innovative approach to the study of technology <strong>and</strong> market structure <strong>and</strong><br />
achieves a very general, albeit somehow loose, result <strong>on</strong> the relati<strong>on</strong>ship between c<strong>on</strong>centrati<strong>on</strong><br />
<strong>and</strong> R&D. He is able to characterize, under very general assumpti<strong>on</strong>s, a lower bound to market<br />
c<strong>on</strong>centrati<strong>on</strong> (in large markets) that holds in those industries where the effectiveness of R&D in<br />
raising c<strong>on</strong>sumers’ willingness to pay is high <strong>and</strong> the product groups within the industry are<br />
sufficiently close substitutes. By c<strong>on</strong>trast, in those industries where R&D is not effective or where<br />
product groups are not close substitutes, very fragmented market structures can also be<br />
equilibrium c<strong>on</strong>figurati<strong>on</strong>s no matter how large the market size is.<br />
The models discussed above c<strong>on</strong>sider the relati<strong>on</strong>ship between product market structure (i.e.<br />
c<strong>on</strong>centrati<strong>on</strong>) <strong>and</strong> innovati<strong>on</strong>. In the c<strong>on</strong>text of models where firms race to innovate another<br />
interesting questi<strong>on</strong> c<strong>on</strong>cerns the relati<strong>on</strong>ship between the number of firms that are part of the<br />
race <strong>and</strong> the pace of technological advance.<br />
17<br />
235
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Loury (1979) c<strong>on</strong>siders a tournament model of innovati<strong>on</strong> where R&D expenditures are<br />
committed upfr<strong>on</strong>t (that is, the probability of success depends <strong>on</strong> the scale of the R&D lab) <strong>and</strong><br />
shows that increasing the number of firms reduces the expected date of inventi<strong>on</strong>. Lee <strong>and</strong> Wilde<br />
(1980) c<strong>on</strong>sider the case where the probability of innovati<strong>on</strong> is related to the research intensity<br />
<strong>and</strong> show that as the number of firms in the industry increases, the equilibrium rate of investment<br />
per firm increases <strong>and</strong> success by any <strong>on</strong>e firm is hastened by an increase in the number of<br />
firms. However, the first success date is hastened by an increase in the number of firms in both<br />
models.<br />
Reinganum (1982) c<strong>on</strong>siders a setting where firms can vary the research intensity as in Lee <strong>and</strong><br />
Wilde (1980) but does not assume that the rate of expenditure is c<strong>on</strong>stant over time. Instead,<br />
firms may adjust R&D intensity in resp<strong>on</strong>se to elapsed time <strong>and</strong> rival progress. Reinganum<br />
shows that, in this setting, when imitati<strong>on</strong> is not possible, an increase in the number of firms<br />
increases the equilibrium rate of investment for each firm <strong>and</strong> decreases the expected time of<br />
innovati<strong>on</strong>. When there is no full patent protecti<strong>on</strong>, the relati<strong>on</strong>ship is ambiguous <strong>and</strong> depends <strong>on</strong><br />
the relative payoffs to the innovator <strong>and</strong> the imitators.<br />
The persistence of dominance<br />
In many industries characterized by fundamental l<strong>on</strong>g-term market dynamics competiti<strong>on</strong> may<br />
take the form of competiti<strong>on</strong> for the market rather than competiti<strong>on</strong> in the market.<br />
In these markets the issue is not whether more or less c<strong>on</strong>centrati<strong>on</strong> is associated with faster<br />
technological progress but whether market dynamics would be characterised by persistent<br />
dominance of the incumbent leader or by acti<strong>on</strong>-reacti<strong>on</strong> whereby incumbents are overtaken by<br />
some rivals whose incumbency is itself short-lived.<br />
The dynamic evoluti<strong>on</strong> of market structure depends <strong>on</strong> both abilities <strong>and</strong> incentives of the<br />
incumbent <strong>and</strong> the rivals to innovate. Game theoretical models are well suited to analyse the<br />
incentives underlying R&D investments <strong>and</strong> the resulting evoluti<strong>on</strong> of market structure. If we<br />
focus <strong>on</strong> ec<strong>on</strong>omic incentives, <strong>and</strong> set aside differences in R&D abilities, persistence of m<strong>on</strong>opoly<br />
or acti<strong>on</strong>-reacti<strong>on</strong> can be related to the different profit incentive <strong>and</strong> competitive threat faced by<br />
the leader <strong>and</strong> the follower(s).<br />
Gilbert <strong>and</strong> Newbery (1982) use the aucti<strong>on</strong> model to examine potential competiti<strong>on</strong> at the R&D<br />
stage <strong>and</strong> reverse Arrow’s (1962) result: potential competiti<strong>on</strong> matters <strong>and</strong> a m<strong>on</strong>opolist may<br />
have more incentives to innovate than a potential entrant because the incumbent m<strong>on</strong>opolist may<br />
face a larger competitive threat.<br />
The profit of a successful incumbent who innovates is that of a m<strong>on</strong>opolist firm, whereas if it were<br />
the entrant to innovate each firm would get the profit of a duopolist. Hence, the competitive threat<br />
of the incumbent can be measured as the difference between the profit of a m<strong>on</strong>opolist firm <strong>and</strong><br />
the profit of a duopolist firm. The incumbent’s competitive threat, instead, is simply equal to the<br />
profit of a duopolist. This implies that the incumbent would have more incentives to innovate (i.e.<br />
the larger competitive threat) if, as it is normally the case, the profit of a m<strong>on</strong>opolist is greater than<br />
the combined profits of two duopolists.<br />
18<br />
236
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Katz <strong>and</strong> Shapiro (1987) study a general aucti<strong>on</strong> model <strong>and</strong> show that for minor innovati<strong>on</strong>s, the<br />
industry leader will typically be the innovator, whether or not imitati<strong>on</strong> <strong>and</strong> licensing are feasible.<br />
For markets where patent protecti<strong>on</strong> is str<strong>on</strong>g, they predict that the major innovati<strong>on</strong>s will be<br />
made by industry leaders. But if imitati<strong>on</strong> is easy, industry followers or entrants will make the<br />
major discoveries.<br />
Reinganum (1983) in a tournament model shows that when the first innovator captures a<br />
sufficiently high share of the post innovati<strong>on</strong> market, then the incumbent firm invests less <strong>on</strong> a<br />
given project that does the potential entrant. This is because the incumbent has less incentive<br />
than the challenger to shorten the period of its incumbency.<br />
Beath, Katsoulacos <strong>and</strong> Ulph (1995) show that in a tournament model under Cournot<br />
competiti<strong>on</strong>, a large innovati<strong>on</strong> or a large initial gap results in persistent dominance while a small<br />
initial gap <strong>and</strong> a small innovati<strong>on</strong> results in acti<strong>on</strong>-reacti<strong>on</strong>. However, if market competiti<strong>on</strong> is<br />
Bertr<strong>and</strong> there is persistent dominance irrespective of the size of the innovati<strong>on</strong> or the initial gap.<br />
This is a result that points towards the possible trade-off between static <strong>and</strong> dynamic efficiency<br />
suggested also by Vickers (1986).<br />
Market dynamics have also been investigated in models that c<strong>on</strong>sider a sequence of innovati<strong>on</strong>s.<br />
Reinganum (1985) c<strong>on</strong>siders a sequence of drastic innovati<strong>on</strong>s <strong>and</strong> shows that market dynamics<br />
are characterised by a process that resembles Schumpeter’s process of creative destructi<strong>on</strong>: the<br />
incumbent invests less than each challenger in each stage.<br />
Vickers (1986) c<strong>on</strong>siders a sequence of n<strong>on</strong>-drastic process innovati<strong>on</strong> in the c<strong>on</strong>text of the<br />
aucti<strong>on</strong> model. He compares market dynamics under Bertr<strong>and</strong> <strong>and</strong> Cournot competiti<strong>on</strong> <strong>and</strong><br />
finds that when the product market is very competitive then there is increasing dominance, but<br />
when it is not very competitive there is acti<strong>on</strong>-reacti<strong>on</strong>. When innovati<strong>on</strong> is drastic, then market<br />
dynamics are characterised by increasing dominance.<br />
Reinganum (1989) observes that the differences in the results obtained in different models can be<br />
ascribed to the different roles that the profit incentive <strong>and</strong> the competitive threat play in stochastic<br />
tournament <strong>and</strong> deterministic aucti<strong>on</strong> models. In a deterministic model the incentive to pre-empt<br />
(larger for the incumbent) dominates the firms’ decisi<strong>on</strong>. When success is stochastic, however,<br />
the threat from the rival innovating is less acute. In the case of drastic innovati<strong>on</strong>s, the<br />
competitive threat is the same for both firms <strong>and</strong> it is the profit incentive (which is larger for<br />
challengers) that determines R&D investments. The relevance of the profit incentive extends to<br />
the case for some n<strong>on</strong>-drastic innovati<strong>on</strong>.<br />
Competiti<strong>on</strong> intensity <strong>and</strong> innovati<strong>on</strong><br />
Market structure is often associated with the c<strong>on</strong>cept of competitiveness: usually, a high level of<br />
c<strong>on</strong>centrati<strong>on</strong> in an industry is interpreted as weak competiti<strong>on</strong>. This view is based <strong>on</strong> a<br />
symmetric Cournot model, where price-cost margins are higher as the number of firms increases.<br />
However, it is preferable to disentangle the noti<strong>on</strong>s of market structure <strong>and</strong> toughness of price<br />
competiti<strong>on</strong> as, am<strong>on</strong>g others, Sutt<strong>on</strong> (1998) <strong>and</strong> Bo<strong>on</strong>e (2000, 2001) have d<strong>on</strong>e.<br />
19<br />
237
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
In a simple world with homogeneous firms, toughness of price competiti<strong>on</strong> can be c<strong>on</strong>sidered as<br />
being related to the level of price-cost margins given any level of c<strong>on</strong>centrati<strong>on</strong>. Hence, a<br />
differentiated Bertr<strong>and</strong> market would be c<strong>on</strong>sidered to be more competitive than a differentiated<br />
Cournot market because price-cost margins would be lower in the former, for any level of<br />
c<strong>on</strong>centrati<strong>on</strong>. As a result, more competitive markets may allow fewer firms to profitably survive.<br />
Recently, Bo<strong>on</strong>e (2000,2001) has offered an interesting c<strong>on</strong>tributi<strong>on</strong> to the study of the<br />
relati<strong>on</strong>ship between toughness of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. Bo<strong>on</strong>e defines intensity of<br />
competiti<strong>on</strong> <strong>on</strong> the basis of an axiomatic approach, which allows the result to be general to a<br />
large class of specific game-theoretical models. Bo<strong>on</strong>e c<strong>on</strong>siders four axioms:<br />
•<br />
•<br />
•<br />
•<br />
the lower limit of competiti<strong>on</strong> is such that firms are not affected by opp<strong>on</strong>ents’ acti<strong>on</strong>s;<br />
the least efficient firm in the market loses as competiti<strong>on</strong> becomes more intense;<br />
if the leader is far enough ahead, he gains as competiti<strong>on</strong> becomes more intense; <strong>and</strong><br />
if all active firms have similar costs, they all lose as competiti<strong>on</strong> becomes more intense.<br />
Bo<strong>on</strong>e (2001) shows that this definiti<strong>on</strong> encompasses, for instance, a switch from Cournot to<br />
Bertr<strong>and</strong> competiti<strong>on</strong> <strong>and</strong> a reducti<strong>on</strong> in travel costs in typical horiz<strong>on</strong>tal product differentiati<strong>on</strong><br />
models.<br />
When the noti<strong>on</strong> of intensity of competiti<strong>on</strong> is defined in such a way, it is normally inversely related<br />
to equilibrium c<strong>on</strong>centrati<strong>on</strong>. For instance, in Bo<strong>on</strong>e’s setting, if intensity of competiti<strong>on</strong> is low,<br />
then a large number of firms can be active in the market because the less efficient firms can<br />
survive as well. On the other h<strong>and</strong>, fierce competiti<strong>on</strong> is associated with high c<strong>on</strong>centrati<strong>on</strong><br />
because <strong>on</strong>ly the most efficient firms can survive.<br />
On the basis of this definiti<strong>on</strong> of product market competiti<strong>on</strong>, Bo<strong>on</strong>e (2001) investigates the<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> firms’ incentives to invest in R&D. The central<br />
finding of this study is that the relati<strong>on</strong>ship between market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> may not<br />
be m<strong>on</strong>ot<strong>on</strong>ic; this is so because varying intensity of competiti<strong>on</strong> is associated with different<br />
identity of the innovator.<br />
Bo<strong>on</strong>e (2001) explains this interesting result as follow. When intensity of competiti<strong>on</strong> is low it is<br />
the follower that would be the next innovator, whereas when intensity of competiti<strong>on</strong> is large, it is<br />
the current technological leader that is likely to innovate. When it is the follower that innovates,<br />
tougher competiti<strong>on</strong> implies lower profits <strong>and</strong> hence lower incentives to innovate. However, when<br />
it is the current leader to innovate, an increase of toughness of price competiti<strong>on</strong> would further<br />
increase the profits related to his technological leadership <strong>and</strong> hence would increase the value of<br />
innovati<strong>on</strong> for the firm. Hence, the relati<strong>on</strong>ship between toughness of price competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> would be U-shaped.<br />
Denicolò (2002) observes that Bo<strong>on</strong>e’s paper generalises the findings of a number of other<br />
studies. Delb<strong>on</strong>o <strong>and</strong> Denicolò (1990) c<strong>on</strong>sidered a homogeneous product market <strong>and</strong> found<br />
that Bertr<strong>and</strong> duopolists have greater incentive to innovate than Cournot duopolists. Bester <strong>and</strong><br />
20<br />
238
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Petrakis (1993) <strong>and</strong> B<strong>on</strong>anno <strong>and</strong> Haworth (1998) showed respectively that horiz<strong>on</strong>tal <strong>and</strong><br />
vertical product differentiati<strong>on</strong> could reverse this result. Qiu (1997) compared the two regimes in<br />
a n<strong>on</strong>-tournament model <strong>and</strong> find that is such a framework the incentive to innovate is greater<br />
with Cournot competiti<strong>on</strong> even in the case of homogenous products.<br />
Other studies that investigate the relati<strong>on</strong>ship between toughness of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
fall within the endogenous growth literature ,Aghi<strong>on</strong> <strong>and</strong> Griffith (2005) <strong>and</strong> are surveyed in the<br />
relevant secti<strong>on</strong> of this<br />
review.<br />
R&D Co-operati<strong>on</strong><br />
Increasingly, research joint ventures (RJVs) have become a widespread form of industrial cooperati<strong>on</strong>.<br />
In parallel, the study of the effects of R&D co-operati<strong>on</strong> has also emerged as an<br />
important research topic when c<strong>on</strong>sidering industry policy.<br />
Co-operative R&D ventures can take many forms, ranging from R&D joint ventures, to crosslicensing<br />
agreements <strong>and</strong> various informal types of technology trading or informati<strong>on</strong>-sharing<br />
agreements.<br />
Research Joint Ventures themselves can cover a variety of arrangements. One type of RJV is<br />
the traditi<strong>on</strong>al joint venture, in which two or more parties create a separate entity in which they all<br />
have equity interests to c<strong>on</strong>duct well-defined R&D projects for their benefits. Another type is the<br />
research c<strong>on</strong>sortium. A third form of RJV is the venture capital investment by firms in a st<strong>and</strong>al<strong>on</strong>e<br />
start-up company.<br />
Ec<strong>on</strong>omists have investigated the extent to which RJVs might allow firms to internalize spillovers,<br />
co-ordinate their research activities <strong>and</strong> achieve higher R&D efficiency. Co-operative R&D is<br />
thought to be socially beneficial for several reas<strong>on</strong>s:<br />
•<br />
•<br />
•<br />
RJVs can alleviate the under-provisi<strong>on</strong> of R&D effort that results from technological<br />
spillovers <strong>and</strong> other sub-optimal R&D decisi<strong>on</strong>s.<br />
Co-operati<strong>on</strong> can lead to greater disseminati<strong>on</strong> of R&D results. It can improve R&D<br />
efficiency through good research design <strong>and</strong> informati<strong>on</strong> sharing, avoiding needless<br />
duplicati<strong>on</strong> of resources.<br />
Co-operative R&D enables the firms to share risks, to exploit synergies, pool different<br />
complementary assets, <strong>and</strong> exploit increasing returns to scale in R&D. It can enable firms<br />
to overcome a cost-of-development barrier impenetrable to any <strong>on</strong>e of them al<strong>on</strong>e.<br />
Adding to these effects, an important issue that is not very developed in the industrial organizati<strong>on</strong><br />
literature but is emphasized in the evoluti<strong>on</strong>ary ec<strong>on</strong>omics <strong>and</strong> the management literature <strong>on</strong><br />
21<br />
239
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
inter-firm co-operati<strong>on</strong> is that cooperative ventures may be an important means by which firms<br />
exchange specialised know-how <strong>and</strong> tacit knowledge.<br />
Countering these social benefits is the fear that firms participating in co-operative R&D might use<br />
the venture to engage in anti-competitive practices or that they might free ride <strong>on</strong> each other.<br />
RJVs <strong>and</strong> R&D investments<br />
Since the early c<strong>on</strong>tributi<strong>on</strong>s to the literature <strong>on</strong> RJVs, the effects of R&D co-operati<strong>on</strong> <strong>on</strong> the<br />
amount of resources invested in innovative activities have been extensively investigated.<br />
This attenti<strong>on</strong> is immediately related to the study of the potential of RJVs to correct the suboptimal<br />
R&D decisi<strong>on</strong>s that may characterise independent decisi<strong>on</strong>-making.<br />
The results of these studies accord support to the presumpti<strong>on</strong> that RJVs, by internalising<br />
externalities which affect the efficiency of n<strong>on</strong>-cooperative decisi<strong>on</strong>-making, can help to correct<br />
the n<strong>on</strong>-optimality of independent investments in R&D. In explaining what impact RJVs may<br />
have <strong>on</strong> R&D investments, knowledge spillovers are widely acknowledged as a crucial factor, <strong>and</strong><br />
their c<strong>on</strong>siderati<strong>on</strong> plays an important role in most ec<strong>on</strong>omic models of RJVs.<br />
In their seminal paper, d’Asprem<strong>on</strong>t <strong>and</strong> Jacquemin (1988) pi<strong>on</strong>eered an influential approach to<br />
study the effects of co-operati<strong>on</strong> <strong>on</strong> R&D investments. They c<strong>on</strong>sider a two-stage model in which<br />
duopolists first c<strong>on</strong>duct partially inappropriable research leading to a reducti<strong>on</strong> in unit cost, <strong>and</strong><br />
then compete à la Cournot in the product market. The focus of their analysis is <strong>on</strong> the<br />
comparis<strong>on</strong> of the magnitude of cost-reducing technical advance achieved when firms c<strong>on</strong>duct<br />
R&D competitively versus co-operatively, in the presence of spillover effects.<br />
The difference between the independent decisi<strong>on</strong>-making equilibrium <strong>and</strong> the co-operative <strong>on</strong>e<br />
results from the balance of the spillover internalisati<strong>on</strong> effect <strong>and</strong> the reduced strategic incentive<br />
effect. Joint decisi<strong>on</strong> <strong>on</strong> the levels of R&D expenditures internalises two externalities disregarded<br />
under independent decisi<strong>on</strong>-making:<br />
• Research knowledge is partially inappropriable, <strong>and</strong> it can benefit other firms without the<br />
innovator being able to fully appropriate the value of the knowledge created. This<br />
externality causes sub-optimality of independent R&D investments. Its internalisati<strong>on</strong><br />
under joint-decisi<strong>on</strong> making (spillover internalisati<strong>on</strong> effect) will tend to increase cooperative<br />
R&D investment relative to the n<strong>on</strong>-cooperative level.<br />
•<br />
The innovative activity of a firm comprises also a negative pecuniary externality <strong>on</strong> the<br />
profits of the rival through a clear strategic market interacti<strong>on</strong>. Joint decisi<strong>on</strong>-making<br />
internalises this externality (reduced strategic incentive effect) <strong>and</strong> tends to decrease the<br />
co-operative level of R&D investment relative to the n<strong>on</strong>-cooperative equilibrium.<br />
The balance of these opposite forces is crucial to underst<strong>and</strong> the effects of co-operati<strong>on</strong> <strong>on</strong> the<br />
level of resources invested in research <strong>and</strong>, in turn, <strong>on</strong> the appropriateness of co-operati<strong>on</strong> to<br />
reduce the gap between the private <strong>and</strong> social incentives for doing R&D.<br />
22<br />
240
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
In case of “high” spillover rates, the spillover internalizati<strong>on</strong> effect would dominate the reduced<br />
strategic incentive effect, the net effect being an increase in the R&D expenditures. By c<strong>on</strong>trast, if<br />
spillovers are “low”, R&D co-operati<strong>on</strong> would cause R&D investments to decrease relative to the<br />
n<strong>on</strong>-cooperative equilibrium.<br />
The c<strong>on</strong>tributi<strong>on</strong>s of Kamien, Muller <strong>and</strong> Zang (1992), Suzumura (1992) Simps<strong>on</strong> <strong>and</strong> V<strong>on</strong>ortas<br />
(1994), have proposed extensi<strong>on</strong>s of the basic modelling framework to cover an arbitrary number<br />
of firm, general dem<strong>and</strong> <strong>and</strong> cost assumpti<strong>on</strong>s.<br />
RJVs <strong>and</strong> R&D efficiency<br />
As we have discussed, c<strong>on</strong>siderable attenti<strong>on</strong> has been devoted to the study of the effects of<br />
RJVs <strong>on</strong> the amount of resources allocated to R&D, <strong>and</strong> of the advantages of technological cooperati<strong>on</strong><br />
in terms of correcting inefficient R&D efforts that independent decisi<strong>on</strong>-making involves.<br />
Another comm<strong>on</strong>ly acknowledged advantage of R&D co-operati<strong>on</strong> is identified in the increased<br />
efficiency of R&D activities: RJVs not <strong>on</strong>ly can affect the absolute levels of R&D spending, but<br />
also the amount of R&D spending per unit of cost reducti<strong>on</strong>/quality improvement achieved.<br />
Indeed, to the extent that co-operative R&D is more widely disseminated than individually<br />
c<strong>on</strong>ducted R&D, ex ante co-operati<strong>on</strong> increases the efficiency of R&D efforts because a single<br />
investment benefits a greater number of firms.<br />
The efficiency gain can have three types of positive effects. First, sharing lowers the cost of<br />
investment for each firm, which may induce them to c<strong>on</strong>duct more R&D. Sec<strong>on</strong>d, for a given<br />
level of R&D investment it might increase the effective amount of R&D. Third, co-operative R&D<br />
can eliminate the wasteful duplicati<strong>on</strong> that would occur if several firms separately undertook the<br />
same projects. Even if several firms c<strong>on</strong>tinue to c<strong>on</strong>duct separate R&D under an ex ante<br />
agreement, they still can improve the efficiency of their efforts by co-ordinating them.<br />
Some studies have formalized these intuiti<strong>on</strong>s, by investigating the effects of co-operati<strong>on</strong> <strong>on</strong> the<br />
level of informati<strong>on</strong>-sharing <strong>and</strong> the co-ordinati<strong>on</strong> of research activities.<br />
Katz (1986) presented an early, albeit isolated, study in which firms choose the level of spillovers<br />
in the RJV, by determining the amount of informati<strong>on</strong>-sharing achieved. However, the<br />
endogeneity of spillovers is restricted to the co-operative equilibrium <strong>on</strong>ly, <strong>and</strong> this limits the<br />
significance of the comparis<strong>on</strong> of the (exogenous) n<strong>on</strong>-cooperative <strong>and</strong> (endogenous) cooperative<br />
informati<strong>on</strong>-sharing levels.<br />
To fully exploit the endogenizati<strong>on</strong> of knowledge spillovers, <strong>and</strong> explain the differences of<br />
informati<strong>on</strong>-sharing levels in the co-operative <strong>and</strong> n<strong>on</strong>-cooperative equilibrium it is necessary to<br />
23<br />
241
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
model both as the result of firms’ choices. Katsoulacos <strong>and</strong> Ulph (1998) have presented a<br />
stochastic model of innovati<strong>on</strong> with this feature.<br />
The authors adopt a structured approach to model knowledge spillovers by distinguishing two<br />
relevant dimensi<strong>on</strong>s: the adaptability of the research to the other firm <strong>and</strong> the amount of<br />
informati<strong>on</strong>-sharing that actually takes place.<br />
They show that the n<strong>on</strong>-cooperative equilibrium generally produces minimal spillovers; in this<br />
case, RJVs might lead to a higher informati<strong>on</strong>-sharing. However, if firms operate in different but<br />
complementary industries, it is possible to achieve maximum informati<strong>on</strong>-sharing also in the n<strong>on</strong>cooperative<br />
equilibrium.<br />
Poyago-Theotoky (1999) analyses endogenous spillovers in the c<strong>on</strong>text of a simple n<strong>on</strong>tournament<br />
model of R&D where firms are engaged in cost-reducing innovati<strong>on</strong>. It is shown that<br />
when spillovers of informati<strong>on</strong> are treated as endogenous firms never disclose any of their<br />
informati<strong>on</strong> when choosing their R&D n<strong>on</strong>-cooperatively. Under co-operative R&D, firms will<br />
always choose to fully share their informati<strong>on</strong>, i.e. a research joint venture will operate with a<br />
maximum spillover value.<br />
The relati<strong>on</strong>ship between the research paths followed by firms can be manifold, spanning a<br />
c<strong>on</strong>tinuum from perfect complementarity to perfect substitutability. Co-operati<strong>on</strong> in R&D can take<br />
advantage of joint research design, by choosing the number of labs to operate for optimally<br />
exploiting such substitutability or complementarity.<br />
The issue of organisati<strong>on</strong> design in an RJV has been further c<strong>on</strong>sidered by Katsoulacos <strong>and</strong> Ulph<br />
(1998) who relate this choice to the nature of research paths followed <strong>and</strong> the endogenous level<br />
of informati<strong>on</strong>-sharing chosen by firms. An important result of this study is that cost<br />
c<strong>on</strong>siderati<strong>on</strong>s, dominant when research outputs are close substitutes <strong>and</strong> firms are willing to<br />
share a lot of informati<strong>on</strong>, are <strong>on</strong>ly <strong>on</strong>e of the factors behind the RJVs decisi<strong>on</strong> as to whether to<br />
operate <strong>on</strong>e or two labs.<br />
Indeed, the RJV may close a lab for anti-competitive reas<strong>on</strong>s, to prevent having to face a very<br />
competitive situati<strong>on</strong> when both firms discover. Or, when there are very str<strong>on</strong>g complementarities<br />
between the research outputs of the two firms, the RJV might prefer to keep both labs open in<br />
order to exploit such complementarities.<br />
1.3 Endogenous Growth Models<br />
Technological change is a fundamental driver of ec<strong>on</strong>omic growth. However, <strong>on</strong>ly recently have<br />
the ec<strong>on</strong>omic forces that underlie technological progress in a dynamic ec<strong>on</strong>omy been<br />
investigated. Advances in this field have been associated with the development of endogenous<br />
growth theories, <strong>on</strong>e of the most fertile grounds for ec<strong>on</strong>omic research in the last 15 years. The<br />
relevant research agenda has been set <strong>on</strong> the study of ec<strong>on</strong>omic growth as the result of<br />
knowledge producti<strong>on</strong> — we shall focus <strong>on</strong> technological knowledge in this survey — which is<br />
explained as the outcome of ec<strong>on</strong>omic decisi<strong>on</strong>s.<br />
24<br />
242
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Before the development of endogenous growth theories ec<strong>on</strong>omic models of growth were<br />
essentially models of capital accumulati<strong>on</strong> developed <strong>on</strong> the basis of Solow’s (1956) seminal<br />
c<strong>on</strong>tributi<strong>on</strong>. These models suggested that the simple accumulati<strong>on</strong> of physical capital cannot<br />
sustain l<strong>on</strong>g-run ec<strong>on</strong>omic growth as l<strong>on</strong>g as its marginal productivity is decreasing <strong>and</strong> not<br />
bounded from zero. L<strong>on</strong>g run ec<strong>on</strong>omic growth had to be attributed to the effect of some other<br />
factor, exogenous to the ec<strong>on</strong>omic choices analyzed in these models, such as technological<br />
change.<br />
In fact, even in older exogenous growth theories that studied capital accumulati<strong>on</strong>, technological<br />
change was acknowledged as <strong>on</strong>e of the major factors that explain the ability of an ec<strong>on</strong>omy to<br />
grow in the l<strong>on</strong>g run. Certainly, advances in technology are a desirable element in any<br />
explanati<strong>on</strong> of growth: “a story of growth that neglects technological progress is both ahistorical<br />
<strong>and</strong> implausible”(Grossman <strong>and</strong> Helpman,1994,p.26). Cliometric studies have also examined the<br />
relati<strong>on</strong>ship between knowledge<br />
<strong>and</strong> ec<strong>on</strong>omic growth as shown in the historical background below.<br />
Nevertheless, for a l<strong>on</strong>g time the forces underlying advances were not explained <strong>on</strong> the basis of<br />
ec<strong>on</strong>omic incentives but were essentially related to the exogenous evoluti<strong>on</strong> of scientific<br />
knowledge <strong>and</strong> its technological applicati<strong>on</strong>s.<br />
Endogenous growth theories or new growth theory have certainly c<strong>on</strong>tributed to fill this gap.<br />
These theories are based <strong>on</strong> product differentiati<strong>on</strong> <strong>and</strong> imperfect competiti<strong>on</strong>, ec<strong>on</strong>omies of<br />
scale <strong>and</strong> increasing returns instead of perfect competiti<strong>on</strong> <strong>and</strong> diminishing returns put forward by<br />
neo-classical theory. Freeman (1994) argues that these theories have been c<strong>on</strong>siderably<br />
influenced by the neo-Schumpeterian school of innovati<strong>on</strong>.<br />
New growth theory not <strong>on</strong>ly emphasizes knowledge accumulati<strong>on</strong> <strong>and</strong> diffusi<strong>on</strong>, in the form<br />
mainly of human capital (Lucas, 1988, Rebelo, 1991) or technological innovati<strong>on</strong> (Romer, 1990;<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt 1992), as fundamental drivers of ec<strong>on</strong>omic growth but also the ec<strong>on</strong>omic<br />
decisi<strong>on</strong>s that underlie both.<br />
The model developed by Rebelo (1991) represents the first generati<strong>on</strong> of endogenous growth<br />
models that underscore the importance of human capital accumulati<strong>on</strong>. In this model, there is no<br />
exogenous productivity growth given that the entirety of output growth is explained by human<br />
capital <strong>and</strong> capital stock accumulati<strong>on</strong>.<br />
As such, this review focuses <strong>on</strong> the sec<strong>on</strong>d generati<strong>on</strong> of endogenous growth models that<br />
c<strong>on</strong>sider technological innovati<strong>on</strong> as well as human capital accumulati<strong>on</strong>. In general, these<br />
models relate ec<strong>on</strong>omic growth to innovati<strong>on</strong>s that improve firms’ productive efficiency <strong>and</strong>/or<br />
lead to the producti<strong>on</strong> of new or better intermediate <strong>and</strong> c<strong>on</strong>sumer goods.<br />
8<br />
9<br />
25<br />
243
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Innovati<strong>on</strong>s result from new knowledge produced as the outcome of firms’ learning-by-doing<br />
<strong>and</strong>/or intenti<strong>on</strong>al R&D activities. In general, R&D models of endogenous growth have drawn<br />
heavily from the earlier game theoretical literature <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> extended those analyses to<br />
more dynamic settings.As such, they have developed the analysis of l<strong>on</strong>g-st<strong>and</strong>ing questi<strong>on</strong>s<br />
investigating, for instance, the relati<strong>on</strong>ship between the intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong><br />
<strong>and</strong> the nature of market dynamics.<br />
Recent ec<strong>on</strong>omic analysis in this field has offered interesting new insights <strong>on</strong> the relati<strong>on</strong>ship<br />
between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. Despite some suggestive c<strong>on</strong>clusi<strong>on</strong>s, however,<br />
endogenous growth models suffer from the same limitati<strong>on</strong>s as their c<strong>on</strong>stituent micro<br />
foundati<strong>on</strong>s: results often depend crucially from the assumpti<strong>on</strong>s adopted <strong>and</strong> the specific<br />
structure of the model studied. Unfortunately, in many cases some of these important<br />
assumpti<strong>on</strong>s relate to features of market envir<strong>on</strong>ments that are hard to observe <strong>and</strong> evaluate.<br />
Ec<strong>on</strong>omic theory provides no general results that can be c<strong>on</strong>sidered to hold in any market situati<strong>on</strong>,<br />
but rather specific c<strong>on</strong>siderati<strong>on</strong>s that suggest which features of a market envir<strong>on</strong>ment are worth<br />
c<strong>on</strong>sidering to assess the nature of competiti<strong>on</strong>, market dynamics ,their relati<strong>on</strong>ship <strong>and</strong> growth.<br />
(Aghi<strong>on</strong> <strong>and</strong> Griffith, 2005).<br />
Learning-by-doing<br />
Knowledge generated by learning-by-doing can be a significant engine of ec<strong>on</strong>omic growth <strong>and</strong><br />
models of endogenous growth with technological progress driven by learning-by-doing have been<br />
studied by Romer (1986), Young (1991,1993), Stokey (1988) <strong>and</strong> Aghi<strong>on</strong> <strong>and</strong> Howitt (1998).<br />
Learning-by-doing, since the seminal work of Arrow (1962), Solow (1997) has been c<strong>on</strong>sidered an<br />
important source of technological knowledge. Firms can produce new knowledge without investing<br />
in instituti<strong>on</strong>alized research <strong>and</strong> development activities but rather, as a by-product of their usual<br />
producti<strong>on</strong> activity. Indeed, experience has been shown to be an important source of<br />
technological improvements in many industries such as artificial fibres, semic<strong>on</strong>ductors <strong>and</strong><br />
memory chips.<br />
26<br />
244
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Generally, in these models learning-by-doing is the unintenti<strong>on</strong>al by-product of producti<strong>on</strong> or<br />
investment activities <strong>and</strong> it is usually c<strong>on</strong>sidered to result in technological knowledge that is <strong>on</strong>ly<br />
partially appropriable. The existence of knowledge spillovers is in fact generally acknowledged: a<br />
firm can, in part at least, benefit of the knowledge generated by the learning-by-doing of other<br />
firms.<br />
Young (1991, 1993) studies a setting where new technologies are discovered through R&D<br />
activities but are initially inferior to the older technologies they seek to replace. Experience,<br />
however, generates incremental improvements over time, which allow the new technology to<br />
supplant older technologies. These incremental improvements are <strong>on</strong>ly achievable to some<br />
extent, in the sense that there is a bound to the technological advance that can be generated by<br />
learning-by-doing. This hybrid model emphasises that inventive activity <strong>and</strong> producti<strong>on</strong><br />
experience are complementary forces in driving technological change.<br />
R&D models<br />
Endogenous growth models have also investigated ec<strong>on</strong>omic growth as driven by the<br />
accumulati<strong>on</strong> <strong>and</strong> diffusi<strong>on</strong> of knowledge generated by Research <strong>and</strong> <strong>Development</strong> activities,<br />
undertaken by profit-maximising firms.<br />
These models investigate the process of discovery of new or better products. Accordingly, a<br />
distincti<strong>on</strong> can be drawn between those models that treat innovati<strong>on</strong> in terms of the development<br />
of new varieties of products <strong>and</strong> those that c<strong>on</strong>sider sequential improvements of the quality of<br />
existing products.<br />
The major difference between these two types of models is that the latter capture an important<br />
characteristic of the innovative process, namely that new inventi<strong>on</strong>s entail an element of<br />
destructi<strong>on</strong> because they make old technologies or products obsolete. For this reas<strong>on</strong>, early<br />
endogenous growth models with sequential improvements in the quality of products are called<br />
Schumpeterian.<br />
Despite these differences however, the two types of models share the same fundamental<br />
structure. In both models, firms invest resources to acquire the exclusive right to manufacture a<br />
new product <strong>and</strong> it is usually assumed that R&D activities generate knowledge spillovers that<br />
benefit other research firms. The producti<strong>on</strong> of new knowledge <strong>and</strong> its diffusi<strong>on</strong> drive ec<strong>on</strong>omic<br />
growth in the l<strong>on</strong>g run.<br />
Variety of products<br />
Romer (1990) provides the seminal model where technological progress is captured in terms of<br />
an increasing variety of products. The increase of the variety of intermediate or final goods is<br />
reflected in higher productivity or utility <strong>and</strong> hence ec<strong>on</strong>omic growth.<br />
27<br />
245
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
New varieties of products are discovered as outcomes of R&D investments <strong>and</strong> remunerated by<br />
the rents that accrue to the successful innovator who can benefit from patent protecti<strong>on</strong> (or, more<br />
generally, of other appropriability factors such as lead time, secrecy, access to complementary<br />
resources).<br />
In these models, it is generally assumed that R&D activities lead both to new knowledge that can<br />
be appropriated by the innovator <strong>and</strong> new general knowledge that cannot be appropriated but<br />
instead c<strong>on</strong>tributes to a stock of publicly available knowledge that facilitates further discoveries.<br />
Not <strong>on</strong>ly is the producti<strong>on</strong> of new knowledge necessary for ec<strong>on</strong>omic growth, but also its<br />
diffusi<strong>on</strong>.<br />
A distinguishing feature of Romer’s (1990) growth model is that it studies horiz<strong>on</strong>tal product<br />
innovati<strong>on</strong>s that involve no obsolescence: new goods are never close substitutes for existing<br />
goods (i.e. each new product finds its own horiz<strong>on</strong>tal niche).<br />
Although this feature of the model may detract from its realism is some cases, it is not necessarily<br />
so. There are some industries where innovati<strong>on</strong> is fundamentally directed at the introducti<strong>on</strong> of<br />
new varieties of products rather than at improving products’ qualities. An example discussed in<br />
the literature is the flowmeter industry where technological advance takes the form of the<br />
development of new types of flowmeters. In this industry there is no escalati<strong>on</strong> of R&D <strong>on</strong> a<br />
single technological trajectory <strong>and</strong> related product group but proliferati<strong>on</strong> of product groups.<br />
Quality of products<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1992) examine growth as driven by industrial innovati<strong>on</strong>s that improve the<br />
quality of products.<br />
Vertical innovati<strong>on</strong> models of this kind introduce, relative to models with exp<strong>and</strong>ing varieties of<br />
products, the c<strong>on</strong>siderati<strong>on</strong> of the process of creative destructi<strong>on</strong>: a new innovati<strong>on</strong> makes, to<br />
some extent, old products obsolete <strong>and</strong> negatively affects the ec<strong>on</strong>omic rents of previous innovators.<br />
For this reas<strong>on</strong>, models with qualitative improvements are also called Schumpeterian because they<br />
embody Schumpeter’s idea of creative destructi<strong>on</strong> by emphasizing the process by which new<br />
products (<strong>and</strong> innovators) displace old products (<strong>and</strong> innovators).<br />
Endogenous growth models with vertical innovati<strong>on</strong>s have very elaborated micro-foundati<strong>on</strong>s<br />
describing the nature of market dynamics, which are similar to <strong>and</strong> develop further earlier gametheoretical<br />
models of innovati<strong>on</strong>. As such these models have developed the study of dynamic<br />
competiti<strong>on</strong>.<br />
28<br />
246
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1992) study a model of ec<strong>on</strong>omic growth based <strong>on</strong> creative destructi<strong>on</strong>.<br />
Innovati<strong>on</strong> c<strong>on</strong>sists of a higher quality intermediate good that can be used to produce output<br />
more efficiently than before. Research firms invest resources in R&D activities that stochastically<br />
determine the time of the new innovati<strong>on</strong>; the arrival rate of new innovati<strong>on</strong>s depends <strong>on</strong>ly up<strong>on</strong><br />
the current flow of input to research. Firms’ R&D investments are motivated by the prospect of<br />
m<strong>on</strong>opoly rents that can be captured when a successful innovati<strong>on</strong> is patented. However, unlike<br />
the horiz<strong>on</strong>tal innovati<strong>on</strong>s c<strong>on</strong>sidered in Romer (1990), these rents would last <strong>on</strong>ly until the next<br />
innovati<strong>on</strong> occurs, when the current innovati<strong>on</strong> would become obsolete.<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt c<strong>on</strong>sider the cases of both drastic <strong>and</strong> n<strong>on</strong>-drastic innovati<strong>on</strong>. In the first case<br />
the new innovator is unc<strong>on</strong>strained by potential competiti<strong>on</strong> from the previous patent; in the latter<br />
case, such a c<strong>on</strong>straint is binding.<br />
In the case of drastic innovati<strong>on</strong>s market dynamics take the form of a c<strong>on</strong>tinuous process of<br />
acti<strong>on</strong> <strong>and</strong> reacti<strong>on</strong> whereby at each point in time the market is dominated by a m<strong>on</strong>opolist whose<br />
incumbency is short lived. In theory, a new innovati<strong>on</strong> could be introduced by either the current<br />
m<strong>on</strong>opolist or an outside research firm. The value of innovati<strong>on</strong> is the expected incremental<br />
present value of the flow of m<strong>on</strong>opoly profits generated by the innovati<strong>on</strong> over an interval whose<br />
length depends <strong>on</strong> the occurrence of the next innovati<strong>on</strong>. Because this incremental profit is lower<br />
for an incumbent m<strong>on</strong>opolist, <strong>on</strong>ly outside research firms would innovate <strong>and</strong> the m<strong>on</strong>opolist<br />
chooses to do no research.<br />
The reas<strong>on</strong> for this result lies in the fact that, given the assumpti<strong>on</strong> <strong>and</strong> the structure of the model,<br />
firms’ incentives are driven <strong>on</strong>ly by the profit incentive, which in the specific case is always larger<br />
for an external firm. This is the replacement effect emphasized by Arrow (1962) whereby a<br />
m<strong>on</strong>opolist would have lower incremental profits from innovati<strong>on</strong> than a perfectly competitive firm.<br />
In the case of drastic innovati<strong>on</strong> the competitive threat does not play a role because the flow of<br />
profit is independent of the identity of the innovator.<br />
Innovati<strong>on</strong>s are n<strong>on</strong>-drastic if the previous incumbent could make a positive profit when the<br />
current <strong>on</strong>e is charging the unc<strong>on</strong>strained optimal m<strong>on</strong>opolistic price. In the case of n<strong>on</strong>-drastic<br />
innovati<strong>on</strong> the competitive threat may potentially play a role in shaping equilibrium investments in<br />
R&D activities. However, Aghi<strong>on</strong> <strong>and</strong> Howitt assume that the m<strong>on</strong>opolist chooses to do no<br />
research also in the case of n<strong>on</strong>-drastic innovati<strong>on</strong>s, i.e. assume that the efficiency effect is<br />
smaller than the replacement effect. As a result market dynamics take the form of c<strong>on</strong>tinuous<br />
acti<strong>on</strong>/reacti<strong>on</strong>.<br />
Finally, Aghi<strong>on</strong> <strong>and</strong> Howitt c<strong>on</strong>sider the case in which firms can affect not <strong>on</strong>ly the frequency but<br />
also the size of innovati<strong>on</strong>s. Their finding is that innovati<strong>on</strong>s would be too small if they were<br />
drastic; in the n<strong>on</strong>-drastic case, the tendency to make innovati<strong>on</strong>s too small is in part mitigated by<br />
the incentive for innovators to move away from their competitive fringe.<br />
Grossman <strong>and</strong> Helpman (1991) have built <strong>on</strong> the model by Aghi<strong>on</strong> <strong>and</strong> Howitt to c<strong>on</strong>sider an<br />
ec<strong>on</strong>omy where a c<strong>on</strong>tinuum of, <strong>and</strong> not <strong>on</strong>ly <strong>on</strong>e, intermediate goods. Each product has its own<br />
quality ladder <strong>and</strong> entrepreneurs target individual products <strong>and</strong> race to bring about the next<br />
generati<strong>on</strong>. In each industry, success occurs with a probability per unit of time that is proporti<strong>on</strong>al<br />
29<br />
247
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
to the total R&D resources invested to improving that product. The authors allow free entry in the<br />
race for the next generati<strong>on</strong> of products <strong>and</strong> assume that potential entrants can learn enough<br />
about the state of knowledge to compete to produce the new state-of-the-art quality. These<br />
assumpti<strong>on</strong>s imply that, without a cost advantage, industry leaders do not invest resources to<br />
improve the quality of their own state-of-the-art products. This is because of the replacement<br />
effect whereby the leader would obtain an incremental flow of profits that is less than that of an<br />
external firm.<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) introduce an element of heterogeneity in innovative activity that<br />
captures the distincti<strong>on</strong> between fundamental <strong>and</strong> sec<strong>on</strong>dary research. They first investigate this<br />
heterogeneity by c<strong>on</strong>sidering fundamental research as deriving from R&D activities <strong>and</strong><br />
sec<strong>on</strong>dary innovati<strong>on</strong> from learning-by-doing. Each innovati<strong>on</strong> resulting from research c<strong>on</strong>sists of<br />
a potential new product, <strong>and</strong> learning-by-doing leads to improvements of the quality of goods that<br />
have been invented. They also c<strong>on</strong>sider the possibility that learning-by-doing c<strong>on</strong>tributed to a<br />
general stock of knowledge that creates new opportunities for research as well. Moreover,<br />
Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) c<strong>on</strong>sider both the case where learning is not appropriated by the firm<br />
but is shared by all firms <strong>and</strong> where the quality enhancement of learning is fully internalized.<br />
Finally, they c<strong>on</strong>sider the case where the distincti<strong>on</strong> between fundamental <strong>and</strong> sec<strong>on</strong>dary<br />
innovati<strong>on</strong>s is captured in terms of research versus development.<br />
A feature of Aghi<strong>on</strong> <strong>and</strong> Howitt’s (1992) influential model, shared by most early models of<br />
endogenous growth with quality ladders <strong>and</strong> leapfrogging, is that more competiti<strong>on</strong> (i.e. higher<br />
elasticity of dem<strong>and</strong> for each firm) is associated with less innovati<strong>on</strong> <strong>and</strong> growth, a very<br />
Schumpeterian result. The reas<strong>on</strong> for this c<strong>on</strong>clusi<strong>on</strong> is that these models capture <strong>on</strong>ly an<br />
increase of ex post product market competiti<strong>on</strong> <strong>and</strong> not of ex ante competiti<strong>on</strong> as well. Some<br />
recent studies modify this feature of earlier models <strong>and</strong> reach a different c<strong>on</strong>clusi<strong>on</strong> <strong>on</strong> the<br />
relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>: tougher competiti<strong>on</strong> may increase the<br />
incentives of firms to innovate in order to escape from a more competitive market state.<br />
In the model developed by Aghi<strong>on</strong>, Harris <strong>and</strong> Vickers (1997) <strong>and</strong> subsequently analysed by<br />
Aghi<strong>on</strong> et al. (2001) <strong>and</strong> Aghi<strong>on</strong> et al. (2002), the study of the relati<strong>on</strong>ship between product<br />
market competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> is carried out in the c<strong>on</strong>text of a model of endogenous growth<br />
with vertical differentiati<strong>on</strong> in a duopolistic industry where the incumbent firm can innovate.<br />
The model c<strong>on</strong>siders a sequence of tournaments with step-by-step innovati<strong>on</strong>s: the technological<br />
follower has first to catch up with the leading-edge technology before being able to become the<br />
new technological leader. This assumpti<strong>on</strong> can be justified <strong>on</strong> the grounds that innovative<br />
knowledge is to some extent tacit so that a rival has to invest in its own R&D to catch up <strong>and</strong><br />
cannot simply rely <strong>on</strong> knowledge spillovers.<br />
Aghi<strong>on</strong>, Harris <strong>and</strong> Vickers (1997) focus <strong>on</strong> the comparis<strong>on</strong> between Cournot <strong>and</strong> Bertr<strong>and</strong><br />
competiti<strong>on</strong> to capture different levels of market competiti<strong>on</strong>. Aghi<strong>on</strong> et al. (2001) <strong>and</strong> Aghi<strong>on</strong> et<br />
al. (2002) use the same model to investigate the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong><br />
where the latter is captured in terms of a substitutability parameter at the product market<br />
competiti<strong>on</strong> level.<br />
30<br />
248
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
The substantial difference of these models compared to earlier Schumpeterian models lies in the<br />
fact that, by c<strong>on</strong>sidering the possibility that the incumbent innovates, it is the incremental profits of<br />
each firm that drive R&D investments <strong>and</strong> the impact of product market competiti<strong>on</strong> has to be<br />
evaluated both ex ante <strong>and</strong> ex post. Tougher product market competiti<strong>on</strong> might reduce a firm’s<br />
pre-innovati<strong>on</strong> rents by more than it reduces post-innovati<strong>on</strong> rents, hence it may increase R&D<br />
firms’ investments.<br />
In these models, technological progress does not involve leapfrogging. Hence, a typical industry<br />
can be in either of two states: it can be a leveled industry where firms compete neck-to-neck or<br />
an unleveled industry where <strong>on</strong>e firm is technologically ahead of the follower.<br />
The effect of an increase in product market competiti<strong>on</strong> would have to be evaluated c<strong>on</strong>sidering<br />
the impact in both leveled <strong>and</strong> unleveled industries. The impact is actually twofold: <strong>on</strong> the <strong>on</strong>e<br />
h<strong>and</strong> the degree of product market competiti<strong>on</strong> affects equilibrium R&D investments in each of<br />
the possible state of the industry (leveled <strong>and</strong> unleveled); additi<strong>on</strong>ally, product market<br />
competiti<strong>on</strong> has an impact <strong>on</strong> how often the industry will be in <strong>on</strong>e of either states.<br />
In a leveled industry increased competiti<strong>on</strong> would spur innovati<strong>on</strong> because it would increase the<br />
incentives of firms’ to get ahead of the rival (the new “escape competiti<strong>on</strong> effect”); in an unleveled<br />
industry, increased competiti<strong>on</strong> would reduce the incentives to innovate for the classic<br />
Schumpeterian effect.<br />
The reas<strong>on</strong>ing in Aghi<strong>on</strong> et al. (2002), which leads to the c<strong>on</strong>clusi<strong>on</strong> that the relati<strong>on</strong>ship between<br />
intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> has an inverted-U shape, is discussed below.<br />
C<strong>on</strong>sider the case of very large innovati<strong>on</strong>s, in which case the leader never innovates <strong>and</strong> the<br />
largest gap between the leader <strong>and</strong> the laggard is <strong>on</strong>e technological step.<br />
When there is not much product market competiti<strong>on</strong>, there is hardly any incentive for firms to<br />
innovate when they compete neck-to-neck, <strong>and</strong> the overall innovati<strong>on</strong> rate will be highest when<br />
the industry is in an asymmetric state. Thus the industry will spend relatively more time in the<br />
levelled state where the escape-competiti<strong>on</strong> effect dominates. An increase in product market<br />
competiti<strong>on</strong> would then result in larger incentives for the firms to innovate. Hence, if the degree of<br />
competiti<strong>on</strong> is initially low, an increase will result in a faster innovati<strong>on</strong> rate.<br />
When initial product market competiti<strong>on</strong> is high, there is relatively little incentive for the laggard<br />
firm in an unleveled state to innovate <strong>and</strong> a relatively large incentive for a neck-to-neck firm to<br />
leave the leveled state. The c<strong>on</strong>sequence is that the industry will spend most of the time in the<br />
unleveled state where it is the Schumpeterian effect that dominates. Tougher product market<br />
competiti<strong>on</strong> reduces the Schumpeterian effect so that when the degree of competiti<strong>on</strong> is initially<br />
high, an increase may result in a slower average innovati<strong>on</strong> rate.<br />
These results extend to the general model with no upper bound to the technological gap.<br />
Aghi<strong>on</strong> et al. (2002) also provide empirical evidence that support their claim of an inverted-U<br />
relati<strong>on</strong>ship between intensity of competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>.<br />
31<br />
249
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Aghi<strong>on</strong> et al. (2001) use the same framework to investigate the relati<strong>on</strong>ship between imitati<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong> <strong>and</strong> c<strong>on</strong>clude that this is also characterized by an inverted-U shape. Again this may<br />
be the case because easier imitati<strong>on</strong> may decrease a firm’s pre-innovati<strong>on</strong> rents by more than it<br />
reduces post-innovati<strong>on</strong> rent.<br />
Denicolò <strong>and</strong> Zanchettin (2002) c<strong>on</strong>sider the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth in the<br />
case of n<strong>on</strong>-drastic innovati<strong>on</strong>s by comparing innovati<strong>on</strong> <strong>and</strong> growth under Bertr<strong>and</strong> <strong>and</strong> Cournot<br />
competiti<strong>on</strong> in a st<strong>and</strong>ard leapfrogging model. The main finding of this study is that:<br />
“[…] when the size of innovati<strong>on</strong>s is sufficiently large, the equilibrium rate of growth is<br />
unambiguously greater with Bertr<strong>and</strong> than with Cournot competiti<strong>on</strong>. For smaller<br />
innovati<strong>on</strong>, Cournot competiti<strong>on</strong> may (but need not) create greater incentive to innovate.<br />
The intuiti<strong>on</strong> is that more competiti<strong>on</strong> entails lower prices but at the same time leads to greater<br />
productive efficiency (hence greater rates of growth) in that it lowers the market shares of less<br />
productive firms (Aghi<strong>on</strong> <strong>and</strong> Shankerman, 2000). When firms are symmetric <strong>on</strong>ly the first effect is<br />
at work. When technical progress creates asymmetries between firms the productive efficiency effect<br />
becomes important <strong>and</strong> the effect of more competiti<strong>on</strong> <strong>on</strong> innovators’ profits is generally ambiguous.<br />
However, the productive efficiency effect must dominate if innovati<strong>on</strong>s are close to being drastic.<br />
D’Asprem<strong>on</strong>t et al. (2002) investigate the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> in a<br />
tournament model under uncertainty where, unlike Aghi<strong>on</strong> et al. (2001), the number of firms in<br />
each industry is endogenously determined <strong>and</strong> the race entails the possibility of several<br />
simultaneous winners <strong>and</strong> spillover effects. The main c<strong>on</strong>clusi<strong>on</strong> of the paper is that the<br />
relati<strong>on</strong>ship between toughness of market competiti<strong>on</strong> <strong>and</strong> the incentives to innovate is n<strong>on</strong>m<strong>on</strong>ot<strong>on</strong>e.<br />
The key for this result lies in the possible multiplicity of successful innovators in the<br />
same period.<br />
Introducti<strong>on</strong> of Agency C<strong>on</strong>siderati<strong>on</strong>s<br />
The adopti<strong>on</strong> of a step-by-step technological assumpti<strong>on</strong> is not the <strong>on</strong>ly means by which the<br />
relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth takes a different nature that in the traditi<strong>on</strong>al<br />
Schumpeterian model. Aghi<strong>on</strong> <strong>and</strong> Howitt (1998) c<strong>on</strong>sider also an alternative approach, based<br />
<strong>on</strong> a study published later by Aghi<strong>on</strong>, Dewatrip<strong>on</strong>t <strong>and</strong> Rey (1999) that leads to c<strong>on</strong>clusi<strong>on</strong>s <strong>on</strong><br />
the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong> growth that are different from those of the classic<br />
Schumpeterian model.<br />
This approach is based <strong>on</strong> relaxing the traditi<strong>on</strong>al behavioural assumpti<strong>on</strong> that firms are profit<br />
maximizing. N<strong>on</strong>-profit maximizing behavior is a comm<strong>on</strong> assumpti<strong>on</strong> in the literature that<br />
emphasises the existence of agency problems between producers <strong>and</strong> their financiers. Aghi<strong>on</strong> et<br />
al (1997) c<strong>on</strong>sider a principal-agent model of the firms where managers are interested in<br />
maximizing some private benefits of c<strong>on</strong>trol while minimizing private costs that innovati<strong>on</strong> would<br />
entail such as the costs of reorganizing the firm or training costs.<br />
32<br />
250
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Within this new setting, the impact of intensity of competiti<strong>on</strong> <strong>on</strong> growth can be positive because<br />
tougher competiti<strong>on</strong> may lead to managers to speed up the adopti<strong>on</strong> of new technologies to avoid<br />
bankruptcy <strong>and</strong> the loss of c<strong>on</strong>trol rights <strong>and</strong> hence to faster ec<strong>on</strong>omic growth.<br />
The literature <strong>on</strong> the importance of competiti<strong>on</strong> for incentives is extensive <strong>and</strong> it suggests that<br />
there are other important mechanisms though which competiti<strong>on</strong> may have a positive impact <strong>on</strong><br />
incentives is through the improvement of the ability of assessing managers’ efforts <strong>and</strong> hence of<br />
m<strong>on</strong>itoring them. At the same time, similarly to other areas of research, theoretical predicti<strong>on</strong>s of<br />
these studies are not as immediate as these simple c<strong>on</strong>siderati<strong>on</strong>s may suggest.<br />
Nickell (1996) provides empirical evidence <strong>on</strong> the relati<strong>on</strong>ship between intensity of competiti<strong>on</strong><br />
<strong>and</strong> innovati<strong>on</strong> by studying the impact of competiti<strong>on</strong> <strong>on</strong> both the level <strong>and</strong> the growth of total<br />
factor productivity of a large number of UK manufacturing companies. The main findings of this<br />
paper are that<br />
•<br />
•<br />
market power, which is captured in terms of market share, generates lower levels of<br />
productivity; <strong>and</strong> that<br />
competiti<strong>on</strong>, which is captured by the number of competitors or by lower levels or rents,<br />
leads to faster total factor productivity growth.<br />
Blundell, Griffith <strong>and</strong> van Reenen (1999) also study the relati<strong>on</strong>ship between competiti<strong>on</strong> <strong>and</strong><br />
innovati<strong>on</strong>. Their study provides evidence that<br />
“[…] within industries in was the high market share firms who tended to<br />
commercializemore innovati<strong>on</strong> although increased product market competiti<strong>on</strong> in the<br />
industry tended to<br />
stimulate innovative activity” (p. 550)<br />
For a very comprehensive survey of empirical studies <strong>on</strong> competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> see the<br />
recent OECD study by Ahn (2002).<br />
1.4 New Ec<strong>on</strong>omic Geography<br />
New ec<strong>on</strong>omic geography was effectively initiated by Krugman (1991). The main goal of this<br />
str<strong>and</strong> of research is to provide a theoretical explanati<strong>on</strong> of why ec<strong>on</strong>omic activity <strong>and</strong> populati<strong>on</strong><br />
tend to c<strong>on</strong>centrate, that is to say, why cities <strong>and</strong> industrial belts come into existence <strong>and</strong> keep <strong>on</strong><br />
exp<strong>and</strong>ing at the expense of the periphery. This approach is similar to new growth theory in that<br />
the models assume imperfect competiti<strong>on</strong> <strong>and</strong> product differentiati<strong>on</strong>, ec<strong>on</strong>omies of scale <strong>and</strong><br />
increasing returns.<br />
33<br />
251
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The underlying story is that manufactured goods will be produced in regi<strong>on</strong>s where nearby<br />
dem<strong>and</strong> is higher, as this allows firms to minimize transport costs. Dem<strong>and</strong> will come either from<br />
agriculture or from the manufacturing sector itself (manufacturing workers). Because of a<br />
possible positive feedback between producti<strong>on</strong> of <strong>and</strong> dem<strong>and</strong> for manufactured goods,<br />
manufacturing producti<strong>on</strong> will tend to relocate into these regi<strong>on</strong>s.<br />
This phenomen<strong>on</strong> is formalized by the can<strong>on</strong>ical modelling of core <strong>and</strong> periphery (Krugman<br />
(1991)) that assumes that the ec<strong>on</strong>omy is split into two sectors - namely manufacturing <strong>and</strong><br />
agriculture – <strong>and</strong> into two regi<strong>on</strong>s (the core <strong>and</strong> periphery). In manufacturing, m<strong>on</strong>opolistic<br />
competiti<strong>on</strong> à la Dixit-Stiglitz (1977) prevails: manufacturing produces differentiated products in<br />
the presence of increasing returns. In additi<strong>on</strong>, labor, specific to manufacturing, is mobile<br />
between the two regi<strong>on</strong>s. Wages, both in nominal <strong>and</strong> real terms, can differ between the two<br />
regi<strong>on</strong>s. The driving force behind labor mobility is indeed differences in real wages.<br />
Transportati<strong>on</strong> costs when shipping from <strong>on</strong>e regi<strong>on</strong> to the other occur in the form of Samuels<strong>on</strong>’s<br />
iceberg cost (Samuels<strong>on</strong> (1954)). This means that transport costs are not explicitly c<strong>on</strong>sidered<br />
but rather it is assumed that <strong>on</strong>ly a fracti<strong>on</strong> of goods arrives, with the difference “smelting down”<br />
during transportati<strong>on</strong>.<br />
By c<strong>on</strong>trast, agriculture is characterized by perfect competiti<strong>on</strong> with homogenous goods <strong>and</strong><br />
c<strong>on</strong>stant returns to scale. Wages equal prices <strong>and</strong> labor, specific to agriculture, is immobile<br />
between the two regi<strong>on</strong>s. Furthermore, no transportati<strong>on</strong> costs are assumed to occur when<br />
goods are moved from <strong>on</strong>e regi<strong>on</strong> to the other. Assuming c<strong>on</strong>stant returns to scale <strong>and</strong> the<br />
absence of transport costs secures wage equalizati<strong>on</strong> in agriculture. It should be also noted that<br />
agriculture is evenly divided between the two regi<strong>on</strong>s.<br />
For c<strong>on</strong>centrati<strong>on</strong> in manufacturing to take place in <strong>on</strong>e of the regi<strong>on</strong>s, several c<strong>on</strong>diti<strong>on</strong>s have to<br />
be fulfilled. First, a large share of populati<strong>on</strong> should be allocated to the manufacturing sector so<br />
as it generates large dem<strong>and</strong> (that triggers the virtuous circle). Sec<strong>on</strong>d, ec<strong>on</strong>omies of scale<br />
should be high enough <strong>and</strong> transportati<strong>on</strong> costs low.<br />
Three cases can be c<strong>on</strong>sidered.<br />
1 First, in the event that transport costs are large, there is an equal divisi<strong>on</strong> of<br />
manufacturing between the two regi<strong>on</strong>s.<br />
2<br />
3<br />
If transportati<strong>on</strong> costs are low, manufacturing still can remain equally divided between<br />
regi<strong>on</strong>s. However, this situati<strong>on</strong> is not a stable <strong>on</strong>e (unstable equilibrium) for the following<br />
reas<strong>on</strong>. If manufacturing were <strong>on</strong>ly slightly more important in <strong>on</strong>e of the regi<strong>on</strong>s, this<br />
would lead to an increasing share of manufacturing in this sector accompanied by a<br />
dwindling of manufacturing in the other regi<strong>on</strong>. At the end of this process, manufacturing<br />
ends up being c<strong>on</strong>centrated in <strong>on</strong>e of the regi<strong>on</strong>s, <strong>and</strong> this can be referred to as the coreperiphery<br />
pattern.<br />
The third case, characterized by intermediate transportati<strong>on</strong> costs, can lead to either<br />
complete specializati<strong>on</strong> or equal divisi<strong>on</strong> between regi<strong>on</strong>s.<br />
34<br />
252
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Extensi<strong>on</strong>s<br />
The three-regi<strong>on</strong> model<br />
A possible extensi<strong>on</strong> of the basic model is to simultaneously c<strong>on</strong>sider three regi<strong>on</strong>s. Keeping the<br />
assumpti<strong>on</strong>s of the two-regi<strong>on</strong> model would yield basically the same results: No c<strong>on</strong>centrati<strong>on</strong> of<br />
manufacturing activity in the presence of high transport costs, extreme c<strong>on</strong>centrati<strong>on</strong> in <strong>on</strong>e of the<br />
three regi<strong>on</strong>s if transport costs are low, <strong>and</strong> the combinati<strong>on</strong> of the two previous situati<strong>on</strong>s in the<br />
case of intermediate transport costs.<br />
Transport costs in agriculture<br />
The sec<strong>on</strong>d extensi<strong>on</strong> of the basic two-regi<strong>on</strong> model c<strong>on</strong>sists in releasing the assumpti<strong>on</strong> related<br />
to the absence of transport costs in agriculture. Empirical evidence clearly indicates that transport<br />
costs exist not <strong>on</strong>ly for differentiated goods produced in the manufacturing sector, but also in the<br />
other sectors, <strong>and</strong> that these costs are likely to be at least as high as the cost for manufacturing<br />
products. For this reas<strong>on</strong>, iceberg-type transport costs are allowed for in agriculture. As a<br />
corollary, wages <strong>and</strong> c<strong>on</strong>sequently prices of agricultural products do not equalise between the two<br />
sectors.<br />
The introducti<strong>on</strong> of transportati<strong>on</strong> costs for agricultural products when shipped from <strong>on</strong>e regi<strong>on</strong> to<br />
the other results in higher agricultural prices in the destinati<strong>on</strong> regi<strong>on</strong> (with highly c<strong>on</strong>centrated<br />
manufacturing at the expenses of the other regi<strong>on</strong>). Hence, higher prices for agricultural<br />
products, translated into higher costs of living would actually discourage labour mobility between<br />
the two regi<strong>on</strong>s, <strong>and</strong> thus hinder the c<strong>on</strong>centrati<strong>on</strong> of manufacturing activity.<br />
It seems very unrealistic to assume that the same homogenous agricultural good is produced in<br />
the two regi<strong>on</strong>s. The next step is therefore to c<strong>on</strong>sider differentiated agricultural products. The<br />
results are similar to the previous <strong>on</strong>es. In general, it is possible to reach the c<strong>on</strong>clusi<strong>on</strong> that<br />
while high transport costs in agriculture tend to prohibit c<strong>on</strong>centrati<strong>on</strong> in manufacturing, a<br />
decrease in costs leads to c<strong>on</strong>centrati<strong>on</strong>.<br />
Urban agglomerati<strong>on</strong>s<br />
A sec<strong>on</strong>d str<strong>and</strong> of the new ec<strong>on</strong>omic geography is c<strong>on</strong>cerned with urban agglomerati<strong>on</strong>s <strong>and</strong><br />
attempts to answer questi<strong>on</strong>s as follows:<br />
•<br />
•<br />
•<br />
•<br />
How do cities come into existence <strong>and</strong> why do they exist in the l<strong>on</strong>ger run?<br />
How do new cities form <strong>and</strong> which are the mechanisms underlying the growth of cities?<br />
What is the reas<strong>on</strong> that cities turn out to be linked <strong>on</strong>e to another through a hierarchy?<br />
What is the role of natural advantages, such as harbors, in the development of cities?<br />
35<br />
253
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Internati<strong>on</strong>al trade<br />
New trade theory, as developed by Krugman (1980), has much in comm<strong>on</strong> with new ec<strong>on</strong>omic<br />
geography. The former aims at investigating internati<strong>on</strong>al specializati<strong>on</strong> phenomena in the case<br />
of two or more countries, which can be viewed as the equivalent of agglomerati<strong>on</strong> <strong>and</strong><br />
c<strong>on</strong>centrati<strong>on</strong> within an individual country. It should be noted that the logic <strong>and</strong> the mechanisms<br />
at work are very similar <strong>and</strong> what makes the difference are different restricti<strong>on</strong>s to the basic model<br />
as discussed below.<br />
Fujita et al. (2001) develops a model with intermediate goods. In this two-country framework,<br />
each individual country is composed of two sectors, namely manufacturing <strong>and</strong> agriculture with<br />
labor being assumed to be perfectly mobile between agriculture <strong>and</strong> the manufacturing sector.<br />
By c<strong>on</strong>trast, labor is assumed to be immobile between the two countries. It is worth drawing<br />
attenti<strong>on</strong> to the fact that this model corresp<strong>on</strong>ds exactly to the can<strong>on</strong>ical model with the difference<br />
that two countries instead of two regi<strong>on</strong>s are c<strong>on</strong>sidered. There is no labor mobility between the<br />
two countries (regi<strong>on</strong>s) <strong>and</strong> intermediate goods are added to the model.<br />
The introducti<strong>on</strong> of intermediary goods is normally associated with input-output matrices. Put<br />
differently, firms are assumed to be linked with each other through input <strong>and</strong> output linkages. The<br />
model avoids the inclusi<strong>on</strong> of new sectors by assuming that the range of goods produced in<br />
manufacturing serves not <strong>on</strong>ly for final c<strong>on</strong>sumpti<strong>on</strong> but also as input for manufacturing itself.<br />
As a result, manufacturing uses labor <strong>and</strong> intermediary goods as inputs. In accordance with the<br />
model, the greater the gamut of intermediary goods, the lower the costs of producti<strong>on</strong>. In additi<strong>on</strong><br />
to that, if those goods are available locally, prices do not include transport costs. As far as sales<br />
are c<strong>on</strong>cerned, what firms produce is sold either to c<strong>on</strong>sumer as final goods or to manufacturing<br />
firms as intermediate goods.<br />
On the <strong>on</strong>e h<strong>and</strong>, high trade costs leave manufacturing equally divided between the two countries<br />
because firms have to supply c<strong>on</strong>sumers. On the other h<strong>and</strong>, low trade costs translate into the<br />
full c<strong>on</strong>centrati<strong>on</strong> of manufacturing activity in <strong>on</strong>e of the countries <strong>on</strong> the grounds of backward<br />
<strong>and</strong> forward linkages. In the presence of intermediate trade costs, c<strong>on</strong>centrati<strong>on</strong> is possible but<br />
not inevitable<br />
This model also manages to give an explanati<strong>on</strong> for differing nominal <strong>and</strong> real wage levels in<br />
different countries. According to the model, the real wage in the country where manufacturing<br />
c<strong>on</strong>centrates rises relative to those in the other country. One reas<strong>on</strong> for this is that dem<strong>and</strong> for<br />
labor increases <strong>and</strong> so does the wage level as a c<strong>on</strong>sequence of c<strong>on</strong>centrati<strong>on</strong> in the first<br />
country. Another reas<strong>on</strong> is that the country without manufacturing producti<strong>on</strong> has to pay the bill of<br />
transport costs of the imported manufactured goods, which in turn decreases the real wage.<br />
36<br />
254
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Ec<strong>on</strong>omic development: a sequential process<br />
In this stage of the analysis, an exogenous growth process is added to the model <strong>and</strong> it is<br />
assumed that dem<strong>and</strong> for manufactured goods rises faster than potential supply. A model<br />
c<strong>on</strong>sisting of two countries is c<strong>on</strong>sidered. The model seeks to explain under which c<strong>on</strong>diti<strong>on</strong>s<br />
industrial c<strong>on</strong>centrati<strong>on</strong> starts spilling out from <strong>on</strong>e country to the other.<br />
The point of departure is to c<strong>on</strong>sider <strong>on</strong>e country with highly c<strong>on</strong>centrated manufacturing. It is<br />
shown that c<strong>on</strong>centrati<strong>on</strong> will c<strong>on</strong>tinue until the gains of backward <strong>and</strong> forward linkages in the<br />
industry exceed the potential loss that could be associated with the fact that wages are<br />
c<strong>on</strong>siderable higher in the industrialized country. If the losses are higher than the gains (that is<br />
the wage gap exceeds a given threshold) firms start moving into the other country. These firms<br />
then create their inter-firm linkages in the other country, which, through a positive feedback<br />
process will attract other firms <strong>and</strong> thus accelerate the c<strong>on</strong>centrati<strong>on</strong> process.<br />
However, there are some interesting questi<strong>on</strong>s to answer. For instance, which industry is the first<br />
to move, <strong>and</strong> how does the industrial structure of countries at different stages of industrializati<strong>on</strong><br />
change? This is studied using a five-country model in which each country is composed of seven<br />
different industries. According to the model, the most labor intensive industry will be the first to<br />
move into the other country, followed progressively by less-labor intensive industries. However,<br />
the latter will move faster.<br />
The authors suggest that the inter-industry structure of the ec<strong>on</strong>omies changes depending <strong>on</strong><br />
several factors, including:<br />
•<br />
•<br />
the higher the share of final c<strong>on</strong>sumpti<strong>on</strong> goods in sales, the higher the probability that<br />
firms move (forward linkage); <strong>and</strong><br />
the lower the intermediate input requirement, the quicker firm moves (backward linkage).<br />
Industrial clustering<br />
Fujita et al. (2001) c<strong>on</strong>struct a two-country, two-industry model with a single factor of producti<strong>on</strong>.<br />
Agriculture is dropped in the analysis <strong>and</strong> two manufacturing industries are used instead, both<br />
m<strong>on</strong>opolistically competitive. The questi<strong>on</strong> is: under which circumstances do c<strong>on</strong>centrati<strong>on</strong> or<br />
dispersi<strong>on</strong> occur?<br />
The model provides us with two c<strong>on</strong>clusi<strong>on</strong>s. First, in the event that <strong>on</strong>e industry is str<strong>on</strong>gly<br />
c<strong>on</strong>nected to the other industry through input-output linkages, c<strong>on</strong>centrati<strong>on</strong> will not happen. In<br />
c<strong>on</strong>trast, in the case of str<strong>on</strong>g intra-industry links, the model predicts a high c<strong>on</strong>centrati<strong>on</strong>, that is,<br />
each country specializes in <strong>on</strong>e of the industrial activities. When c<strong>on</strong>sidering trade costs, the<br />
results are similar to the previous <strong>on</strong>es: the two industries remain present in both countries if<br />
trade costs are high, but a dwindling in trade costs allows for <strong>and</strong> subsequently leads to<br />
specializati<strong>on</strong> of the countries.<br />
37<br />
255
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The authors suggest that with deeper ec<strong>on</strong>omic integrati<strong>on</strong> bringing about lower trade costs,<br />
European ec<strong>on</strong>omies are expected to undergo a c<strong>on</strong>centrati<strong>on</strong> process <strong>and</strong> will c<strong>on</strong>verge<br />
towards the ec<strong>on</strong>omic structure of the US.<br />
The two-country, two-industry framework is then replaced by a model including more than two<br />
industrial sectors. Although the model cannot predict why <strong>on</strong>e industry c<strong>on</strong>centrates in <strong>on</strong>e<br />
country <strong>and</strong> not another, it can provide a theoretical underpinning in favor of the argument that in<br />
case <strong>on</strong>e country loses a particular industry as a result of some sort of exogenous shock, there<br />
exists no mechanism that could ensure that, after the shock, the industry in questi<strong>on</strong> regains<br />
ground in the country under c<strong>on</strong>siderati<strong>on</strong>.<br />
Empirical works related to industrial clustering show that the US experienced a high c<strong>on</strong>centrati<strong>on</strong><br />
of ec<strong>on</strong>omic activities <strong>and</strong> a high degree of industrial clustering chiefly during the period prior to<br />
WW-I. (see e.g. Kim (1995)). Patterns of c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong> industrial clustering remained<br />
relatively unchanged afterwards. Studies focusing <strong>on</strong> Europe indicate a str<strong>on</strong>g c<strong>on</strong>centrati<strong>on</strong> <strong>and</strong><br />
clustering which has occurred during the last 20 years or so. Even so, the level of c<strong>on</strong>centrati<strong>on</strong><br />
<strong>and</strong> sectoral clustering in Europe is not as high as in the US (e.g. Midelfart-Knarvik et al. (2000)).<br />
In a recent study with a str<strong>on</strong>g cliometric influence (see secti<strong>on</strong> 2), Davis <strong>and</strong> Weinstein (2002)<br />
set out to investigate agglomerati<strong>on</strong> patterns in Japan based <strong>on</strong> a unique data set stretching over<br />
8000 years. Data drawn from archaeological findings <strong>and</strong> populati<strong>on</strong> census indicate that<br />
dispersi<strong>on</strong> in populati<strong>on</strong> has been historically high in Japan. However, variati<strong>on</strong> in populati<strong>on</strong><br />
density c<strong>on</strong>siderably widened from the outset of the industrial revoluti<strong>on</strong>. The authors argue that<br />
although increasing returns can hardly explain the origins of geographical agglomerati<strong>on</strong>, they<br />
play a prominent role in explaining temporal dynamics, i.e. the accelerati<strong>on</strong> of geographical<br />
c<strong>on</strong>centrati<strong>on</strong> throughout the last 200 years. New light is also shed <strong>on</strong> the temporary nature of<br />
even very large shocks. For instance, the massive bombardment of Japanese cities at the end of<br />
WW-II does not appear to have permanently altered the spatial structure of the Japanese industrial<br />
structure.<br />
38<br />
256
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
2 THE EVOLUTIONARY ECONOMICS OF INNOVATION<br />
This secti<strong>on</strong> discusses the main findings from the evoluti<strong>on</strong>ary approach to the study of<br />
innovati<strong>on</strong> <strong>and</strong> technical change. It then discusses the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
instituti<strong>on</strong>s (known as “systems of innovati<strong>on</strong>”).<br />
2.1 Introducti<strong>on</strong><br />
The evoluti<strong>on</strong>ary approach to the study of innovati<strong>on</strong> <strong>and</strong> technological change arose largely<br />
because the mainstream theory was c<strong>on</strong>sidered inadequate to tackle the inherent disequilibrium<br />
dynamics that characterise technological innovati<strong>on</strong>. Nels<strong>on</strong> <strong>and</strong> Winter (1984) point out that<br />
both the ec<strong>on</strong>omy as a whole <strong>and</strong> the ec<strong>on</strong>omic actors taken individually can be c<strong>on</strong>sidered to<br />
evolve c<strong>on</strong>tinuously, similar to the evoluti<strong>on</strong>ary process described in biology. Hence the name of<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics.<br />
Dissatisfacti<strong>on</strong> with the traditi<strong>on</strong>al approach is almost pervasive <strong>and</strong> embraces the definiti<strong>on</strong> of<br />
innovati<strong>on</strong> <strong>and</strong> knowledge, the descripti<strong>on</strong> of the R&D process <strong>and</strong> more generally of the<br />
innovati<strong>on</strong> process, the modelling approach based <strong>on</strong> the noti<strong>on</strong> of full rati<strong>on</strong>ality <strong>and</strong> equilibrium,<br />
<strong>and</strong> the lack of c<strong>on</strong>siderati<strong>on</strong> of instituti<strong>on</strong>al factors.<br />
Though these themes trace back to earlier researchers — in particular, the Schumpeterian<br />
perspective, which places emphasis <strong>on</strong> dynamics <strong>and</strong> competiti<strong>on</strong> through innovati<strong>on</strong> — dynamic<br />
theorizing in evoluti<strong>on</strong>ary ec<strong>on</strong>omics is often seen to start with the models of Nels<strong>on</strong> <strong>and</strong> Winter<br />
(1982).<br />
Although evoluti<strong>on</strong>ary ec<strong>on</strong>omics is perhaps not a coherent theory with a comm<strong>on</strong> methodology,<br />
it may be categorised as a group of various approaches that share the same criticisms of the neoclassical<br />
approach. For example:<br />
• First, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat ec<strong>on</strong>omic agents as rati<strong>on</strong>al, in the sense<br />
they are in neo-classical ec<strong>on</strong>omics. They do not maximize returns through calculating<br />
the optimal strategy from a range of probabilistic outcomes.<br />
•<br />
Sec<strong>on</strong>d, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not seek to analyse ec<strong>on</strong>omic phenomena from<br />
equilibrium outcomes. In particular, dynamic change within the ec<strong>on</strong>omy is c<strong>on</strong>sidered<br />
best analyzed in terms of the processes that govern it, which are inherently disequilibrium<br />
dynamics.<br />
Evoluti<strong>on</strong>ary models draw their theoretical foundati<strong>on</strong>s <strong>on</strong> different sources <strong>and</strong> particularly <strong>on</strong><br />
biological models, evoluti<strong>on</strong>ary game theory <strong>and</strong> instituti<strong>on</strong>alism in order to capture essentially<br />
dynamic <strong>and</strong> irreversible processes.<br />
The literature tends to place emphasis <strong>on</strong> innovati<strong>on</strong> <strong>and</strong> selecti<strong>on</strong> from variety as the force of<br />
ec<strong>on</strong>omic development. Of particular importance is the need to see innovati<strong>on</strong> as more than<br />
simple technological development (<strong>and</strong> much more than the producti<strong>on</strong> of knowledge as a public<br />
good) <strong>and</strong> to analyze ec<strong>on</strong>omic development through technological <strong>and</strong> organizati<strong>on</strong>al<br />
improvements (which lead to the introducti<strong>on</strong> of new products, inputs <strong>and</strong> techniques).<br />
39<br />
257
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
In evoluti<strong>on</strong>ary theories innovati<strong>on</strong> is in fact defined more broadly than in mainstream models <strong>and</strong><br />
as a multidimensi<strong>on</strong>al phenomen<strong>on</strong>. Innovati<strong>on</strong> is not <strong>on</strong>ly a reducti<strong>on</strong> in producti<strong>on</strong> costs or the<br />
commercializati<strong>on</strong> of new products but also a more efficient reorganizati<strong>on</strong> of a firm’s activities,<br />
the adopti<strong>on</strong> of more sophisticated managerial techniques, access to new markets, etc.<br />
This richer definiti<strong>on</strong> emphasizes that innovati<strong>on</strong> extends outside the realm of technology in two<br />
ways: many innovati<strong>on</strong>s are not technological but involve other aspects of a firm’s activities such<br />
as its organizati<strong>on</strong>al structure; moreover, even technological innovati<strong>on</strong>s often affect n<strong>on</strong>technological<br />
activities of a firm: technological <strong>and</strong> organizati<strong>on</strong>al changes are often importantly<br />
intertwined.<br />
Another important difference with mainstream models of innovati<strong>on</strong> is that evoluti<strong>on</strong>ary models<br />
emphasize the tacit <strong>and</strong> specific nature of knowledge, as opposed to the idea of knowledge as<br />
informati<strong>on</strong> that somehow underlies mainstream models of innovati<strong>on</strong>.<br />
In terms of the process that leads to the discovery of new products, evoluti<strong>on</strong>ary models describe<br />
the importance of learning, <strong>and</strong> of organizati<strong>on</strong>al learning often embodied in organizati<strong>on</strong>al<br />
routines as a source of knowledge. The process through which innovati<strong>on</strong>s emerge is by no<br />
means linear, but it is characterized by complex feedback mechanisms <strong>and</strong> interactive relati<strong>on</strong>s<br />
involving science, technology, learning, producti<strong>on</strong>, instituti<strong>on</strong>s, organizati<strong>on</strong>, policy makers <strong>and</strong><br />
dem<strong>and</strong> (Edquist 1999).<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics can be looked at both from a microec<strong>on</strong>omic <strong>and</strong> macroec<strong>on</strong>omic<br />
perspective. The microec<strong>on</strong>omic aspects of evoluti<strong>on</strong>ary ec<strong>on</strong>omics are mainly c<strong>on</strong>cerned with<br />
how innovati<strong>on</strong> arises at the firm <strong>and</strong> market level. In c<strong>on</strong>trast, the macroec<strong>on</strong>omic str<strong>and</strong> of<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics places innovati<strong>on</strong> into the social <strong>and</strong> ec<strong>on</strong>omic c<strong>on</strong>text <strong>and</strong> examines the<br />
dynamic process through which the innovati<strong>on</strong> systems generate innovati<strong>on</strong>s through a complex<br />
<strong>and</strong> reciprocal relati<strong>on</strong>ship network c<strong>on</strong>necting different comp<strong>on</strong>ents of the system.<br />
2.2 Microec<strong>on</strong>omics of Innovati<strong>on</strong><br />
Knowledge <strong>and</strong> rati<strong>on</strong>ality<br />
At the heart of evoluti<strong>on</strong>ary ec<strong>on</strong>omics is a definiti<strong>on</strong> of knowledge <strong>and</strong> of its characteristics that is<br />
substantially different for the neo-classical assimilati<strong>on</strong> of innovati<strong>on</strong> to informati<strong>on</strong>. In each<br />
technology there are elements of tacit <strong>and</strong> specific knowledge that cannot be formalised <strong>and</strong><br />
cannot, therefore, be diffused in the form of informati<strong>on</strong>. These firm-specific, local <strong>and</strong> cumulative<br />
features of knowledge are, to various extents, essential characteristics of innovati<strong>on</strong>.<br />
The term tacit knowledge is used to encapsulate the idea of knowledge that is found to be useful<br />
(e.g. in an organizati<strong>on</strong>al routine or a skill) without being directly accessible to c<strong>on</strong>sciousness<br />
(Pavitt, 2002) or articulable (Winter, 1987). Thus individuals can be c<strong>on</strong>sidered not to underst<strong>and</strong><br />
all that they know — tacit knowledge cannot be fully explained, written down or otherwise<br />
diffused; yet it remains valuable. By c<strong>on</strong>trast the term informati<strong>on</strong> can be taken to mean ‘explicit’<br />
knowledge, which can be codified <strong>and</strong> transmitted.<br />
40<br />
258
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Informati<strong>on</strong>, as explicit knowledge, can be diffused around the ec<strong>on</strong>omy (<strong>and</strong> fits most closely<br />
with the neo-classical c<strong>on</strong>cept of knowledge as a public good). Tacit knowledge cannot be<br />
passed from agent to agent in this way, but it is transferred <strong>and</strong> transmitted implicitly. Firms, <strong>and</strong><br />
other ec<strong>on</strong>omic organizati<strong>on</strong>s, can be seen as vehicles through which tacit knowledge is<br />
replicated through routines <strong>and</strong> techniques (Foster, 2001).<br />
Furthermore, as menti<strong>on</strong>ed in secti<strong>on</strong> 1.2.5, drawing mutual benefits from tacit knowledge that<br />
cannot be explicitly transferred between firms may bring benefits to inter-firm co-operati<strong>on</strong> in<br />
innovative activity that complement those benefits identified in the industrial organizati<strong>on</strong><br />
literature.<br />
Support is found for this c<strong>on</strong>cept of tacit knowledge from psychological experiments. Nels<strong>on</strong> <strong>and</strong><br />
Winter (2002) cite studies that examine the difference between processes acquired through<br />
rati<strong>on</strong>ality (underst<strong>and</strong>ing the underlying nature of the problem <strong>and</strong> calculating strategies for<br />
success) <strong>and</strong> learning-by-doing (learning successful strategies through trial <strong>and</strong> error without a<br />
complete theory <strong>on</strong> why such strategies are successful). They note that “a close logical<br />
c<strong>on</strong>necti<strong>on</strong> between a learned task <strong>and</strong> a newly presented task does not necessarily indicate a<br />
potential for easy transfer”. Thus there is an important difference between knowledge from<br />
underst<strong>and</strong>ing <strong>and</strong> knowledge from practice (tacit knowledge).<br />
Such results indicate that the more knowledge that is embedded in firms, through working<br />
practices, research programs <strong>and</strong> <strong>on</strong>going developments, is tacit, the less transferable this<br />
knowledge is to outsiders. To some extent, if a firm cannot rati<strong>on</strong>alize why a particular routine has<br />
proved beneficial, but c<strong>on</strong>tinues with it <strong>on</strong> the basis of past performance, then it will be less able<br />
to adapt to a slight change in c<strong>on</strong>diti<strong>on</strong>s <strong>and</strong> less able to apply successful strategies to slightly<br />
different problems. Thus, it is perhaps unrealistic to expect firms to re-optimize their strategies<br />
immediately when circumstances change — by its very nature a process of trial <strong>and</strong> error (or<br />
imitati<strong>on</strong>) takes time.<br />
In a similar manner, the idea of tacit knowledge helps explain the large internati<strong>on</strong>al differences in<br />
living st<strong>and</strong>ards <strong>and</strong> ec<strong>on</strong>omic development that seem inc<strong>on</strong>sistent with knowledge being a<br />
public good transferred <strong>and</strong> adopted at minimal cost. At a more micro level, it may be expected<br />
that different organizati<strong>on</strong>s may obtain different levels of output from ostensibly the same inputs.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omists not <strong>on</strong>ly disagree with the neoclassical traditi<strong>on</strong> <strong>on</strong> the definiti<strong>on</strong> of<br />
technological knowledge but also <strong>on</strong> fundamental assumpti<strong>on</strong>s <strong>on</strong> the behavior of ec<strong>on</strong>omic<br />
agents.<br />
Evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat ec<strong>on</strong>omic agents as rati<strong>on</strong>al, in the sense they are in neoclassical<br />
ec<strong>on</strong>omics. This is not solely because ec<strong>on</strong>omic agents do not have full informati<strong>on</strong>, but<br />
more importantly because they have limited cognitive powers; they do not profit-maximize (given<br />
their informati<strong>on</strong> set) but rather aim at sufficiently high levels of success to enable survival in a<br />
dynamic envir<strong>on</strong>ment.<br />
41<br />
259
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Related to this, Dosi (1988) emphasizes that innovative search is characterized by str<strong>on</strong>g<br />
uncertainty whereby the list of possible events is unknown <strong>and</strong> <strong>on</strong>e does not know either the<br />
c<strong>on</strong>sequence of particular acti<strong>on</strong>s for any given event.<br />
Instead of being rati<strong>on</strong>al <strong>and</strong> thriving for maximizing profit, ec<strong>on</strong>omic actors are assumed to<br />
develop routine behaviors, which they follow as l<strong>on</strong>g as they turn out to work well. In case the<br />
routine behaviour fails to functi<strong>on</strong>, they will be revised <strong>and</strong> changed so that new routines may be<br />
developed. As a result of the evoluti<strong>on</strong>ary process, viable routines prevail whereas the less<br />
successful <strong>on</strong>es are eliminated. It must be menti<strong>on</strong>ed, however, that these changes are not<br />
taking place c<strong>on</strong>tinuously but are rather unevenly distributed over time <strong>and</strong> tend to occur<br />
sporadically.<br />
Innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong><br />
Variety is an important part of evoluti<strong>on</strong>ary theories of innovati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong> <strong>and</strong> can be<br />
traced back to the assumpti<strong>on</strong>s <strong>on</strong> limits to ec<strong>on</strong>omic agents’ rati<strong>on</strong>ality.<br />
A classic issue raised by the c<strong>on</strong>cept of Schumpeterian competiti<strong>on</strong> as a process of creative<br />
destructi<strong>on</strong> is that this, <strong>on</strong> its own, would imply that ec<strong>on</strong>omic growth leads to suppressi<strong>on</strong> of<br />
variety <strong>and</strong> ultimately to m<strong>on</strong>opoly (Nels<strong>on</strong> <strong>and</strong> Winter, 2002, p 35). Thus, to counter this<br />
predicti<strong>on</strong>, there is need to c<strong>on</strong>sider how new variety originates.<br />
Metcalfe (1998, p98) draws the link between the rejecti<strong>on</strong> of the neo-classical assumpti<strong>on</strong> of<br />
rati<strong>on</strong>al maximizing agents, <strong>and</strong> the generati<strong>on</strong> (<strong>and</strong> re-generati<strong>on</strong>) of variety:<br />
“…the positive side of bounded rati<strong>on</strong>ality is that it frees the imaginati<strong>on</strong> from the limits of<br />
calculati<strong>on</strong>. When problems become too complex to be well-defined, let al<strong>on</strong>e solved<br />
analytically, <strong>on</strong>e is inevitably dependent <strong>and</strong> judgment, c<strong>on</strong>jecture <strong>and</strong> the guiding h<strong>and</strong><br />
of experience.”<br />
The point here is that while rati<strong>on</strong>al agents, profit-maximizing under identical resource <strong>and</strong><br />
informati<strong>on</strong> c<strong>on</strong>straints should, by definiti<strong>on</strong>, reach the same strategy, more realistic agents,<br />
guided by judgment <strong>and</strong> experience, will reach a range of preferred strategies. This process<br />
can be expected to lead to significant variety in innovative soluti<strong>on</strong>s, within which there are<br />
elements of path dependence as agents bring past experience to bear <strong>on</strong> current problems.<br />
From these factors leading to the generati<strong>on</strong> of heterogeneity, competiti<strong>on</strong> proceeds to bring<br />
ec<strong>on</strong>omic development. As discussed below, this is because competiti<strong>on</strong> leads to improvement<br />
through selecti<strong>on</strong>, adaptati<strong>on</strong> <strong>and</strong> imitati<strong>on</strong>.<br />
Foster (2001) emphasizes how the traditi<strong>on</strong>al (mainstream) discussi<strong>on</strong> of competiti<strong>on</strong> in<br />
ec<strong>on</strong>omics has much to do with static outcomes <strong>and</strong> little to do with dynamics. By modelling<br />
competiti<strong>on</strong> in terms of equilibrium, competiti<strong>on</strong> is seen as timeless, rather than a process of<br />
development. A key feature of evoluti<strong>on</strong>ary ec<strong>on</strong>omics is its emphasis <strong>on</strong> competiti<strong>on</strong> as a<br />
process rather than an outcome.<br />
42<br />
260
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Adopting a more dynamic approach, evoluti<strong>on</strong>ary ec<strong>on</strong>omics is in part associated with the use of<br />
analogies to evoluti<strong>on</strong>ary biology to explain ec<strong>on</strong>omic growth <strong>and</strong> the process of competiti<strong>on</strong>. At<br />
a very basic level, the noti<strong>on</strong> of competiti<strong>on</strong> through natural selecti<strong>on</strong> associated with Darwin<br />
seems similar to the process by which ec<strong>on</strong>omic competiti<strong>on</strong> selects more fit (efficient <strong>and</strong><br />
profitable) firms at the expense of less fit firms.<br />
In this c<strong>on</strong>text, Mokyr (1999a) defines the Darwinian model “as a system of self-reproducing units<br />
(techniques) that changes over time.” Such a system appears to be best characterized by the<br />
following three properties:<br />
• The relati<strong>on</strong>ship through which the underlying structure determines the manifested<br />
entities. From an ec<strong>on</strong>omic viewpoint, an underlying structure is the “useful knowledge”<br />
that provides the framework for the “feasible” techniques (manifested entity) i.e. what<br />
society potentially is able to do.<br />
•<br />
•<br />
A Darwinian system is likely to undergo dramatic changes over time. Techniques<br />
reproduce themselves between two time periods through either repetiti<strong>on</strong> or learning <strong>and</strong><br />
imitati<strong>on</strong>. This process then entails c<strong>on</strong>stant changes part of which cannot be foreseen<br />
<strong>and</strong> which can turn out to destabilize the current system. An important questi<strong>on</strong> that this<br />
raises is the pace of innovati<strong>on</strong>. Is the innovati<strong>on</strong> process a gradual <strong>on</strong>e or does it occur<br />
in a stepwise manner? This is indeed a crucial issue (for example, in cliometrics as<br />
discussed below, the debate <strong>on</strong> whether the first industrial revoluti<strong>on</strong> as it occurred in Britain<br />
created disc<strong>on</strong>tinuity or whether it can be regarded as a smooth transiti<strong>on</strong>).<br />
Excessive variety in a Darwinian system implies that the actual number of techniques<br />
exceeds the number sustainable in the system. This results in a selecti<strong>on</strong> process.<br />
Mokyr (1999a) argues that selecti<strong>on</strong> can operate through three channels. The first <strong>on</strong>e is<br />
the st<strong>and</strong>ard neo-classical mechanism where techniques are chosen in the event they<br />
maximize an objective functi<strong>on</strong> incorporating supply <strong>and</strong> dem<strong>and</strong> side c<strong>on</strong>siderati<strong>on</strong> as<br />
well as externalities. Sec<strong>on</strong>d, there is inertia in the system that eliminates useless<br />
techniques. Finally, <strong>and</strong> more importantly, selecti<strong>on</strong> not <strong>on</strong>ly happens in the market but<br />
also at the social level.<br />
Inherent in this model of competiti<strong>on</strong> is the associati<strong>on</strong> between competiti<strong>on</strong> <strong>and</strong> experimentati<strong>on</strong>.<br />
A variety of experiments allows for ec<strong>on</strong>omic progress bey<strong>on</strong>d the scope that would be<br />
c<strong>on</strong>sidered achievable through comprehensive calculati<strong>on</strong> <strong>and</strong> strategy (i.e. if each agent had to<br />
specify all permutati<strong>on</strong>s of outcomes, the likelihood of each, <strong>and</strong> the interrelati<strong>on</strong>ships involved).<br />
Nels<strong>on</strong> <strong>and</strong> Winter (2002, p 28) identify that the ec<strong>on</strong>omic development ‘puzzle’ is how the vast<br />
technological progress <strong>and</strong> efficient forms of organizati<strong>on</strong> seen in the modern age could have<br />
arisen given the cognitive limits of humans <strong>and</strong> organizati<strong>on</strong>s. To this end, they propose that<br />
competiti<strong>on</strong> is the means through which the deficiency of human rati<strong>on</strong>ality <strong>and</strong> ability are<br />
mitigated to achieve favorable ec<strong>on</strong>omic outcomes:<br />
“Neo-classical ec<strong>on</strong>omics discovers social value in human selfishness, but not virtue<br />
that<br />
is robust against the human limitati<strong>on</strong> of incompetence — <strong>and</strong> the possible role of the<br />
market process in achieving that robustness is not featured.”<br />
43<br />
261
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Adopting this positi<strong>on</strong>, limited rati<strong>on</strong>ality, rather than being an impediment to well-functi<strong>on</strong>ing<br />
markets, is actually seen as part of the process of competiti<strong>on</strong>, <strong>and</strong> through it, innovati<strong>on</strong>. If<br />
agents possessed far greater cognitive powers <strong>and</strong> rati<strong>on</strong>ality, the importance of competiti<strong>on</strong> in<br />
delivering ec<strong>on</strong>omic development would be reduced.<br />
Arguably there is an interesting corresp<strong>on</strong>dence between this c<strong>on</strong>cept of competiti<strong>on</strong> <strong>and</strong> the role<br />
of competitive markets in c<strong>on</strong>veying informati<strong>on</strong>, most famously associated with Hayek (1945):<br />
“The marvel is that in a case like that of a scarcity of <strong>on</strong>e raw material, without an order<br />
being issued, without more than perhaps a h<strong>and</strong>ful of people knowing the cause, tens of<br />
thous<strong>and</strong>s of people whose identity could not be ascertained by m<strong>on</strong>ths of investigati<strong>on</strong>,<br />
are made to use the material or its products more sparingly; that is, they move in the right<br />
directi<strong>on</strong>.”<br />
In this example, the focus is <strong>on</strong> competiti<strong>on</strong> as a force for distilling informati<strong>on</strong> in an efficient<br />
manner.<br />
Thus the process of competiti<strong>on</strong> can be seen to bring benefits exactly because agents in the<br />
ec<strong>on</strong>omy do not have full informati<strong>on</strong> — for example, through “price signals” competiti<strong>on</strong> tackles<br />
the deficiencies <strong>and</strong> asymmetries of informati<strong>on</strong> across the ec<strong>on</strong>omy. Similarly, as discussed<br />
above, the process of competiti<strong>on</strong> can be seen to bring benefits exactly because agents in the<br />
ec<strong>on</strong>omy are not fully rati<strong>on</strong>al — for example, selecti<strong>on</strong> from experimentati<strong>on</strong> tackles the inability<br />
of ec<strong>on</strong>omic agents to make optimal use of all the resources <strong>and</strong> informati<strong>on</strong> they have at their<br />
disposal at a point in time.<br />
Yet the Darwinian reference to competitive selecti<strong>on</strong> does not quite mean the same in<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics as in biology. Foster (2001, p120) argues that competiti<strong>on</strong> in ec<strong>on</strong>omics<br />
should allow for “inheritance of behavioral characteristics acquired from experience in particular<br />
envir<strong>on</strong>ments.” In other words, competiti<strong>on</strong> is not just about selecti<strong>on</strong> between firms (raising<br />
industry st<strong>and</strong>ards through eliminati<strong>on</strong> of weaker firms) but also the benefits of learning through<br />
creative endeavors, experience <strong>and</strong> also imitati<strong>on</strong> that allows individual firms to make progress.<br />
(In fact Foster sees this process as more reflective of the evoluti<strong>on</strong>ary theories of Lamarck than<br />
Darwin — a significant difference being that Lamarck c<strong>on</strong>sidered that organisms evolve because<br />
they are willing or need to evolve, rather than <strong>on</strong> the chance-orientated basis of Darwin).<br />
In particular, any competitive advantage <strong>on</strong>e firm has over another depends <strong>on</strong> the extent to<br />
which the factors underlying that advantage are imitable. While species may not be able to<br />
imitate other species (imitati<strong>on</strong> is fundamentally different from the process through species may<br />
adapt in a similar way to similar circumstances), firms are often able to imitate their competitors at<br />
low cost. Therefore an evoluti<strong>on</strong>ary approach to ec<strong>on</strong>omic should not be so Darwinian as to miss<br />
the importance of imitati<strong>on</strong> in ec<strong>on</strong>omic development. Furthermore, the role of imitati<strong>on</strong> in<br />
ec<strong>on</strong>omic development will be governed by the ability of firms to maintain their competitive<br />
advantage through patents, <strong>and</strong> secrets, as well as relative speeds of innovati<strong>on</strong> where timing<br />
matters to success. (How firms seek competitive advantage is discussed further in secti<strong>on</strong> 3<br />
within the c<strong>on</strong>text of the management <strong>and</strong> strategy literature.)<br />
44<br />
262
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Thus ec<strong>on</strong>omic development <strong>and</strong> innovati<strong>on</strong> can be seen as a combined effect of selecti<strong>on</strong> (via<br />
competiti<strong>on</strong>) from a variety of competing routines <strong>and</strong> practices as well as the more endogenous<br />
process of agents seeking improved routines <strong>and</strong> practices. The role of ‘dynamic’ competiti<strong>on</strong> in<br />
this regard is quite different from its “static” force in c<strong>on</strong>straining prices.<br />
This emphasis <strong>on</strong> competiti<strong>on</strong> as a selecti<strong>on</strong> process is supported by a quite separate str<strong>and</strong> of<br />
empirical literature that seeks to decompose industry-level productivity growth to examine the<br />
‘microstructure’ of this growth. Such studies use panel data of firms (or plants within multi-plant<br />
firms) to examine the extent to which ec<strong>on</strong>omic development at the industry level is driven by<br />
productivity improvements within firms (‘internal restructuring’) versus expansi<strong>on</strong> (<strong>and</strong> entry) of<br />
high-productivity firms <strong>and</strong> c<strong>on</strong>tracti<strong>on</strong> (<strong>and</strong> exit) of low-productivity firms (‘external<br />
restructuring’).<br />
For example, Disney et al. (2000) analyze a dataset which covers the period from 1980 to 1992<br />
<strong>and</strong> c<strong>on</strong>tains data for 140,000 UK manufacturing establishments a year. They examine the<br />
relati<strong>on</strong>ship between industry-level <strong>and</strong> firm-level dynamics. Of particular significance they find<br />
that external restructuring explains approximately 50 per cent of changes in labor productivity,<br />
<strong>and</strong> 90 per cent of changes in total factor productivity, over the period. These results indicate that<br />
in the industries studied, competiti<strong>on</strong> in terms of selecti<strong>on</strong> from variety was a very important driver<br />
of productivity growth.<br />
Note that such studies focus more <strong>on</strong> technological <strong>and</strong> organizati<strong>on</strong> improvements than drastic<br />
innovati<strong>on</strong>s (major new lines of products) that would not be fully captured in the productivity<br />
studies. But it is clear to see how evidence of internal versus external restructuring, in the<br />
c<strong>on</strong>text of industry-level ec<strong>on</strong>omic development fits in with the evoluti<strong>on</strong>ary c<strong>on</strong>cepts of<br />
competiti<strong>on</strong> as a selecti<strong>on</strong> process.<br />
Industry structure <strong>and</strong> innovati<strong>on</strong><br />
Schumpeter is often associated with the hypothesis that large firms with market power are more<br />
innovative than small <strong>on</strong>es, although Fagerberg (2002) argues that Schumpeter was more<br />
interested in the difference between new firms <strong>and</strong> old firms. Evoluti<strong>on</strong>ary ec<strong>on</strong>omics offers an<br />
interesting perspective <strong>on</strong> the relati<strong>on</strong>ship between firm size <strong>and</strong> innovati<strong>on</strong>.<br />
Fagerberg (2002) cites models by Nels<strong>on</strong> <strong>and</strong> Winter from the early 1980s that seek to explain<br />
why larger firms may be more innovative than smaller firms. On the <strong>on</strong>e h<strong>and</strong>, if firms use<br />
retained profits to finance R&D, large successful firms will have advantages over smaller firms.<br />
On the other h<strong>and</strong>, a large firm is able to derive greater benefit from finding a better “routine”<br />
because it can put it to use across a larger number of units of producti<strong>on</strong>. Such analysis would<br />
suggest that large firms are at a competitive advantage, although Nels<strong>on</strong> <strong>and</strong> Winter are reported<br />
to see this effect counteracted by large firms (with more market power) having a higher profit<br />
target (price/cost rati<strong>on</strong>) <strong>and</strong> this may provide some restraint in these dominating smaller firms.<br />
These models also suggest that if large firms are aggressive in pursuing their advantage over<br />
small firms, <strong>and</strong> there is in turn a tendency towards c<strong>on</strong>centrati<strong>on</strong>, large “imitating” firms may do<br />
well at the expense of smaller innovators. Thus where technological progress is endogenous,<br />
productivity may be hindered compared to a situati<strong>on</strong> of a more varied industry structure.<br />
45<br />
263
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
An alternative way of c<strong>on</strong>sidering whether large firms are more innovative than small firms is to<br />
focus more <strong>on</strong> heterogeneity in ability. To the extent that competiti<strong>on</strong> leads to the growth of<br />
innovative firms, we might simply expect causality to run the other way: it is not that large firms<br />
are inherently better innovators than smaller firms, just that the better innovators will be<br />
successful <strong>and</strong> hence grow in size. Theories based al<strong>on</strong>g these lines would therefore be treating<br />
innovati<strong>on</strong> as exogenous to firm size, with causality running from the former to the latter.<br />
This expositi<strong>on</strong> between exogenous <strong>and</strong> endogenous innovati<strong>on</strong> is valid more generally than firm<br />
size. Nels<strong>on</strong> <strong>and</strong> Winter (2002) draw a distincti<strong>on</strong> between industries that are characterized by<br />
“science-based” innovati<strong>on</strong> <strong>and</strong> industries that are characterized by “cumulative” innovati<strong>on</strong>.<br />
Science-based industries are c<strong>on</strong>sidered to be those where the thrust of innovati<strong>on</strong> comes from<br />
R&D activities outside that industry. For example, firms in such an industry might benefit<br />
principally from external scientific developments <strong>and</strong> innovative activities by their suppliers. By<br />
c<strong>on</strong>trast, in cumulative industries, innovati<strong>on</strong> stems from R&D activity within each of the firms in<br />
that industry.<br />
An important difference may be expected between science-based industries <strong>and</strong> cumulative<br />
industries. In the former, where firms reap the benefits of innovative activities from outside their<br />
own industry, new entrants <strong>and</strong> previous laggards may be expected to have similar scope for<br />
development than industry leaders. Just because a firm has performed poorly in the past, or is a<br />
newcomer to an industry, should not totally undermine its ability to take advantage of external<br />
innovati<strong>on</strong>s if these are exogenous.<br />
However, if the innovative activity within an industry comes predominantly from the efforts of firms<br />
within that industry, we may expect a significant degree of path dependency. Successful<br />
innovators in <strong>on</strong>e period will gain advantage over other firms that allows them greater potential for<br />
innovati<strong>on</strong> the next period. The cumulative effect of this is that less innovative firms fall further<br />
<strong>and</strong> further behind, <strong>and</strong> incumbents have some degree of advantage over new entrants.<br />
High-level policy implicati<strong>on</strong>s<br />
In discussing policy implicati<strong>on</strong>s, Fagerberg (2002) draws distincti<strong>on</strong> between evoluti<strong>on</strong>ary<br />
approaches <strong>and</strong> neo-classical approaches. Neo-classical ec<strong>on</strong>omics places emphasis <strong>on</strong> the<br />
failure of markets to deliver “public goods”, <strong>and</strong> makes the case for government support to<br />
provide the public good of knowledge. By c<strong>on</strong>trast, evoluti<strong>on</strong>ary ec<strong>on</strong>omics does not treat<br />
knowledge as a public good in this sense; for example tacit knowledge exists in routines <strong>and</strong><br />
practices rather than the c<strong>on</strong>sciousness <strong>and</strong> cannot be transferred from <strong>on</strong>e party to another in<br />
the way that a formula or design can. Adopting a perspective more in line with the evoluti<strong>on</strong>ary<br />
approach would therefore provide les support for government funding of knowledge as a public<br />
good.<br />
Furthermore, evoluti<strong>on</strong>ary ec<strong>on</strong>omics argues that variety is an important part of the scope for<br />
innovati<strong>on</strong> <strong>and</strong> development, while the role of heterogeneity per se is underplayed in neoclassical<br />
ec<strong>on</strong>omics. Policy proposals stem from this stance. For example:<br />
46<br />
264
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
!Rather than subsidizing R&D in well-established firms in traditi<strong>on</strong>al sectors, <strong>on</strong>e<br />
might put the resources into new types of activities or actors, not necessarily with the<br />
expectati<strong>on</strong> that these would do extremely well, but because the entire system<br />
(including thetraditi<strong>on</strong>al sectors) might benefit from such increased diversity”<br />
(Fagerberg, p 41).<br />
2.3 Systems of Innovati<strong>on</strong> <strong>and</strong> the Role of Instituti<strong>on</strong>s<br />
The emphasis <strong>on</strong> the noti<strong>on</strong> of “systems” of innovati<strong>on</strong> suggests that innovati<strong>on</strong> is not a<br />
phenomen<strong>on</strong> which results from the activities of isolated innovative firms, but the result of<br />
complex important interacti<strong>on</strong>s between firms <strong>and</strong> other organizati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s. Each<br />
innovative activity by a certain organizati<strong>on</strong> involves an element of reliance <strong>on</strong> external sources<br />
<strong>and</strong> innovati<strong>on</strong> in the ec<strong>on</strong>omy results from the complementary c<strong>on</strong>tributi<strong>on</strong>s of different<br />
organizati<strong>on</strong>s in a particular instituti<strong>on</strong>al envir<strong>on</strong>ment.<br />
This str<strong>and</strong> of the literature draws together many comm<strong>on</strong> themes between two related<br />
approaches — evoluti<strong>on</strong>ary ec<strong>on</strong>omics <strong>and</strong> cliometrics. What is cliometrics?<br />
Cliometrics was initiated by Alfred C<strong>on</strong>rad <strong>and</strong> John Meyer (C<strong>on</strong>rad <strong>and</strong> Meyer<br />
1957,1958) <strong>and</strong> has been popularized by Robert Fogel <strong>and</strong> Douglass North.<br />
Cliometrics is also often labelled as new ec<strong>on</strong>omic history or quantitative ec<strong>on</strong>omic<br />
history. It attempts to provide new insights into ec<strong>on</strong>omic history, mainly by<br />
c<strong>on</strong>sidering issues related to ec<strong>on</strong>omic development <strong>and</strong> growth.<br />
A more precise definiti<strong>on</strong> for cliometrics is given by McCloskey (1978): “a<br />
cliometrician is an ec<strong>on</strong>omist applying ec<strong>on</strong>omic theory (usually simple) to historical<br />
facts (not always quantitative) in the interest of history (not ec<strong>on</strong>omics)”. Despite<br />
substantial initial criticism, cliometrics has become a well-established branch of<br />
ec<strong>on</strong>omics <strong>and</strong> in 1993, the NobelPrize in Ec<strong>on</strong>omics was awarded to Robert Fogel<br />
<strong>and</strong> Douglass North.<br />
According to Claude Diebolt , cliometrics can be broken down into three major<br />
categories. First, descriptive cliometrics, which seeks to disentangle the ec<strong>on</strong>omic<br />
<strong>and</strong> social influence that might have exerted an influence <strong>on</strong> historical events <strong>and</strong><br />
processes.<br />
Sec<strong>on</strong>d, positive cliometrics, which is c<strong>on</strong>cerned with organizing data based <strong>on</strong><br />
which then different theoretical c<strong>on</strong>jectures can be tested. Third, normative<br />
cliometrics, which cannot be tested empirically <strong>and</strong> which deal with ethical questi<strong>on</strong>s<br />
such as efficiency,distributive <strong>and</strong> social justice <strong>and</strong> with how to ameliorate social<br />
welfare.<br />
Cliometrics is largely based <strong>on</strong> rec<strong>on</strong>structed l<strong>on</strong>g (secular) time series or large panel<br />
data related to historic events. The theoretical background underlying cliometrics was<br />
initially neo-classical theory, but it has been evolving c<strong>on</strong>tinuously with<br />
developments in<br />
general ec<strong>on</strong>omics over the past 40 years or so. Recent works in this field are<br />
essentially related to evoluti<strong>on</strong>ary ec<strong>on</strong>omics as a theoretical background.<br />
47<br />
265
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Defining ‘Innovati<strong>on</strong> Systems’<br />
Different definiti<strong>on</strong>s of the c<strong>on</strong>cept of innovati<strong>on</strong> systems have been offered in the literature, <strong>and</strong>,<br />
despite the inevitable differences, all point towards the importance of the noti<strong>on</strong> that inventi<strong>on</strong>,<br />
innovati<strong>on</strong> <strong>and</strong> its diffusi<strong>on</strong> result from interacti<strong>on</strong>s between different complementary<br />
organisati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s.<br />
Freeman (1987) defines a system of innovati<strong>on</strong> as:<br />
“[…] a network of instituti<strong>on</strong>s in the public <strong>and</strong> private sectors whose activities <strong>and</strong><br />
interacti<strong>on</strong>s initiate, import, modify <strong>and</strong> diffuse new technology”.<br />
Lundvall (1992) defines a system of innovati<strong>on</strong> as a system that includes all parts <strong>and</strong> aspects of<br />
the ec<strong>on</strong>omic structure <strong>and</strong> the instituti<strong>on</strong>al set-up affecting learning as well searching <strong>and</strong><br />
exploring. Alternatively, Gregers<strong>on</strong> <strong>and</strong> Johns<strong>on</strong> (1998, pp. 5) provides the following descripti<strong>on</strong><br />
for system of innovati<strong>on</strong>:<br />
“…overall innovati<strong>on</strong> performance of an ec<strong>on</strong>omy depends not <strong>on</strong>ly <strong>on</strong> how specific<br />
organizati<strong>on</strong>s like firms <strong>and</strong> research institutes perform, but also <strong>on</strong> how they interact<br />
with each other <strong>and</strong> with the government sector in knowledge producti<strong>on</strong> <strong>and</strong><br />
distributi<strong>on</strong>. Innovating firms operate within a comm<strong>on</strong> instituti<strong>on</strong>al set-up <strong>and</strong> they<br />
jointly depend <strong>on</strong>, c<strong>on</strong>tribute to <strong>and</strong> utilize a comm<strong>on</strong> knowledge infrastructure. It can<br />
be thought of as a system which creates <strong>and</strong> distributes knowledge, utilizes this<br />
knowledge by introducing itinto the ec<strong>on</strong>omy in the form of innovati<strong>on</strong>s, diffuses it<br />
<strong>and</strong> transforms it into something valuable, for example, internati<strong>on</strong>al competitiveness<br />
<strong>and</strong> ec<strong>on</strong>omic growth.”<br />
In general a system of innovati<strong>on</strong> is defined in terms of its comp<strong>on</strong>ents <strong>and</strong> the relati<strong>on</strong>s through<br />
which the comp<strong>on</strong>ents are linked with other another.<br />
The elements of the system of innovati<strong>on</strong> are organizati<strong>on</strong>s <strong>and</strong> instituti<strong>on</strong>s. Instituti<strong>on</strong>s exert a<br />
c<strong>on</strong>siderable influence <strong>on</strong> organizati<strong>on</strong>s, <strong>and</strong> simultaneously, organizati<strong>on</strong>s also impact <strong>on</strong> the<br />
instituti<strong>on</strong>al envir<strong>on</strong>ment. Instituti<strong>on</strong>s may give rise to new organizati<strong>on</strong>s as well as being the<br />
origin of new instituti<strong>on</strong>s.<br />
Edquist (2001) argues that organizati<strong>on</strong>s are formal structures that are created with a well defined<br />
goal in mind <strong>on</strong> the <strong>on</strong>e h<strong>and</strong>, <strong>and</strong> instituti<strong>on</strong>s such comm<strong>on</strong> habits, routines, established<br />
practices, rules or laws <strong>on</strong> the other. Organizati<strong>on</strong>s within the system can be broken down into<br />
two categories - primary <strong>and</strong> sec<strong>on</strong>dary. The activities listed above are associated with primary<br />
organizati<strong>on</strong>s whereas sec<strong>on</strong>dary organizati<strong>on</strong>s impact <strong>on</strong> how primary organizati<strong>on</strong>s behave<br />
when performing fundamental activities.<br />
Schoser (1999) argues that the instituti<strong>on</strong>al setting can be broken down into formal <strong>and</strong> informal<br />
instituti<strong>on</strong>s <strong>and</strong> that the factors identified by Gregers<strong>on</strong> <strong>and</strong> Johns<strong>on</strong> (1998) – instituti<strong>on</strong>al set-up,<br />
knowledge infrastructure, specializati<strong>on</strong> patterns, public <strong>and</strong> private dem<strong>and</strong> structure <strong>and</strong><br />
government policy – may impact <strong>on</strong> innovati<strong>on</strong> either directly or indirectly.<br />
Typical organizati<strong>on</strong>s in an innovati<strong>on</strong> system are private firms, universities, government<br />
laboratories <strong>and</strong> industrial research associati<strong>on</strong>s, am<strong>on</strong>g which a divisi<strong>on</strong> of labor is realized,<br />
48<br />
266
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
mainly coordinated by n<strong>on</strong>-market means. These various organizati<strong>on</strong>s undertake different<br />
activities within the system, depending <strong>on</strong> their particular objectives, research mechanisms <strong>and</strong><br />
incentive structures.<br />
The divisi<strong>on</strong> of labor within a system of innovati<strong>on</strong> explains why the systemic view holds that not<br />
<strong>on</strong>ly the comp<strong>on</strong>ents of a system of innovati<strong>on</strong> are important but crucially also their links <strong>and</strong><br />
interacti<strong>on</strong>s. Metcalfe (1995) observes that c<strong>on</strong>nectivity within a system is achieved by a variety<br />
of mechanisms as for instance:<br />
•<br />
•<br />
•<br />
mobility of employers in the labor market;<br />
grants <strong>and</strong> c<strong>on</strong>tracts for research; <strong>and</strong><br />
informal networks, e.g. the links between user firms <strong>and</strong> their suppliers<br />
How ‘Innovati<strong>on</strong> Systems’ Work<br />
There are various activities undertaken within a system. Liu <strong>and</strong> White (2001), identify five<br />
activities: research; implementati<strong>on</strong>, i.e. manufacturing; end-use; educati<strong>on</strong> <strong>and</strong> linkage. Bergek<br />
<strong>and</strong> Jacobss<strong>on</strong> (2002) put forward the following functi<strong>on</strong>s of a system of innovati<strong>on</strong>: creati<strong>on</strong> of<br />
new knowledge; guidance of the research process; supply of resources; creati<strong>on</strong> of positive<br />
externalities <strong>and</strong> creati<strong>on</strong> of markets.<br />
Pearce (2002) offer a different approach to defining the activities of a system of innovati<strong>on</strong>.<br />
Pearce identifies basic research as the first element <strong>on</strong> which applied research is based. Product<br />
development including market research is then built <strong>on</strong> both basic <strong>and</strong> applied research. The<br />
final stage of the system of innovati<strong>on</strong> is adaptati<strong>on</strong> <strong>and</strong> marketing of the products. The first two<br />
stages depend heavily <strong>on</strong> research instituti<strong>on</strong>s, government support as well as firms. By c<strong>on</strong>trast,<br />
the last two stages mainly c<strong>on</strong>cern private firms.<br />
N<strong>on</strong>etheless, Nels<strong>on</strong> (1996) notes that the noti<strong>on</strong> of systems of innovati<strong>on</strong> does not imply that the<br />
system was c<strong>on</strong>sciously designed or that interacti<strong>on</strong>s between the comp<strong>on</strong>ents of the system<br />
work are smoothly <strong>and</strong> that the system is coherent as a whole.<br />
Systems of innovati<strong>on</strong> are often c<strong>on</strong>sidered to be nati<strong>on</strong>al, although it is debated whether nati<strong>on</strong>al<br />
boundaries can be deemed relevant.<br />
On the <strong>on</strong>e h<strong>and</strong> it is argued that the element of nati<strong>on</strong>ality may derive from different factors such<br />
as the nati<strong>on</strong>al focus of technological <strong>and</strong> other policies, laws <strong>and</strong> regulati<strong>on</strong>s that have an impact<br />
<strong>on</strong> the innovative envir<strong>on</strong>ment (Metcalfe, 1995). Nels<strong>on</strong> (1996) argues that because the<br />
educati<strong>on</strong> <strong>and</strong> research system, public infrastructure, laws <strong>and</strong> financial instituti<strong>on</strong>s keep some of<br />
their nati<strong>on</strong>al characteristics, differences across different nati<strong>on</strong> innovati<strong>on</strong> systems are likely to<br />
persist even in the l<strong>on</strong>ger run.<br />
On the other h<strong>and</strong>, the boundaries of systems of innovati<strong>on</strong> in different countries are getting<br />
blurred with the rise of internati<strong>on</strong>al firms present in several countries <strong>and</strong> increasingly intensifying<br />
cross-country inter-firm c<strong>on</strong>necti<strong>on</strong>s to share knowledge <strong>and</strong> innovate.<br />
49<br />
267
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
It may also be argued that systems of innovati<strong>on</strong> may be local <strong>and</strong> regi<strong>on</strong>al rather than nati<strong>on</strong>al<br />
<strong>and</strong> that <strong>on</strong>e may think of different systems of innovati<strong>on</strong>s at various levels of aggregati<strong>on</strong>.<br />
It is widely recognized in the literature that there is a dilemma in nati<strong>on</strong>al systems of innovati<strong>on</strong>.<br />
This dilemma is whether to create incentives for innovati<strong>on</strong> or whether to foster diffusi<strong>on</strong> of<br />
innovati<strong>on</strong>. At the heart of this problem is the patent system that encourages innovati<strong>on</strong> by<br />
allowing for reaping the benefits <strong>and</strong> which at the same time hinders diffusi<strong>on</strong>.<br />
The role patents play in innovati<strong>on</strong> is c<strong>on</strong>troversial. For instance, Levin et al (1987) report a fairly<br />
limited role of patents when it comes to protecting new technology. A brief discussi<strong>on</strong> of some of<br />
the findings of the literature review regarding patents is presented in the box below.<br />
Mokyr (1992, 1999a) also suggests that some instituti<strong>on</strong>al arrangements may hinder innovati<strong>on</strong>.<br />
He points out that self-organizing systems become the cornerst<strong>on</strong>e of evoluti<strong>on</strong>ary ec<strong>on</strong>omics<br />
<strong>and</strong> notes that:<br />
“…the most interesting property of these systems is that they resist change. Resistance<br />
to change is essential for any system if it is to functi<strong>on</strong> <strong>and</strong> not degenerate into chaos.”<br />
However, he also emphases that resistance in society might have not <strong>on</strong>ly a stabilizing effect but<br />
can also act as a break <strong>on</strong> innovati<strong>on</strong>. Indeed, high resistance can slow down or even eliminate<br />
the introducti<strong>on</strong> of innovati<strong>on</strong>s or even deter innovati<strong>on</strong> activity.<br />
Oppositi<strong>on</strong> against new innovati<strong>on</strong>s might come from both owners <strong>and</strong> employees. Mokyr<br />
(1999a) enumerates the following sources of resistance:<br />
•<br />
•<br />
•<br />
•<br />
ec<strong>on</strong>omically motivated resistance;<br />
ideologically motivated resistance;<br />
systemic resistance; <strong>and</strong><br />
frequency resistance.<br />
However, the success of an innovati<strong>on</strong> hinges <strong>on</strong> two factors. On the <strong>on</strong>e h<strong>and</strong>, the greater the<br />
value of the equipment <strong>and</strong> the more specific the skill that are threatened by a new innovati<strong>on</strong>,<br />
the higher the incentive to resist. On the other h<strong>and</strong>, in the event that the innovati<strong>on</strong> generates<br />
high social gains, the incentive to establish the innovati<strong>on</strong> will be equally high.<br />
Sec<strong>on</strong>d, the introducti<strong>on</strong> of the new technology requires time for people to become familiar with it<br />
<strong>and</strong> to make best use of it. Atkens<strong>on</strong> <strong>and</strong> Kehoe (2002) build a quantitative model of technology<br />
diffusi<strong>on</strong> in order to simulate this slow transiti<strong>on</strong> from a low towards a high-productivity regime.<br />
The model seems to fit actual data rather well for the late-19th century US manufacturing sector.<br />
One important feature of the model is that the new technology can be adopted <strong>on</strong>ly by building<br />
new plants. This assumpti<strong>on</strong> is c<strong>on</strong>sistent with empirical literature <strong>on</strong> the microstructure of<br />
productivity growth that finds that multi-plant firms often achieve productivity gains by closing<br />
50<br />
268
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
down inefficient plants rather than improving efficiency within those plants (e.g. Disney et al,<br />
2000).<br />
For the model to produce the slow transiti<strong>on</strong>, another assumpti<strong>on</strong>s are needed. Manufacturers<br />
using the old technology dispose of a slowly accumulated stock of knowledge. At the outset of<br />
the transiti<strong>on</strong> (from old to new technology), they are hesitant to relinquish this stock of knowledge<br />
in favour of the new technology. The higher the stock of knowledge compared to the <strong>on</strong>e related<br />
to the new technology, the slower they are to go for the new technology.<br />
51<br />
269
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
The role of patents<br />
A number of papers attempt to analyze the history of patents <strong>and</strong> the patent system in<br />
theUS. For instance, Sokoloff (1988) examines 4,500 patents accepted in the US<br />
during the period 1790-1846 <strong>and</strong> sheds light <strong>on</strong> the fact that the number of patents<br />
changed at an uneven pace over time. Apart from an important rise of roughly 300<br />
per cent in patents that can be observed during 1805-10 <strong>and</strong> 1820-36, the number of<br />
patents recorded stayed fairly c<strong>on</strong>stant during the rest of the period.<br />
Using a different set of US patents, Phillips (1992) is c<strong>on</strong>cerned with the period<br />
spanning from 1831 to 1879 <strong>and</strong> c<strong>on</strong>firms the finding of Sokoloff (1988). Accepted<br />
patents rose sharply over the periods 1950-60 <strong>and</strong> 1962-68, but remained relatively<br />
stable in other years. If the number of patents is viewed as a good proxy for<br />
innovative activity, it could be argued that innovati<strong>on</strong> follows a stochastic path.<br />
Sokoloff (1988) suggested that the surge in the number of patents occurred between<br />
1820 <strong>and</strong> 1936 was due to major changes in the patent system that enforced<br />
incentives for innovative activities. In fact, patent theory tells us that the higher the<br />
expected return to an innovati<strong>on</strong>, the higher the incentive to innovate.<br />
Khan (1995) puts forward that the US patent system, which was introduced in 1790<br />
,<strong>and</strong> especially the pro-innovati<strong>on</strong> attitude of judges when accepting patent dem<strong>and</strong>s<br />
fostered substantially innovative activity in 19 century US.<br />
Phillips (1992) emphasizes the dem<strong>and</strong>-driven nature of patent activity. He argues<br />
that not <strong>on</strong>ly accumulati<strong>on</strong> of human capital, but also the increasing dem<strong>and</strong> for<br />
patentable ideas <strong>and</strong> new technology in the c<strong>on</strong>text of increasing industrializati<strong>on</strong>,<br />
which lead to an increase in accepted patents.<br />
However, some others have argued that patents do not truly reflect innovati<strong>on</strong>s.<br />
MacLeod (1986) points out that patentees are likely to react sensitively to the<br />
ec<strong>on</strong>omic <strong>and</strong> financial c<strong>on</strong>juncture. She sets out to show that the increase in<br />
registered patents between 1691-1693 was influenced by the increased availability of<br />
risk capital. It remains an unanswered questi<strong>on</strong>, however, whether innovati<strong>on</strong>s were<br />
also that dependent <strong>on</strong> the ec<strong>on</strong>omic envir<strong>on</strong>ment. MacLeod also notes that the<br />
<strong>on</strong>going war gave a significant impetus to military innovati<strong>on</strong>. Another important<br />
questi<strong>on</strong> here is the degree of development of the patent system. It is argued that<br />
weaknesses in the system probably exerted a deterrent or at least not c<strong>on</strong>ductive for<br />
patenting new innovati<strong>on</strong>.<br />
MacLeod (1992) investigates the 19 century British mechanical engineering industry<br />
<strong>and</strong> argues that the existence of a patent system was <strong>on</strong>e key to innovati<strong>on</strong>. The<br />
system provided certainty for innovators so that they could reap the benefits of their<br />
innovati<strong>on</strong>. Therefore, the uncertainty prevailing before 1830 due to the unreformed<br />
<strong>and</strong> inappropriate patent system encouraged secrecy, which in turn worked against<br />
the diffusi<strong>on</strong> of new technology.<br />
52<br />
270
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Do Different “Systems” Matter?<br />
There are differences of opini<strong>on</strong> as to whether differences between “Innovati<strong>on</strong> Systems “ matter<br />
in terms of promoting innovati<strong>on</strong> <strong>and</strong> technical change.<br />
One factor that can promote technology diffusi<strong>on</strong> is the presence of multinati<strong>on</strong>al companies<br />
worldwide. However, Saviotti (2000) documents substantial differences that persist between<br />
nati<strong>on</strong>al systems of innovati<strong>on</strong> of different countries. He observes that different instituti<strong>on</strong>al <strong>and</strong><br />
organisati<strong>on</strong>al c<strong>on</strong>figurati<strong>on</strong>s result in different specializati<strong>on</strong> patterns across countries (see also<br />
secti<strong>on</strong> 1.4 <strong>on</strong> new ec<strong>on</strong>omic geography.)<br />
Nels<strong>on</strong> (1996) also sets out to analyze the cause of differences in different nati<strong>on</strong>al systems of<br />
innovati<strong>on</strong>. Innovati<strong>on</strong> systems are analyzed in three groups of countries, namely large highincome<br />
countries such as the US, Germany, Japan, small high-income countries such as Sweden<br />
<strong>and</strong> developing, low-income countries, e.g. South Korea <strong>and</strong> Argentina.<br />
Based <strong>on</strong> this categorizati<strong>on</strong> of countries, it is argued that the size <strong>and</strong> the affluence of a<br />
particular country mainly determine its capability to innovate. In additi<strong>on</strong>, ec<strong>on</strong>omic <strong>and</strong> political<br />
circumstances <strong>and</strong> priorities also have a significant role to play in shaping the innovati<strong>on</strong> system.<br />
Further to this, Nels<strong>on</strong> (1996) emphasis the importance of competiti<strong>on</strong>. Str<strong>on</strong>g competiti<strong>on</strong> is<br />
believed to promote innovati<strong>on</strong>s. However, he argues that domestic markets especially in small<br />
countries fail to play this role <strong>and</strong> therefore internati<strong>on</strong>al competiti<strong>on</strong> encountered in world<br />
markets, that is <strong>on</strong> export <strong>and</strong> import markets, is crucial in fostering innovative activity.<br />
A str<strong>on</strong>g <strong>and</strong> competitive export sector is thought to be the engine of ec<strong>on</strong>omic growth in many<br />
ec<strong>on</strong>omies. However, for the export sector to withhold internati<strong>on</strong>al competiti<strong>on</strong>, the nati<strong>on</strong>al<br />
system of innovati<strong>on</strong> should be shaped to increase innovati<strong>on</strong> <strong>and</strong> thus support exports.<br />
N<strong>on</strong>etheless, competitiveness can be viewed <strong>and</strong> defined differently in high-income <strong>and</strong> lowincome<br />
countries. High wage level in developed countries implies that <strong>on</strong>ly new, improved<br />
products <strong>and</strong> the c<strong>on</strong>stant ameliorati<strong>on</strong> of the underlying technological process offer possibilities<br />
to boost competitiveness. By c<strong>on</strong>trast, in less developed countries with lower wage levels,<br />
competitiveness is to be improved by diffusing <strong>and</strong> learning foreign technology.<br />
Security policy c<strong>on</strong>siderati<strong>on</strong>s also seem to forge the shape of the innovati<strong>on</strong> system. Nels<strong>on</strong><br />
(1996) argues that the bulk of government backed research <strong>and</strong> development is c<strong>on</strong>centrating <strong>on</strong><br />
the defense industry in the US <strong>and</strong> Britain. Military inventi<strong>on</strong>s are then applied to civil problems<br />
giving birth to new products.<br />
Furthermore, the current German <strong>and</strong> Japanese systems of innovati<strong>on</strong> deeply root in the past<br />
when the system was designed to back the producti<strong>on</strong> of better weap<strong>on</strong>ry. However, after World<br />
War II, those innovative capacities were turned toward civilian producti<strong>on</strong>, <strong>and</strong> this led to<br />
flourishing exports.<br />
This is a clear sign for the str<strong>on</strong>g c<strong>on</strong>tinuity in nati<strong>on</strong>al innovati<strong>on</strong> systems over time. Nels<strong>on</strong><br />
(1996) points out that the German <strong>and</strong> French innovati<strong>on</strong> systems in 1990 were very similar to<br />
53<br />
271
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
those in 1890. This phenomen<strong>on</strong> seems to hold true for the majority of developed countries<br />
except for the US where a major change occurred after 1945.<br />
Grupp et al. (2001) set out to investigate the system of innovati<strong>on</strong> in Germany throughout the past<br />
150 years <strong>and</strong> come to the c<strong>on</strong>clusi<strong>on</strong> that the innovati<strong>on</strong> system has been very stable over the<br />
period under investigati<strong>on</strong>. According to the authors the industrial research system can be<br />
associated with an increase in innovative activity. When analyzing how R&D expenditures <strong>and</strong><br />
innovati<strong>on</strong> activity have changed from 1850 to 1999, they found that the German industrial<br />
research system has been very stable. In particular, they provide some more in depth<br />
ec<strong>on</strong>ometric analysis, which reveals that during 1850-1913, patent activity can be predicted by<br />
growth in human capital. Furthermore, there is causal link from patent activity to st<strong>and</strong>ard of<br />
livings whereas st<strong>and</strong>ard of living influenced human capital.<br />
These results clearly indicate the presence of a virtuous circle that results in substantial ec<strong>on</strong>omic<br />
growth. Results for the sec<strong>on</strong>d period (1951-99) are less straightforward: the st<strong>and</strong>ard of living<br />
variable appears exogenous <strong>and</strong> turns out to explain patent activity as well as <strong>on</strong> public <strong>and</strong><br />
private expenditures in R&D.<br />
The educati<strong>on</strong> system <strong>and</strong> especially universities have a significant role to play in the formati<strong>on</strong><br />
<strong>and</strong> the successful functi<strong>on</strong>ing of the innovati<strong>on</strong> system. One important questi<strong>on</strong> to be addressed<br />
is how <strong>and</strong> to what extent the research <strong>and</strong> teaching orientati<strong>on</strong> of universities satisfies the need<br />
for technological innovati<strong>on</strong>.<br />
In fact, the design of a good interface between public research organizati<strong>on</strong>s <strong>and</strong> private firms<br />
has been central to innovati<strong>on</strong> policy in most advanced Countries. This interface is c<strong>on</strong>sidered<br />
important both because it may allow research c<strong>on</strong>ducted in public organizati<strong>on</strong>s to be more easily<br />
translated into ec<strong>on</strong>omic growth <strong>and</strong> competitiveness <strong>and</strong> because it may result in fertile<br />
combinati<strong>on</strong> of complementary knowledge-basis <strong>and</strong> skills. Countering these benefits, the<br />
c<strong>on</strong>cern that closer linkages between public <strong>and</strong> private research organizati<strong>on</strong>s, <strong>and</strong> an increased<br />
privatizati<strong>on</strong> of university research, may diminish the variety of reward systems in the ec<strong>on</strong>omy<br />
<strong>and</strong> in the l<strong>on</strong>g-term the potential of public research organizati<strong>on</strong>s to c<strong>on</strong>tribute to the innovati<strong>on</strong><br />
system.<br />
Evidence from Studies using Cliometric techniques<br />
Cliometrics also offers some observati<strong>on</strong>s <strong>on</strong> the role of the relati<strong>on</strong>ship between innovati<strong>on</strong> <strong>and</strong><br />
instituti<strong>on</strong>s.<br />
Neo-classical theory posits that ec<strong>on</strong>omic agents maximise aggregate income in the absence of<br />
transacti<strong>on</strong> costs. However, since Coase (1960), it is clear that transacti<strong>on</strong> costs do matter. For<br />
instance, Wallis <strong>and</strong> North (1986) show that in 1970, 45 per cent of US GNP was linked to the<br />
transacti<strong>on</strong> sector. As a corollary, the instituti<strong>on</strong> setting, which, to a large extent, determines the<br />
magnitude of transacti<strong>on</strong> costs, has a c<strong>on</strong>siderable impact <strong>on</strong> the final outcome.<br />
What matters in the l<strong>on</strong>g- run is the c<strong>on</strong>tinuous interacti<strong>on</strong> between instituti<strong>on</strong>s <strong>and</strong> organizati<strong>on</strong>s<br />
that exerts an influence <strong>on</strong> developments in the instituti<strong>on</strong>al envir<strong>on</strong>ment. Organizati<strong>on</strong>s are born<br />
<strong>and</strong> are operating within the given framework of the instituti<strong>on</strong>s <strong>and</strong> their activities reflect the<br />
54<br />
272
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
opportunities provided by instituti<strong>on</strong>s. If, for example, security is lax, piracy will come into<br />
existence, whereas if, say, property rights are properly enforced, actors may engage in more<br />
innovative activities. (This approach is indeed very close to the systems of innovati<strong>on</strong>)<br />
Ols<strong>on</strong> (1996) puts forward that instituti<strong>on</strong>s are essential for efficiently functi<strong>on</strong>ing markets.<br />
N<strong>on</strong>etheless, it does not necessarily hold true that instituti<strong>on</strong>s provide per se the c<strong>on</strong>diti<strong>on</strong>s<br />
required for efficiency in markets <strong>and</strong> therefore they are subject to changes. If the rules of the<br />
game are perceived by the organizati<strong>on</strong>s as too restrictive, they may attempt to modify the<br />
instituti<strong>on</strong>al setting. The driving force behind changes is, according to North, learning by<br />
individuals. And it is argued that competiti<strong>on</strong> c<strong>on</strong>tributes significantly to the learning process<br />
given that learning is indispensable for survival. That is, the greater the degree of competiti<strong>on</strong>,<br />
the bigger the incentive for learning.<br />
In a series of seminal studies, Douglass North (1990,1991) underscores the importance of social<br />
<strong>and</strong> ec<strong>on</strong>omic instituti<strong>on</strong>s, which are likely to affect innovati<strong>on</strong> <strong>and</strong> hence l<strong>on</strong>g-term ec<strong>on</strong>omic<br />
development. According to North, instituti<strong>on</strong>s form the incentive structure of a society, <strong>and</strong> as a<br />
c<strong>on</strong>sequence, the political <strong>and</strong> ec<strong>on</strong>omic instituti<strong>on</strong>s are the underlying determinants of ec<strong>on</strong>omic<br />
performance.<br />
The instituti<strong>on</strong>al envir<strong>on</strong>ment provides a general framework for social <strong>and</strong> ec<strong>on</strong>omic interacti<strong>on</strong>s.<br />
Instituti<strong>on</strong>s fall into two categories. The first category can be referred to as formal c<strong>on</strong>straints <strong>and</strong><br />
is composed of rules, laws, c<strong>on</strong>stituti<strong>on</strong>s, property rights, whereas the sec<strong>on</strong>d <strong>on</strong>e is labeled as<br />
informal c<strong>on</strong>straints <strong>and</strong> is made up of customs, traditi<strong>on</strong>s, behavior, c<strong>on</strong>venti<strong>on</strong>s, codes of<br />
c<strong>on</strong>ducts etc.<br />
In North (1994), the evoluti<strong>on</strong> of instituti<strong>on</strong>s is analyzed in a historical perspective. It is argued that<br />
there are some primitive instituti<strong>on</strong>al settings that are most unlikely to experience changes<br />
triggered by pressures coming from the inside. In three types of exchanges under c<strong>on</strong>siderati<strong>on</strong>,<br />
namely tribal society, regi<strong>on</strong>al ec<strong>on</strong>omy <strong>and</strong> l<strong>on</strong>g-distance caravan trade, learning, <strong>and</strong> hence the<br />
accumulati<strong>on</strong> of knowledge <strong>and</strong> skill will not lead to changes, simply because innovati<strong>on</strong> is<br />
perceived as something endangering the system’s survival. This is also borne out by Posner<br />
(1980).<br />
North (1968) also explores the impacts that changes in the instituti<strong>on</strong>al envir<strong>on</strong>ment have <strong>on</strong><br />
technological progress. He analyses the case of ocean transportati<strong>on</strong> <strong>and</strong> finds that from 1600 to<br />
1860, the costs of ocean shipping dropped by 50 per cent due to huge increases in total factor<br />
productivity. The questi<strong>on</strong> is whether productivity gains are driven by technological advances or<br />
by other factors.<br />
North argues that in the period 1600-1784 productivity advances were slow, <strong>and</strong> virtually all the<br />
productivity growth came from decreased crew size <strong>and</strong> the better use of shipping time, that is the<br />
reducti<strong>on</strong> of time spent in harbors. This productivity advance may be understood within the<br />
evoluti<strong>on</strong>ary ec<strong>on</strong>omics framework as an improvement in organizati<strong>on</strong>al routines <strong>and</strong> working<br />
practices.<br />
55<br />
273
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Between 1814 <strong>and</strong> 1860, North finds that total factor productivity experienced a c<strong>on</strong>siderable<br />
accelerati<strong>on</strong> compared with the previous period, <strong>and</strong> grew roughly 10 times faster than before.<br />
North argues that the bulk of these productivity gains can be attributed to larger ships <strong>and</strong> to an<br />
increased load factor.<br />
Essentially, an increase in the size of the market <strong>and</strong> a sharp reducti<strong>on</strong> in piracy explain these<br />
changes in the characteristics of ships. On <strong>on</strong>e h<strong>and</strong>, shipping capacity was significantly better<br />
used because of immigrants pouring in America: Ships that used to return to America without<br />
carrying anything but the ballast were transporting men. On the other h<strong>and</strong>, security became<br />
better because of the eliminati<strong>on</strong> of piracy. This led to a reducti<strong>on</strong> in the need for military pers<strong>on</strong>al<br />
<strong>and</strong> equipment <strong>on</strong> board. All in all, North emphasizes that productivity gains were not due to<br />
technological progress but rather the c<strong>on</strong>sequence of organizati<strong>on</strong>al innovati<strong>on</strong>s <strong>and</strong> changes in<br />
the instituti<strong>on</strong>al envir<strong>on</strong>ment.<br />
With regards to the first industrial revoluti<strong>on</strong> in 18th <strong>and</strong> 19th century Britain, L<strong>and</strong>es (1994) argues<br />
that it was not pure accident that the first industrial revoluti<strong>on</strong> occurred in Britain <strong>and</strong> not<br />
elsewhere, e.g. in France. This is because of the structure of the industry <strong>and</strong> the instituti<strong>on</strong>al<br />
setting, which made it more likely that major inventi<strong>on</strong>s would be produced in Britain. However,<br />
Crafts (1995a,b) disagrees with this view <strong>and</strong> suggests that major inventi<strong>on</strong>s follow a stochastic<br />
process <strong>and</strong> thus those inventi<strong>on</strong>s could have occurred with roughly the same probability in<br />
France than in Britain given that the things in comm<strong>on</strong> between the two countries outweighs the<br />
differences between them.<br />
According to Mokyr (1993), two types of inventi<strong>on</strong>s can be distinguished. The first is<br />
“macro inventi<strong>on</strong>” <strong>and</strong> describes inventi<strong>on</strong>s of paramount importance for the ec<strong>on</strong>omy as a whole.<br />
The sec<strong>on</strong>d is “micro inventi<strong>on</strong>” <strong>and</strong> c<strong>on</strong>sists of small steps by which already extant techniques<br />
are ameliorated. Mokyr (1993) points out that macro inventi<strong>on</strong>s “… do not seem to obey obvious<br />
laws, do not necessarily resp<strong>on</strong>d to incentives, <strong>and</strong> defy most attempts to relate them to<br />
exogenous ec<strong>on</strong>omic variables. Many of them resulted from strokes of genius, luck or<br />
serendipity.” Mokyr notes that Britain was by no means in a better positi<strong>on</strong> in the realm of<br />
macro inventi<strong>on</strong>s compared with France but the instituti<strong>on</strong>al <strong>and</strong> ec<strong>on</strong>omic envir<strong>on</strong>ment were<br />
more c<strong>on</strong>ducive as regards micro inventi<strong>on</strong>s.<br />
51<br />
56<br />
274
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
3 STRATEGIES OF INNOVATIVE FIRMS<br />
This secti<strong>on</strong> sets out some of the strategies applied by companies to enter new markets <strong>and</strong><br />
introduce new products. This is addressed by looking at the strategies employed to gain <strong>and</strong><br />
protect such an advantage.<br />
The secti<strong>on</strong> then looks how the optimal timing of innovati<strong>on</strong> can be used as a strategy <strong>and</strong><br />
includes a discussi<strong>on</strong> of life cycles more generally. Finally, the secti<strong>on</strong> c<strong>on</strong>siders competiti<strong>on</strong> in<br />
network industries.<br />
Innovati<strong>on</strong> <strong>and</strong> Competitive Advantage<br />
In c<strong>on</strong>sidering the strategies employed by innovative firms, it is useful to draw a distincti<strong>on</strong><br />
between two broad categories:<br />
•<br />
•<br />
how a firm gains a competitive advantage over rivals; <strong>and</strong><br />
how a firm can protect <strong>and</strong> maintain that competitive advantage.<br />
In the following secti<strong>on</strong>s, we discuss further these strategies <strong>and</strong> their relati<strong>on</strong>ship with<br />
innovati<strong>on</strong>,although it is necessary to be aware that the distincti<strong>on</strong> between these comp<strong>on</strong>ents is not<br />
clear-cut.<br />
Value creati<strong>on</strong>: how firms gain a competitive advantage<br />
The first strategy focuses <strong>on</strong> how firms can create more value, which can be achieved in principle<br />
through:<br />
•<br />
•<br />
•<br />
cost reducti<strong>on</strong> underpinned by efficiency improvements;<br />
increasing quality <strong>and</strong> value of existing products <strong>and</strong> services via differentiati<strong>on</strong> relative to<br />
competitors; <strong>and</strong><br />
developing new products <strong>and</strong> services.<br />
At the centre of value creati<strong>on</strong> is the noti<strong>on</strong> of competitive advantage popularized by Porter<br />
(1985). Porter’s view is this:<br />
“Competitive advantage is at the heart of the firm’s performance in competitive<br />
markets.After several decades of vigorous expansi<strong>on</strong> <strong>and</strong> prosperity, however, many<br />
firms lost sight of competitive advantage in their scramble for growth <strong>and</strong> pursuit of<br />
diversificati<strong>on</strong>. Today the importance of competitive advantage could hardly be<br />
greater. Firms throughout the world face slower growth as well as domestic <strong>and</strong> global<br />
competitors that are no l<strong>on</strong>ger acting as if the exp<strong>and</strong>ing pie were big enough for all.”<br />
In accordance with Porter (1985), competiti<strong>on</strong> is a c<strong>on</strong>tinuous quest for competitive advantage.<br />
57<br />
275
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
External <strong>and</strong> internal sources of competitive advantage can be identified. External sources such<br />
as changing c<strong>on</strong>sumer patterns or changes in technology are exogenous to firms. Since firms<br />
are heterogeneous in their resources <strong>and</strong> capabilities, they are not equally able to adapt to such<br />
changes, <strong>and</strong> thus some firms are able to realize competitive advantage through more timely <strong>and</strong><br />
efficient resp<strong>on</strong>ses to exogenous factors.<br />
The internal source of competitive advantage is essentially the capability to innovate, by which we<br />
mean inventive effort, rather than adaptive or imitative behavior.<br />
Cost leadership<br />
As highlighted above, value to customers can be created by means of decreasing costs, through<br />
which a firm is able to reduce prices relative to competitors.<br />
At the extreme, focus <strong>on</strong> this strategy can be described as thriving for cost leadership, which<br />
refers to the situati<strong>on</strong> when a firm has the capability to produce similar products at significantly<br />
lower cost than its competitors. There are several routes through which a firm can achieve cost<br />
advantage, for example:<br />
•<br />
•<br />
Ec<strong>on</strong>omies of scale <strong>and</strong> scope yield benefits to the extent that an increase in producti<strong>on</strong><br />
volume will entail a reducti<strong>on</strong> in unit cost.<br />
Ec<strong>on</strong>omies of learning describe experience-based learning. The more complex the<br />
process technology or the product, especially in terms of tacit knowledge (i.e. that which<br />
cannot be codified <strong>and</strong> transferred) the larger the scope for benefits from learning-bydoing.<br />
This provides means for established companies to experience cost advantage<br />
over new firms.<br />
Enhancing process technology <strong>and</strong> process design allows for greater productive<br />
efficiency <strong>and</strong> lower producti<strong>on</strong> costs for the same output.<br />
Similarly, enhancing the organizati<strong>on</strong>al efficiency <strong>and</strong> routines (for a given technology) will<br />
increase productive efficiency.<br />
Reducing input costs allows for lower unit costs for the same level of productivity.<br />
Because of the use of different suppliers, geographical differences in input prices <strong>and</strong><br />
different bargaining power, different firms may pay a different price for the same set of<br />
inputs.<br />
Augmenting capacity utilizati<strong>on</strong>: In the presence of high fixed costs, unit costs are to be<br />
decreased by increasing capacity utilizati<strong>on</strong>. On the other h<strong>and</strong>, overcapacity resulting in<br />
overtime pay, premiums for night shifts <strong>and</strong> rising maintenance costs also increase unit<br />
costs.<br />
58<br />
276
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Product quality <strong>and</strong> differentiati<strong>on</strong><br />
Changes to product quality (in particular quality-price trade-offs) <strong>and</strong> product differentiati<strong>on</strong><br />
compared to rival products represent the sec<strong>on</strong>d facet of value creati<strong>on</strong>.<br />
A firm can set out to differentiate its product according to the tangible <strong>and</strong> intangible aspects of<br />
the products it manufactures. The former c<strong>on</strong>cerns physical characteristics <strong>and</strong> performance of<br />
the product, while the latter relates to perceived, mainly social <strong>and</strong> psychological faculties.<br />
The nature of competiti<strong>on</strong> in a market will be associated with whether firms strive to gain<br />
competitive advantage through innovati<strong>on</strong> in physical attributes of the products they supply, or the<br />
intangible characteristics. For example, where an industry is mature <strong>and</strong> there appears little<br />
scope for changes in product innovati<strong>on</strong> <strong>and</strong> cost leadership, competitive advantage may be<br />
achieved through a successful br<strong>and</strong>ing exercise that provides differentiati<strong>on</strong> relative to rivals.<br />
Product innovati<strong>on</strong><br />
While changes to the quality of a product represent some degree of product innovati<strong>on</strong>,<br />
innovative behavior is likely to be greater with regard to the introducti<strong>on</strong> of new products <strong>and</strong><br />
services that have features to mark them as distinct from current products.<br />
A strategy of product innovati<strong>on</strong> may have the effect of creating new markets (in competiti<strong>on</strong><br />
policy terms, the introducti<strong>on</strong> or products for which there are no good substitutes <strong>on</strong> the dem<strong>and</strong>side)<br />
but may also, though not necessarily, destroy old markets (as c<strong>on</strong>sumers cease<br />
c<strong>on</strong>sumpti<strong>on</strong> of previous generati<strong>on</strong> of products in favor of the new products).<br />
When it comes to assessing how to exploit an innovati<strong>on</strong> to the maximum, that is to maximize<br />
profits related to the innovati<strong>on</strong>, firms can choose from several opti<strong>on</strong>s depending <strong>on</strong><br />
•<br />
•<br />
How much risk can or are they willing to take, <strong>and</strong> thus how high a return do they expect;<br />
How many resources can or are they willing to put at disposal when exploiting the<br />
innovati<strong>on</strong>.<br />
Risks are high in emerging, innovative industries. First, firms have to face technological<br />
uncertainties (such as the directi<strong>on</strong> technology develops in <strong>and</strong> which technical st<strong>and</strong>ards are<br />
established in the industry). Sec<strong>on</strong>d, there are substantial market uncertainties. In other words, it<br />
is difficult to predict the potential size <strong>and</strong> expansi<strong>on</strong> of the market. N<strong>on</strong>etheless, these risks are<br />
manageable to some extent via:<br />
•<br />
•<br />
•<br />
Co-operati<strong>on</strong> with lead users<br />
Limiting financial risks related to financing innovati<strong>on</strong><br />
Maintaining a high degree of flexibility <strong>and</strong> reactivity.<br />
59<br />
277
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Grant (2002) enumerates five different strategies through which innovative firms can reap the<br />
benefits of their innovati<strong>on</strong>s. These strategies, shown in Table 3.2, can be associated with<br />
differing involvement of he firms, different risk-taking, resource requirements <strong>and</strong> firm size.<br />
Strategies in Innovative <strong>Markets</strong><br />
Strategy<br />
Internal<br />
commercializati<strong>on</strong><br />
Joint venture<br />
Strategic alliance<br />
Outsourcing<br />
Licensing<br />
Risk<br />
Necessary firm size Resource requirement<br />
High risk, high returns <strong>and</strong><br />
Medium-large High<br />
complete c<strong>on</strong>trol<br />
Sharing investment <strong>and</strong> risk but<br />
also risk of c<strong>on</strong>flict with the partner Medium-large Resource <strong>and</strong><br />
capability synergies<br />
company<br />
Medium am<strong>on</strong>g several firms<br />
Limits risk but increases<br />
dependence <strong>on</strong> the outside world<br />
Small risk <strong>and</strong> small returns<br />
Small<br />
Small<br />
Access to outside<br />
resources<br />
Low<br />
Source: Adapted from Grant (2002 p342).<br />
Each of these strategies is described below:<br />
• The strategy of internal commercializati<strong>on</strong> means that the firm commercializes the new<br />
product — that is, the output of the innovati<strong>on</strong> — without any external productive help.<br />
This is the highest level of involvement that goes h<strong>and</strong> in h<strong>and</strong> with the heaviest<br />
investments <strong>and</strong> therefore the highest risk. Nevertheless, the firms can have the whole<br />
process under its c<strong>on</strong>trol. In doing so, the resource requirements are substantial, which<br />
implies that the firm is expected to be fairly large to be able to bear the burdens.<br />
•<br />
•<br />
•<br />
•<br />
The decisi<strong>on</strong> to start a joint venture still implies a great deal of involvement. By c<strong>on</strong>trast, it<br />
also helps to share the cost <strong>and</strong> the risk of the investment. In additi<strong>on</strong> to that, the<br />
participating firms can pool their resources together <strong>and</strong> thus benefit from synergies. On<br />
the other h<strong>and</strong>, however, they run the risk of potential disagreement <strong>and</strong> possible<br />
differences in corporate culture.<br />
Strategic alliance represents a middle way between total <strong>and</strong> very small involvement.<br />
Risk <strong>and</strong> resources can be shared.<br />
Outsourcing is suited to both large <strong>and</strong> small companies so as to enable access to<br />
outside resources without too much commitment.<br />
Licensing is typically — but not exclusively — employed by small firms, which do not<br />
possess the necessary resources to exploit the innovati<strong>on</strong> <strong>on</strong> their own.<br />
Protecting competitive advantage<br />
The sec<strong>on</strong>d strategy can be understood as efforts to protect <strong>and</strong> maintain a competitive<br />
advantage that has been achieved through value creati<strong>on</strong>. In particular, it c<strong>on</strong>cerns efforts firms<br />
60<br />
278
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
may take to prevent imitati<strong>on</strong> by rivals, since in competitive envir<strong>on</strong>ments, imitati<strong>on</strong> is a crucial<br />
force that undermines the competitive advantage of a firm over time. Such behavior may be<br />
both legal (e.g. patenting) <strong>and</strong> illegal (e.g. certain types of anticompetitive behavior that inhibit<br />
rivals access to a market).<br />
Once a firm has been able to gain competitive advantage in some respect, it will have incentives<br />
to protect <strong>and</strong> maintain this. In particular, where a firm has benefited relative to competitors from<br />
an important innovati<strong>on</strong> it will be keen that rival firms do not undermine this advantage through<br />
imitati<strong>on</strong>.<br />
One way in which firms achieve such protecti<strong>on</strong> is through various legal instituti<strong>on</strong>s put in place<br />
for this purpose (e.g. patent policy). These are highlighted below before we turn to other ways in<br />
which a source of competitive advantage can be maintained.<br />
The use of legal instituti<strong>on</strong>s against imitati<strong>on</strong><br />
Intellectual property rights are legal instituti<strong>on</strong>s that may protect the competitive advantage a firm<br />
has realized through innovati<strong>on</strong> by restricting the degree of imitati<strong>on</strong> of the innovati<strong>on</strong> that can<br />
take place.<br />
Patent policy provides a firm with protecti<strong>on</strong> against direct imitati<strong>on</strong> of certain types of inventive<br />
innovati<strong>on</strong>, for a period of time, subject to disclosure of informati<strong>on</strong> relating to the innovati<strong>on</strong>.<br />
Similarly, copyright protecti<strong>on</strong> prevents the direct copying of new c<strong>on</strong>tent. In both cases the aim<br />
of the legal protecti<strong>on</strong> is to provide incentives for the innovati<strong>on</strong> to take place, exactly because in<br />
the absence of this protecti<strong>on</strong> such incentives for such innovati<strong>on</strong> would be mitigated by the<br />
threat of direct imitati<strong>on</strong> <strong>and</strong> risk of immediate erosi<strong>on</strong> of any competitive advantage accruing to<br />
the innovati<strong>on</strong>.<br />
N<strong>on</strong>etheless, there are ways in which patent policy may be used by firms against the interests of<br />
c<strong>on</strong>sumers. Pre-emptive patenting describes the case where firms gather a series of patent<br />
around an initial innovati<strong>on</strong>, not for the purposes of using those patents in the introducti<strong>on</strong> of new<br />
products, but rather to stop other firms form inventing around the original innovati<strong>on</strong>. Thus the<br />
new patented products or technologies are neither used by the patentee nor licensed to other<br />
firms, but their value derives from the fact that rivals cannot use them rather than the use the<br />
holder makes of them.<br />
Alternatively, where a particular product or service protected by patent represents a bottleneck to<br />
a range of related markets, the patent holder may be able to use its protecti<strong>on</strong> in the supply of the<br />
patented product to enjoy competitive advantage in the supply of products in these related<br />
markets.<br />
In some sense, there is overlap between creati<strong>on</strong> of competitive advantage through product<br />
differentiati<strong>on</strong> <strong>and</strong> maintenance of this advantage through br<strong>and</strong>ing. Where firms can link an<br />
innovative product to a successful br<strong>and</strong> image, <strong>and</strong> protect that br<strong>and</strong> from imitati<strong>on</strong> (e.g. via<br />
trademark law), the advantage from the initial innovati<strong>on</strong> can be maintained in the face of<br />
imitati<strong>on</strong> as to the products physical characteristics. For example, in the pharmaceutical industry,<br />
61<br />
279
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
successful creati<strong>on</strong> of goodwill, reputati<strong>on</strong> <strong>and</strong> br<strong>and</strong> recogniti<strong>on</strong> allow firms to enjoy some (albeit<br />
less) competitive advantage over a drug that they have created <strong>on</strong>ce the patent has expired.<br />
Other protecti<strong>on</strong> against innovati<strong>on</strong><br />
Besides the use of intellectual property rights, the competitive advantage associated with an<br />
innovati<strong>on</strong> can be obtained through other mechanisms.<br />
The first strategy is secrecy. Where an innovati<strong>on</strong> is hard to observe or hard to codify, the<br />
innovative firm may able to maintain competitive advantage simply by not facilitating the flow of<br />
informati<strong>on</strong> about the innovati<strong>on</strong> to competitors.<br />
A sec<strong>on</strong>d strategy to protect competitive advantage is to merge with rivals. This could provide<br />
protecti<strong>on</strong> where a particular rival is capable of reducing the competitive advantage a firm has<br />
gained through innovati<strong>on</strong>. Of course, the merger could bring other benefits that are associated<br />
with creati<strong>on</strong> rather than maintenance of competitive advantage. For example, the combinati<strong>on</strong><br />
of tacit knowledge in the merged parties may bring greater capability to innovate.<br />
Finally, a firm may be able to with sufficient market power may be able to protect its competitive<br />
advantage by using this power to restrict rivals’ ability to compete with it. Besides predati<strong>on</strong>, a<br />
firm may be able to use competitive advantage with regard to <strong>on</strong>e good or service, <strong>and</strong> spread<br />
this advantage to related services, for example:<br />
• denying the rival access to some important input service, ranging from necessary raw<br />
materials <strong>and</strong> comp<strong>on</strong>ents to services that allow supply of a service to particular<br />
customers, or even IP rights; or<br />
•<br />
bundling or tying products such that rivals are forced to compete against the bundle of<br />
services offered by a firm <strong>and</strong> the competitive advantage the firm holds is transferred<br />
across the range of products supplied together.<br />
Not <strong>on</strong>ly could such a strategy help maintain the competitive advantage from the original line of<br />
innovati<strong>on</strong>, but it may also serve to exp<strong>and</strong> the ec<strong>on</strong>omic area of which returns from the<br />
competitive advantage can be enjoyed. However, it is essential to recognize that such strategies<br />
may also be c<strong>on</strong>ducive to value generati<strong>on</strong>. For instance, with respect to bundling, this pricing<br />
opti<strong>on</strong> may be the <strong>on</strong>ly way in which sufficient revenue could be generated to cover the initial<br />
outlay required for an innovati<strong>on</strong>.<br />
3.2 Timing of Innovati<strong>on</strong> <strong>and</strong> Life Cycles<br />
Having set out firm strategies according to two broad categories, we discuss below issues of<br />
timing. The first sub-secti<strong>on</strong> c<strong>on</strong>siders how the success of innovati<strong>on</strong>, <strong>and</strong> thus the optimal<br />
innovati<strong>on</strong> strategy, may be dependant <strong>on</strong> timing. The sec<strong>on</strong>d sub-secti<strong>on</strong> c<strong>on</strong>siders product<br />
‘life-cycles’ more generally.<br />
62<br />
280
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
First-mover advantage<br />
An essential questi<strong>on</strong> in innovative markets is to choose a strategy whether to enter early in the<br />
market <strong>and</strong> invest heavily in research <strong>and</strong> development in the hope of future profits <strong>and</strong> market<br />
leadership or to adopt a wait-<strong>and</strong>-see policy <strong>and</strong> to follow the technology leader by copying the<br />
innovati<strong>on</strong>. The benefits of a first-mover strategy relies <strong>on</strong> three c<strong>on</strong>diti<strong>on</strong>s:<br />
•<br />
•<br />
•<br />
Length of the lead time (i.e. the time over which the innovati<strong>on</strong> is protected from imitators<br />
<strong>and</strong> followers) either through IP rights or less formally.<br />
The extent to which complementary resources are required while exploiting the<br />
innovati<strong>on</strong>. If the need for complementary resources is relatively low, the initial investment<br />
<strong>and</strong> therefore the risk is lower. This may stimulate firms to try to be the first-mover.<br />
The possibility to set a st<strong>and</strong>ard is also likely to give an impetus for first-movers.<br />
The reas<strong>on</strong> why st<strong>and</strong>ards are set is closely c<strong>on</strong>nected with positive network externalities<br />
(discussed further below in the c<strong>on</strong>text of the ec<strong>on</strong>omics of networks). The higher the number of<br />
users of the st<strong>and</strong>ardized product, the more valuable the product to individual user. Positive<br />
network externalities have three major sources:<br />
•<br />
•<br />
•<br />
users of the product are c<strong>on</strong>nected with each other through networks (such as<br />
transportati<strong>on</strong> <strong>and</strong> the internet);<br />
the ease with which complementary products are available (for example, software<br />
applicati<strong>on</strong>s for Windows); <strong>and</strong><br />
switching cost between different networks: the larger the network the customer uses, the<br />
lower the costs to switch to another network (for example, mobile teleph<strong>on</strong>y networks).<br />
Generally, technological st<strong>and</strong>ards tend to result in ever increasing positive network externalities.<br />
If a technological st<strong>and</strong>ard appears to be dominant in the industry, more <strong>and</strong> more users are likely<br />
to choose this <strong>on</strong>e that may easily lead to winner-takes-all situati<strong>on</strong>. This is <strong>on</strong>e reas<strong>on</strong> why<br />
innovative firms try hard to establish their technology as the industry st<strong>and</strong>ard.<br />
Shapiro <strong>and</strong> Varian (1999) identify three possible <strong>and</strong> highly complementary strategies how firms<br />
could establish their own technology as the st<strong>and</strong>ard. The first strategy builds <strong>on</strong> obtaining the<br />
support of other, rival companies <strong>and</strong> firms producing complementary products, possibly through<br />
tacit agreements. The sec<strong>on</strong>d approach c<strong>on</strong>sists of market pre-empti<strong>on</strong>. Finally, <strong>and</strong> importantly,<br />
the firm aiming at winning the st<strong>and</strong>ard war has to make the impressi<strong>on</strong> from the very outset that<br />
its st<strong>and</strong>ard will be accepted at the end of the st<strong>and</strong>ard war. Put it another way, marketing <strong>and</strong><br />
other communicati<strong>on</strong> techniques can be employed so as to create self-fulfilling expectati<strong>on</strong>s<br />
am<strong>on</strong>g customers <strong>and</strong> rivals.<br />
However, different firms may have different strategic windows. These are defined as the time<br />
period during which they are able to seize market opportunities in line with their resources <strong>and</strong><br />
capabilities. Smaller firms tend to have shorter strategic windows because they cannot wait for a<br />
63<br />
281
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
l<strong>on</strong>g time to introduce new innovati<strong>on</strong>s <strong>and</strong> procure competitive advantage as opposed to large<br />
corporati<strong>on</strong> that do not tend to rush in introducing new technologies <strong>and</strong> take unnecessary risks<br />
given their solid financial background.<br />
The role of the life-cycle of an industry<br />
The positi<strong>on</strong> in the life cycle of a particular product, market or industry is likely to have a role to<br />
play as regards the strategy adopted by any firm towards the creati<strong>on</strong> <strong>and</strong> protecti<strong>on</strong> of<br />
competitive advantage. Life cycle theory sets out a useful analytical framework for analyzing this.<br />
In this c<strong>on</strong>text, there is some ambiguity in the relevant literature regarding terminology between<br />
product life cycles <strong>and</strong> industry life cycles. In principle these are different c<strong>on</strong>cepts since industry<br />
usually refers to a wider set of entities that a product, but these terms refer broadly to the same<br />
approach <strong>and</strong> are sometimes used interchangeably. In some sense the c<strong>on</strong>cepts of an industry<br />
life cycle is associated with the industrial organizati<strong>on</strong> literature (B<strong>on</strong>accorsi <strong>and</strong> Giuri (2000),<br />
Horvath et al. (2000) <strong>and</strong> Klepper <strong>and</strong> Kenneth (2001)), while marketing literature may be more<br />
likely to use term product life-cycle. In this overview we refer to product life cycles.<br />
Based <strong>on</strong> a series of case studies <strong>and</strong> empirical papers, Klepper (1996, pp. 564-565) comes up<br />
with five major regularities that best characterize the product life cycle:<br />
•<br />
•<br />
•<br />
•<br />
•<br />
The number of new entrants rises very rapidly at the beginning <strong>and</strong> then starts declining<br />
<strong>and</strong> eventually stabilizes at a very low level.<br />
The number of incumbents increases in the initial phase <strong>and</strong> then starts falling<br />
accompanied by a steady increase in output.<br />
Initially, market shares change quickly, i.e. some firms grow fast <strong>and</strong> others loose ground.<br />
However, after a while, market shares stabilize.<br />
Simultaneously, there is an increase in the diversity of products competing in the market<br />
because of a rise in product innovati<strong>on</strong>.<br />
However, as the market gets more mature <strong>and</strong> with the c<strong>on</strong>solidati<strong>on</strong> of the firms,<br />
incumbents spill more energy <strong>on</strong> process innovati<strong>on</strong>.<br />
Werker (2003) provides two more stylized facts associated with the product life cycle complement<br />
the analysis above. First that prices fall sharply at the outset of the life cycle with st<strong>and</strong>ardizati<strong>on</strong><br />
<strong>and</strong> a growing number of customers. However, later <strong>on</strong>, prices decrease at a lower pace <strong>and</strong> the<br />
decrease finally comes to a halt. Sec<strong>on</strong>d, that the beginning of the life cycle favors entry<br />
whereas established firms are favored to new entrants at later stages of the life cycle.<br />
More generally, the life cycle of a product may typically have four phases: (i) introducti<strong>on</strong>; (ii)<br />
shakeout; (iii) maturity; <strong>and</strong> (iv) decline.<br />
During the first phase, i.e. the introducti<strong>on</strong>, a few companies produce a wide range of goods with<br />
substantial differences in product features partly <strong>on</strong> the grounds of differences in the underlying<br />
64<br />
282
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
technology. The competiti<strong>on</strong> am<strong>on</strong>g technologies yields rapid product innovati<strong>on</strong>. Werker (2003)<br />
notes that at this stage of the product cycle, c<strong>on</strong>sumers’ preferences are not well established.<br />
Therefore, firms exert effort to discover the unclear preferences through high rates of innovati<strong>on</strong>.<br />
The sec<strong>on</strong>d stage of the life cycle usually witnesses a shakeout <strong>and</strong> the emergence of a<br />
dominant technology, accompanied by st<strong>and</strong>ardizati<strong>on</strong> of around this technology. This lays the<br />
ground for further rapid innovati<strong>on</strong>. The competitive envir<strong>on</strong>ment forces unsuccessful firms to exit<br />
the market <strong>and</strong> leads to mergers <strong>and</strong> acquisiti<strong>on</strong>s. Product differentiati<strong>on</strong> in this phase can be<br />
achieved mainly through new product design <strong>and</strong> by c<strong>on</strong>tinually improving quality.<br />
For instance, Klepper <strong>and</strong> Sim<strong>on</strong>s (2001) provide empirical evidence for dramatic shakeouts that<br />
occurred at some point in the automobiles, tires, televisi<strong>on</strong>s <strong>and</strong> penicillin industries in the US.<br />
They suggest that earlier entrants are more likely to survive shakeouts. Shakeouts do not seem<br />
to be triggered by drastic events but rather the c<strong>on</strong>solidati<strong>on</strong> in these industries was a result of the<br />
process of competiti<strong>on</strong> during which early entrants gain high market shares through innovati<strong>on</strong>. A<br />
similar c<strong>on</strong>clusi<strong>on</strong> is drawn by Horvath et al. (2000), who find that the shakeout in the US beer<br />
brewing industry occurred from 1880 to 1890 was a c<strong>on</strong>sequence of earlier mass entry <strong>and</strong> the<br />
subsequent competiti<strong>on</strong>.<br />
When an industry reaches maturity, products tend to become increasingly homogeneous <strong>and</strong><br />
could be viewed as commodities. In this c<strong>on</strong>text, <strong>and</strong> to counter this tendency, br<strong>and</strong>ing <strong>and</strong> the<br />
introducti<strong>on</strong> of complementary services represent an effective way to differentiate products. Price<br />
competiti<strong>on</strong> also tends to become more intense in this phase of the life cycle. Hence, cost<br />
advantage is a crucial aspect of competitive advantage. Simultaneously, the technology<br />
developed in earlier stages becomes diffused <strong>and</strong> firms seek to improve this technology<br />
incrementally.<br />
The final phase can be referred to as the decline of the industry. During this phase fierce <strong>and</strong><br />
possibly destructive price wars become the norm given that product differentiati<strong>on</strong> is extremely<br />
difficult. The lack or at least the very low pace of product <strong>and</strong> process innovati<strong>on</strong> is <strong>on</strong>e reas<strong>on</strong><br />
for this. As a result, some firms decide to exit from the market <strong>and</strong> this further c<strong>on</strong>tributes to the<br />
decline.<br />
Different industries may be very different as regards the intensity of competiti<strong>on</strong> <strong>and</strong> the pace of<br />
innovati<strong>on</strong> in the same stage of the life cycle. Accordingly, three types of industries can be<br />
identified:<br />
• In local m<strong>on</strong>opoly markets, products are very specialized <strong>and</strong> target narrow market<br />
segments. Products are of high quality with high unit values, which implies that volume of<br />
producti<strong>on</strong> is low. Because of substantial differentiati<strong>on</strong>, competiti<strong>on</strong> is very limited.<br />
Examples for this kind of market are found am<strong>on</strong>gst the following areas: highway <strong>and</strong><br />
residential c<strong>on</strong>structi<strong>on</strong>, surgical appliances, defense industry.<br />
•<br />
Traditi<strong>on</strong>al industrial markets are characterized by homogenous products where<br />
competitive advantage can be usually achieved via cost leadership, br<strong>and</strong>ing or product<br />
65<br />
283
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
variety. Passenger cars <strong>and</strong> household equipment are examples of products bel<strong>on</strong>ging to<br />
this category<br />
• Schumpeterian markets are dominated by product innovati<strong>on</strong>, where incumbents can be<br />
completely displaced by entrants. These markets might be seen within the computer<br />
software industry <strong>and</strong> within c<strong>on</strong>sumer electr<strong>on</strong>ics.<br />
3.3 Issues Relating to Network Industries<br />
Network industries give rise to a number of issues that are likely to affect — or at least influence<br />
— the strategies used by firms to innovate. The characteristic features of these industries<br />
originate from network externalities. A positive network externality arises when “a good is more<br />
valuable to a user the more users adopt the same good or compatible <strong>on</strong>es”. A comm<strong>on</strong><br />
distincti<strong>on</strong> is that between direct <strong>and</strong> indirect network externalities.<br />
Direct externalities typically occur in a two-way (physical) network. This externality reflects the<br />
fact, for instance, that a teleph<strong>on</strong>e user benefits from others being c<strong>on</strong>nected to the same<br />
network: an additi<strong>on</strong>al teleph<strong>on</strong>e increases the number of potential communicati<strong>on</strong> within the<br />
system, <strong>and</strong> thus the value of membership.<br />
Indirect externalities arise from the fact that a network of users increases the incentives to<br />
produce compatible products that are complementary with the network good. These externalities<br />
arise in systems, which can be defined as “[…] collecti<strong>on</strong>s of two or more comp<strong>on</strong>ents together<br />
with an interface that allows the comp<strong>on</strong>ents to work together” (Katz <strong>and</strong> Shapiro, 1994, p93).<br />
The typical system comprises of a “platform” (or hardware) <strong>and</strong> of “applicati<strong>on</strong>s” (or software) that<br />
can <strong>on</strong>ly be used with a platform. Compatibility refers to the possibility that software can be used<br />
with a particular platform. The network externality arises when users make their purchase over<br />
time, because in the presence of ec<strong>on</strong>omies of scale a greater number of complementary<br />
products can be supplied at a lower price when the network grows.<br />
The existence of network externalities has a substantial impact <strong>on</strong> the way technologies are<br />
chosen <strong>and</strong> promoted. In this secti<strong>on</strong> we follow closely Katz <strong>and</strong> Shapiro (1994) overview of<br />
systems competiti<strong>on</strong>s <strong>and</strong> their identificati<strong>on</strong> of three main issues that have been addressed in<br />
the literature:<br />
•<br />
•<br />
technology adopti<strong>on</strong> decisi<strong>on</strong>s;<br />
product selecti<strong>on</strong> decisi<strong>on</strong>s; <strong>and</strong><br />
.<br />
66<br />
284
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
• compatibility decisi<strong>on</strong>s.<br />
<strong>Technology</strong> adopti<strong>on</strong> decisi<strong>on</strong>s<br />
A number of issues that arise in markets characterized by network externalities can be addressed<br />
by c<strong>on</strong>sidering a system <strong>on</strong>ly, setting aside competiti<strong>on</strong> between different systems.<br />
The existence of direct network externalities drives a wedge between private <strong>and</strong> social<br />
incentives to join the network. A single user does not c<strong>on</strong>sider the benefits that would accrue to<br />
others by his joining the network <strong>and</strong> as a result the market may lead to network of inefficient<br />
small size. In fact, the benefits that an individual would take into c<strong>on</strong>siderati<strong>on</strong> when choosing<br />
whether to join a network depends not <strong>on</strong>ly <strong>on</strong> the current size of the network, but also <strong>on</strong> its<br />
expected future size.<br />
C<strong>on</strong>sumers’ expectati<strong>on</strong>s play a central role in driving market outcomes where there are<br />
technology decisi<strong>on</strong>s in markets with important network effects. An important c<strong>on</strong>sequence is that<br />
multiple equilibria can easily arise. For instance, if all c<strong>on</strong>sumers expect no <strong>on</strong>e else to join the<br />
network, then its size would be zero, even if the network may be valuable for c<strong>on</strong>sumers; if all<br />
expect every<strong>on</strong>e to join, the network may achieve a large size.<br />
Indirect externalities may have a different impact <strong>on</strong> the size of the network. In a system market,<br />
<strong>on</strong>e c<strong>on</strong>sumer’s adopti<strong>on</strong> decisi<strong>on</strong> would does not affect other c<strong>on</strong>sumers, given the prices <strong>and</strong><br />
varieties of products available. The externalities arise indirectly through the impact of <strong>on</strong>e<br />
c<strong>on</strong>sumers’ adopti<strong>on</strong> decisi<strong>on</strong> of the future variety or prices of applicati<strong>on</strong>s. Katz <strong>and</strong> Shapiro<br />
(1994) note that in such an envir<strong>on</strong>ment, if cost c<strong>on</strong>diti<strong>on</strong>s are c<strong>on</strong>sistent with the existence of a<br />
competitive equilibrium, a perfectly competitive equilibrium arises in system markets. When, at<br />
the other extreme, a m<strong>on</strong>opolist is the sole supplier of the comp<strong>on</strong>ents of a system, there are<br />
issues regarding multi-product pricing <strong>and</strong> inter-temporal pricing.<br />
We follow Katz <strong>and</strong> Shapiro (1994) closely to discuss these issues. C<strong>on</strong>sider the case where the<br />
m<strong>on</strong>opolist can commit to the prices of both comp<strong>on</strong>ents of the system:<br />
•<br />
•<br />
with fixed proporti<strong>on</strong>s technology, the m<strong>on</strong>opolist simply sets the price that maximizes<br />
profits in the usual way.<br />
with a variable proporti<strong>on</strong>s technology, issues of bundling <strong>and</strong> price discriminati<strong>on</strong> arise<br />
but again the inefficiencies are attributable to m<strong>on</strong>opoly power rather than externalities.<br />
If the supplier cannot commit to prices in advance, buyer expectati<strong>on</strong>s are again crucial: a<br />
m<strong>on</strong>opolist would like to c<strong>on</strong>vince c<strong>on</strong>sumers applicati<strong>on</strong>s will be available at low price in the<br />
future. After c<strong>on</strong>sumers are locked in, however, the m<strong>on</strong>opolist may raise price to the m<strong>on</strong>opoly<br />
level.<br />
In the literature, it is usually assumed that c<strong>on</strong>sumers’ expectati<strong>on</strong>s <strong>on</strong> the future prices of<br />
applicati<strong>on</strong>s are affected by the quantity of hardware currently sold which may act as a signal: a<br />
larger base may lead to cheaper <strong>and</strong>/or more varied applicati<strong>on</strong>s. Hence, the m<strong>on</strong>opolist may<br />
want to lower the price of hardware to create a larger network <strong>and</strong> software aftermarket.<br />
67<br />
285
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Product selecti<strong>on</strong> decisi<strong>on</strong>s<br />
Competiti<strong>on</strong> in network industries may take the form of competiti<strong>on</strong> between incompatible<br />
systems.<br />
Network externalities may lead to a de facto st<strong>and</strong>ardizati<strong>on</strong> whereby every<strong>on</strong>e uses the same<br />
system. Due to positive feedback elements, a system may become dominant <strong>on</strong>ce it has gained<br />
an initial edge, a phenomen<strong>on</strong> which is usually referred to as “tipping”.<br />
Two main factors, however, may limit tipping <strong>and</strong> sustain multiple networks:<br />
•<br />
•<br />
network externalities being exhausted at a smaller network size;<br />
c<strong>on</strong>sumers’ heterogeneity <strong>and</strong> product differentiati<strong>on</strong>.<br />
In the latter case multiple networks would reflect c<strong>on</strong>sumers’ love for variety <strong>and</strong> there is a typical<br />
trade-off between variety <strong>and</strong> st<strong>and</strong>ardizati<strong>on</strong>.<br />
Competiti<strong>on</strong> between systems may be very intense, at least before a dominant system, if any,<br />
would emerge as dominant. For instance, firms may be engaged in very intense price<br />
competiti<strong>on</strong> at an early stage for seeking to establish an installed base <strong>and</strong> achieve leadership.<br />
This competiti<strong>on</strong> can be interpreted as firms bidding for future m<strong>on</strong>opoly profits.<br />
Competiti<strong>on</strong> may also be played in trying to affect c<strong>on</strong>sumers’ expectati<strong>on</strong>s about the dominant<br />
system that would emerge.<br />
An important implicati<strong>on</strong> of the network externalities literature is that the market may settle in an<br />
equilibrium where the dominant system is that with a lower social evaluati<strong>on</strong>. <strong>Markets</strong> can exhibit<br />
“excess inertia” <strong>and</strong> remain locked into an obsolete st<strong>and</strong>ard, even though a better <strong>on</strong>e is<br />
available.<br />
However, markets can also exhibit “excess momentum” whereby the market tips inefficiently to<br />
new technologies. This could, for instance, arise when competiti<strong>on</strong> is between an older<br />
technology which is competitively supplied <strong>and</strong> a new technology which is sp<strong>on</strong>sored. The new<br />
technology that is sp<strong>on</strong>sored may have an advantage over an older technology that is more<br />
competitively supplied because the sp<strong>on</strong>sor of the technology may engage in pricing below costs<br />
or other types of investments (with the hope of recouping these later <strong>on</strong>ce the technology is<br />
established). N<strong>on</strong>etheless, this perceived risk is not straightforward — pricing very low from the<br />
outset may be the <strong>on</strong>ly way of introducing an efficient new technology, <strong>and</strong> tipping may be<br />
inevitable.<br />
Compatibility choices<br />
Compatibility between different networks/systems is often a choice variable of firms. Two types of<br />
compatibility can be c<strong>on</strong>sidered:<br />
• horiz<strong>on</strong>tal compatibility: between two comparable rival systems; <strong>and</strong><br />
68<br />
286
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
vertical compatibility: between successive generati<strong>on</strong>s of a technology.<br />
Compatibility entails both social benefits <strong>and</strong> costs:<br />
•<br />
•<br />
•<br />
•<br />
•<br />
compatibility exp<strong>and</strong>s the size of both networks thereby avoiding the cost of participating<br />
to two different networks (e.g. duplicate equipment);<br />
compatibility in systems may lead to lower producti<strong>on</strong> costs for ec<strong>on</strong>omics of scale,<br />
learning effects, etc;<br />
compatibility enhances variety by allowing c<strong>on</strong>sumers to combine comp<strong>on</strong>ents from<br />
various systems;<br />
the risks of adopting a particular technology are lower; <strong>and</strong><br />
costs derive from the mechanism by which compatibility is achieved. St<strong>and</strong>ardizati<strong>on</strong><br />
may lead to a loss of variety <strong>and</strong> may prevent the development of new incompatible<br />
systems; adapters that allow interfacing have a cost themselves <strong>and</strong> may work<br />
imperfectly.<br />
The nature of competiti<strong>on</strong> in the market is affected by compatibility decisi<strong>on</strong>s. For systems that<br />
are compatible, competiti<strong>on</strong> is essentially at the level of each comp<strong>on</strong>ent. For incompatible<br />
systems, competiti<strong>on</strong> is at the network or system level.<br />
Katz <strong>and</strong> Shapiro (1986) c<strong>on</strong>sider the impact of compatibility <strong>on</strong> pricing competiti<strong>on</strong> over time.<br />
Price competiti<strong>on</strong> is relaxed at earlier stages of the product life cycle because the market loses its<br />
winner-take-all feature. For the same reas<strong>on</strong>, however, competiti<strong>on</strong> may be intensified in later<br />
stages of the product life cycle.<br />
Compatibility decisi<strong>on</strong>s may be affected by a number of factors:<br />
•<br />
•<br />
•<br />
asymmetries in the probability of being the winner, e.g. reputati<strong>on</strong>;<br />
product differentiati<strong>on</strong>; <strong>and</strong><br />
installed base<br />
Side payments may help firms to reach an agreement <strong>on</strong> compatibility. When they are not<br />
feasible, it is useful to distinguish markets where a firm can unilaterally impose compatibility <strong>and</strong><br />
those where a firm can unilaterally impose incompatibility.<br />
In general, there may be different mechanisms <strong>and</strong> instituti<strong>on</strong>s that govern compatibility choices:<br />
st<strong>and</strong>ards committees, unilateral acti<strong>on</strong> <strong>and</strong> hybrid mechanisms. (Farrell <strong>and</strong> Sal<strong>on</strong>er, 1988)<br />
69<br />
287
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
4<br />
4.1<br />
BIBLIOGRAPHY<br />
Industrial Organizati<strong>on</strong><br />
Arrow, K J (1962) “Ec<strong>on</strong>omic Welfare <strong>and</strong> the Allocati<strong>on</strong> of Resources for Inventi<strong>on</strong>”, in<br />
Universities Nati<strong>on</strong>al Bureau Committee for Ec<strong>on</strong>omic Research. The rate <strong>and</strong> directi<strong>on</strong> of<br />
inventive activity Princet<strong>on</strong>, Princet<strong>on</strong> University Press.<br />
Baumol,William (2002) The Free Market Innovati<strong>on</strong> Machine,Princet<strong>on</strong>,N.J.,Princet<strong>on</strong> Univ.Press<br />
Beath, J, Poyago-Theotoki, J, Ulph, D, (1998), “Organizati<strong>on</strong> design <strong>and</strong> informati<strong>on</strong>-sharing in a<br />
research joint venture with spillovers”, Bulletin of Ec<strong>on</strong>omic Research, 50:1, pp.47-59.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1987) “Sequential product innovati<strong>on</strong> <strong>and</strong><br />
industry evoluti<strong>on</strong>”, Ec<strong>on</strong>omic Journal (Supplement), Vol 99, p32-43.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1994) “Strategic R&D innovati<strong>on</strong>” in Cable,<br />
John (ed) Current Issues in Industrial Ec<strong>on</strong>omics, Basingstoke: Macmillan.<br />
Beath, John, Katsoulacos, Yannis <strong>and</strong> Ulph, David (1995) “Game-theoretical approaches to the<br />
modelling of technological change” in St<strong>on</strong>eman, Paul (ed) H<strong>and</strong>book of the Ec<strong>on</strong>omics of<br />
Innovati<strong>on</strong> <strong>and</strong> Technological Change, Oxford UK <strong>and</strong> Cambridge USA: Blackwell.<br />
Bester, H. <strong>and</strong> Petrakis, E. (1993) “The incentives for cost reducti<strong>on</strong> in a differentiated industry”,<br />
Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, No16, p519-534.<br />
Bo<strong>on</strong>e, Jan (2000) “Competitive pressure: the effects <strong>on</strong> investments in product <strong>and</strong> process<br />
innovati<strong>on</strong>”, RAND Journal of Ec<strong>on</strong>omics, Vol 31, No 3, p549-569.<br />
Bo<strong>on</strong>e, Jan (2001) “Intensity of competiti<strong>on</strong> <strong>and</strong> the incentive to innovate”, Internati<strong>on</strong>al Journal of<br />
Industrial Organizati<strong>on</strong>, Vol 19, p705-726.<br />
Cohen, W M <strong>and</strong> Levinthal, D A (1989a) “The Implicati<strong>on</strong>s of Spillovers for R&D Investment <strong>and</strong><br />
Welfare: A New Perspective” in a Link <strong>and</strong> K. Smith eds, Advances in applied micro-ec<strong>on</strong>omics<br />
vol. 5: The factors affecting technological change. Greenwich, C<strong>on</strong>n: JAI Press.<br />
Cohen, W M <strong>and</strong> Levinthal, D A (1989b) “Innovati<strong>on</strong> <strong>and</strong> learning: the two faces of R&D –<br />
implicati<strong>on</strong>s for the analysis of R&D investment”, Ec<strong>on</strong>omic Journal, Vol 99,p569-96.<br />
d’Asprem<strong>on</strong>t, C, Jacquemin, A, (1988), “Cooperative <strong>and</strong> n<strong>on</strong>-cooperative R&D in a duopoly with<br />
spillovers”, American Ec<strong>on</strong>omic Review, 78, pp.1133-1137.<br />
Dasgupta, P <strong>and</strong> Stiglitz, J E (1980a) “Industrial structure <strong>and</strong> the nature of innovative activity”,<br />
Ec<strong>on</strong>omic Journal, 90:266-293.<br />
Dasgupta, P <strong>and</strong> Stiglitz, J E (1980b) “Uncertainty, industrial structure <strong>and</strong> the speed of R&D”,<br />
Bell Journal of Ec<strong>on</strong>omics, 11:1 28.<br />
Dasgupta, P. (1986) “The theory of technological competiti<strong>on</strong>”, in J.E. Stiglitz <strong>and</strong> F. Mathews<strong>on</strong><br />
eds., New <strong>Development</strong>s in the Analysis of Market Structure, L<strong>on</strong>d<strong>on</strong>: Macmillan.<br />
70<br />
288
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Delb<strong>on</strong>o, F. <strong>and</strong> Denicolò, Vincenzo (1990) “R&D investment in a symmetric <strong>and</strong> homogeneous<br />
oligopoly: Bertr<strong>and</strong> vs Cournot”, Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, p297-313<br />
Fudenberg, D & Tirole, J (1983) “Learning-by-doing <strong>and</strong> market performance”, Bell Journal,<br />
14:522-530.<br />
Fudenberg, D., Gilbert, R, Stiglitz, J. <strong>and</strong> Tirole, Jean (1983) “Preempti<strong>on</strong>, leapfrogging <strong>and</strong><br />
competiti<strong>on</strong> in patent races”, European Ec<strong>on</strong>omic Review, Vol 22, p3-31<br />
Gilbert, R J <strong>and</strong> Newbery, D M G (1982) “Preemptive patenting <strong>and</strong> the persistence of m<strong>on</strong>opoly”,<br />
American Ec<strong>on</strong>omic Review, 72:514-526.<br />
Greenlee, P, Cassiman, B, (1999), “Product market objectives <strong>and</strong> the formati<strong>on</strong> of research joint<br />
ventures”, Managerial <strong>and</strong> Decisi<strong>on</strong> Ec<strong>on</strong>omics, 20(3), pp.115-130<br />
Grossman, G M <strong>and</strong> Shapiro, C (1987) “Dynamic R&D competiti<strong>on</strong>”, Ec<strong>on</strong>omic Journal,<br />
97:372-387.<br />
Hagedoorn, J, Schakenraad, J, (1992), “Leading companies <strong>and</strong> networks of strategic alliances<br />
in informati<strong>on</strong> technologies”, Research Policy, 21, pp. 163-190.<br />
Harris. Christopher <strong>and</strong> Vickers, John (1987) “Racing with uncertainty”, Review of Ec<strong>on</strong>omic<br />
Studies, Vol 54, No 1, p1-21<br />
Jacquemin, A, Soete, L, (1994), “Cooperati<strong>on</strong> in R&D, efficiency <strong>and</strong> European antitrust policy”,<br />
European Review, 2, pp.65-72.<br />
Judd, K L (1985b) “On The Performance of Patents”, Ec<strong>on</strong>ometrica, 53:567-585.<br />
Kamien, M, Muller, E, <strong>and</strong> Zang, I, (1992), “Research joint ventures <strong>and</strong> R&D cartels”, American<br />
Ec<strong>on</strong>omic Review, 82, pp.1293-1306.<br />
Katsoulacos, Y, Ulph, D, (1998), “Endogenous spillovers <strong>and</strong> the performance of research joint<br />
ventures”, The Journal of Industrial Ec<strong>on</strong>omics, 3, pp.333-357.<br />
Katz, L M <strong>and</strong> Shapiro, C (1986) “<strong>Technology</strong> Adopti<strong>on</strong> in the Presence of Network Externalities”,<br />
Journal of Political Ec<strong>on</strong>omy, 94: 822-841.<br />
Katz, M L <strong>and</strong> Shapiro, C (1986) “How to License Intangible Property”, Quarterly Journal of<br />
Ec<strong>on</strong>omics, 101:567-590.<br />
Katz, M L <strong>and</strong> Shapiro, C (1986) “<strong>Technology</strong> Adopti<strong>on</strong> in the Presence of Network Externalities”,<br />
Journal of Political Ec<strong>on</strong>omy, 94: 822-841.<br />
Katz, M L <strong>and</strong> Shapiro, C (1987) “R&D Rivalry with Licensing or Imitati<strong>on</strong>”, American Ec<strong>on</strong>omic<br />
Review, 77:402-420.<br />
71<br />
289
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Katz, M L, (1986), “An analysis of cooperative research <strong>and</strong> development”, RAND Journal of<br />
Ec<strong>on</strong>omics, 17 (Winter), pp.527-543.<br />
Katz, M L, Ordover, J A, (1990), “R&D cooperati<strong>on</strong> <strong>and</strong> competiti<strong>on</strong>”, Brookings Papers <strong>on</strong><br />
Ec<strong>on</strong>omic Activity, Special Issue, pp. 137-91.<br />
Kleinknecht, A, Reijnen, J, (1992), “Why do firms cooperate <strong>on</strong> R&D? an empirical study”,<br />
Research Policy, 21, 347-360.<br />
Lee, T <strong>and</strong> Wilde, I (1980) “Market Structure <strong>and</strong> Innovati<strong>on</strong>: A Reformulati<strong>on</strong>”, Quarterly Journal<br />
of Ec<strong>on</strong>omics, 94:429, 436.<br />
Levin R, Klevorick, A K, Nels<strong>on</strong>, RR <strong>and</strong> Winter, S G (1987) “Appropriating the returns from<br />
industrial research <strong>and</strong> development”, Brookings Papers <strong>on</strong> Ec<strong>on</strong>omic Activity, vol. 3:783-820.<br />
Loury, G C (1979) “Market Structure <strong>and</strong> Innovati<strong>on</strong>”. Quarterly Journal of Ec<strong>on</strong>omics. 93(3):<br />
395-410.<br />
Perloff, Jeffrey , Karp, Larry <strong>and</strong> Golan, Amos (2007) Estimating Market Power <strong>and</strong> Strategies,<br />
Cambridge,Ma.: Cambridge Univ. Press<br />
Poyago-Theotoki, J, (1999), “A note <strong>on</strong> endogenous spillovers in a n<strong>on</strong>-tournament R&D<br />
duopoly”, Review of Industrial Organizati<strong>on</strong>, 15(3), November, pp.253-262.<br />
Qiu, L. (1997) “On the dynamic efficiency of Bertr<strong>and</strong> <strong>and</strong> Cournot equilibria”, Journal of<br />
Ec<strong>on</strong>omic Theory, NO 75, p213-229<br />
Reinganum, J F (1981a) “Dynamic Games of Innovati<strong>on</strong>”, Journal of Ec<strong>on</strong>omic Theory, 25:21-41;<br />
1985 “Corrigendum” 35:196-197.<br />
Reinganum, J F (1981b) “On the Diffusi<strong>on</strong> of New <strong>Technology</strong>: A Game Theoretic Approach”,<br />
Review of Ec<strong>on</strong>omic Studies, 48:395-405.<br />
Reinganum, J F (1981c) “Market Structure <strong>and</strong> the Diffusi<strong>on</strong> of New <strong>Technology</strong>”, Bell Journal,<br />
12:618-624.<br />
Reinganum, J F (1982a) “A Dynamic Game of R <strong>and</strong> D: Patent Protecti<strong>on</strong> <strong>and</strong> Competitive<br />
Behaviour”, Ec<strong>on</strong>ometrica, 50:671-688.<br />
Reinganum, J F (1982b) “Strategic Search Theory”, Internati<strong>on</strong>al Ec<strong>on</strong>omic Review, 23:1-15.<br />
Reinganum, Jennifer F. (1983) “Uncertain innovati<strong>on</strong> <strong>and</strong> the persistence of m<strong>on</strong>opoly”, American<br />
Ec<strong>on</strong>omic Review, Vol 73, p741-48<br />
Reinganum, Jennifer F. (1985) “Innovati<strong>on</strong> <strong>and</strong> industry evoluti<strong>on</strong>”, Quarterly Journal of<br />
Ec<strong>on</strong>omics, p81-99<br />
Segerstrom, P., 2007. Intel ec<strong>on</strong>omics. Internati<strong>on</strong>al Ec<strong>on</strong>omic Review 48 (1), 247–280<br />
Simps<strong>on</strong>, R D, V<strong>on</strong>ortas, N S, (1994), “Cournot equilibrium with imperfectly appropriable R&D”,<br />
The Journal of Industrial Ec<strong>on</strong>omics, 1, pp.79-92<br />
72<br />
290
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Spence, A M (1984) “Cost Reducti<strong>on</strong>, Competiti<strong>on</strong>, <strong>and</strong> Industry Performance”, Ec<strong>on</strong>ometrica<br />
52:101-121.<br />
Sutt<strong>on</strong>, John (1998) <strong>Technology</strong> <strong>and</strong> Market Structure, Cambridge, Massachusetts <strong>and</strong> L<strong>on</strong>d<strong>on</strong>,<br />
Engl<strong>and</strong>: The MIT Press.<br />
Suzumura, K, (1992), “Cooperative <strong>and</strong> n<strong>on</strong>-cooperative R&D in an oligopoly with spillovers”,<br />
American Ec<strong>on</strong>omic Review, 82, pp.1307-1320.<br />
Vickers, John (1986) “The evoluti<strong>on</strong> of industry structure when there is a sequence of<br />
innovati<strong>on</strong>s”, Journal of Industrial Ec<strong>on</strong>omics, Vol 35,p1-12<br />
4.2 Endogenous Growth<br />
Aghi<strong>on</strong>,Philippe <strong>and</strong> Giffith,Rachel (2005) Competiti<strong>on</strong> <strong>and</strong> Growth, Cambridge,Ma., The MIT<br />
Press<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1988) “Growth <strong>and</strong> Cycles through Creative Destructi<strong>on</strong>”, MIT<br />
Mimeo.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1992) “A model of growth through creative destructi<strong>on</strong>”,<br />
Ec<strong>on</strong>ometrica, Vol 60,p323-51.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Howitt, Peter (1998) Endogenous Growth Theory, Cambridge,<br />
Massachusetts <strong>and</strong> L<strong>on</strong>d<strong>on</strong>, Engl<strong>and</strong>: The MIT Press.<br />
Aghi<strong>on</strong>, Philippe <strong>and</strong> Shankerman, Mark (2000), “A model of market enhancing infrastructure”<br />
CEPR Discussi<strong>on</strong> Paper, NO.2461.<br />
Aghi<strong>on</strong>, Philippe, Bloom, Nicholas, Blundell, Richard, Griffith, Rachel <strong>and</strong> Howitt, Peter (2002)<br />
“Competiti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>: an inverted U relati<strong>on</strong>ship” NBER Working Paper, no.9269.<br />
Aghi<strong>on</strong>, Philippe, Dewatrip<strong>on</strong>t, M. <strong>and</strong> Rey, P. (1997) “Competiti<strong>on</strong>, financial discipline <strong>and</strong><br />
growth”, Review of Ec<strong>on</strong>omic Studies, Vol 66, No 4, p825-852.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher <strong>and</strong> Vickers, John (1995) “Competiti<strong>on</strong> <strong>and</strong> Growth with<br />
Step-by-Step Technological Progress”, mimeo.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher <strong>and</strong> Vickers, John (1997) “Competiti<strong>on</strong> <strong>and</strong> growth with<br />
step-by-step innovati<strong>on</strong>: an example”, European Ec<strong>on</strong>omic Review, Vol 41, p771-782.<br />
Aghi<strong>on</strong>, Philippe, Harris, Christopher, Howitt, Peter <strong>and</strong> Vickers, John (2001) “Competiti<strong>on</strong>,<br />
imitati<strong>on</strong> <strong>and</strong> growth with step-by-step innovati<strong>on</strong>”, Review of Ec<strong>on</strong>omic Studies, Vol 68, p467-<br />
492.<br />
Aghi<strong>on</strong>,Philippe <strong>and</strong> Griffith,Rachel (2005) Competiti<strong>on</strong> <strong>and</strong> Growth, Cambridge,Ma., The MIT Pr<br />
Ahn, Sangho<strong>on</strong>, “Competiti<strong>on</strong>, innovati<strong>on</strong> <strong>and</strong> productivity growth: a review of theory <strong>and</strong><br />
evidence”, OECD Ec<strong>on</strong>omic Department Working Paper, no 317.<br />
Arrow, K J (1962) “Ec<strong>on</strong>omic Welfare <strong>and</strong> the Allocati<strong>on</strong> of Resources for Inventi<strong>on</strong>”, in<br />
Universities Nati<strong>on</strong>al Bureau Committee for Ec<strong>on</strong>omic Research. The rate <strong>and</strong> directi<strong>on</strong> of<br />
inventive activity Princet<strong>on</strong>, Princet<strong>on</strong> University Press.<br />
73<br />
291
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Blundell, Richard, Griffith, Rachel <strong>and</strong> Van Reenen, John (1999) “Market share, market value <strong>and</strong><br />
innovati<strong>on</strong> in a panel of British manufacturing firms”, Review of Ec<strong>on</strong>omic Studies, Vol 66, No 3, p<br />
529-554.<br />
Chen,Derek H.C. <strong>and</strong> Kee,H.L. (2005),”A Model <strong>on</strong> Knowledge <strong>and</strong> Endogeneous Growth”, World<br />
Bank Policy Research Working Paper 3539, March<br />
D’Asprem<strong>on</strong>t, Claude, Dos Santos Ferreira, Rodolphe <strong>and</strong> Gérard-Varet, Louis-André (2002)<br />
“Strategic R&D investment, competiti<strong>on</strong> toughness <strong>and</strong> growth”, mimeo.<br />
Gilbert, Richard J. <strong>and</strong> Newbery, David M.G. (1982) “Preemptive Patenting <strong>and</strong> the Persistence<br />
of M<strong>on</strong>opoly”, American Ec<strong>on</strong>omic Review, 72:514-526.<br />
Grossman, Gene M. <strong>and</strong> Helpman, Elhanan. (1991) “Quality ladders <strong>and</strong> product cycles”,<br />
Quarterly Journal of Ec<strong>on</strong>omics, Vol 106, p557-586.<br />
Lucas, R E, (1988). “On the Mechanics of Ec<strong>on</strong>omic <strong>Development</strong>”, Journal of M<strong>on</strong>etary<br />
Ec<strong>on</strong>omics 22(1): 3-42.<br />
Nickell, Stephen, J. (1996) “Competiti<strong>on</strong> <strong>and</strong> corporate performance”, The Journal of Political<br />
Ec<strong>on</strong>omy, Vol 104, No 4, p724-746.<br />
Romer, Paul M. (1986) “Increasing Returns <strong>and</strong> L<strong>on</strong>g Run Growth”, The Journal of Political<br />
Ec<strong>on</strong>omy 94(5): 1002-1037.<br />
Romer, Paul M. (1990) “Endogenous Technological Change”, The Journal of Political Ec<strong>on</strong>omy<br />
98(5) part 2: 71-102.<br />
Solow, R M (1956) “A C<strong>on</strong>tributi<strong>on</strong> to the Theory of Ec<strong>on</strong>omic Growth”, Quarterly Journal of<br />
Ec<strong>on</strong>omics 70(1): 65-94.<br />
Solow,Robert (1997) Learning from ‘Learning by Doing’, Stanford: Stanford Univ. Press<br />
Stokey, Nancy L. (1988) “Learning by doing <strong>and</strong> the introducti<strong>on</strong> of new goods”, Journal of<br />
Political Ec<strong>on</strong>omy, Vol 96, No 4 (August), p701-171.<br />
Young, Alwyn (1991) “Learning by Ding <strong>and</strong> the Dynamic Effects of Internati<strong>on</strong>al Trade”, Quarterly<br />
Journal of Ec<strong>on</strong>omics 106(2): 369-406.<br />
Young, Alwyn (1993) “Inventi<strong>on</strong> <strong>and</strong> Bounded Learning by Doing”, Journal of Political Ec<strong>on</strong>omy<br />
101(3): 443-472.<br />
4.3 New Ec<strong>on</strong>omic Geography<br />
Davis, R<strong>on</strong>ald R. <strong>and</strong> David E.Weinstein (2002) B<strong>on</strong>es, Bombs, <strong>and</strong> Break Points: The<br />
Geography of Ec<strong>on</strong>omic Activity, 92(5), pp. 1269-1289.<br />
Fujita, Masahisa, Paul Krugman <strong>and</strong> Anth<strong>on</strong>y J. Venables (2001): The Spatial Ec<strong>on</strong>omy: Cities,<br />
Regi<strong>on</strong>s <strong>and</strong> Internati<strong>on</strong>al Trade, MIT Press.<br />
74<br />
292
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Fujita, Masahisa <strong>and</strong> Jacques-François Thisse, (2002) Agglomerati<strong>on</strong> <strong>and</strong> growth with migrati<strong>on</strong><br />
<strong>and</strong> knowledge externalities, mimeo, April.<br />
Kim, S. (1995) “Expansi<strong>on</strong> of markets <strong>and</strong> the geographic distributi<strong>on</strong> of ec<strong>on</strong>omic activities. The<br />
trends in US regi<strong>on</strong>al manufacturing structure: 1860-1987” Quarterly Journal of Ec<strong>on</strong>omics,<br />
110(4), pp. 881-908.<br />
Krugman, Paul Anth<strong>on</strong>y J. Venables (1995) “Globalizati<strong>on</strong> <strong>and</strong> the Inequality of Nati<strong>on</strong>s”<br />
Quarterly Journal of Ec<strong>on</strong>omics, 110(4), pp. 857-880.<br />
Krugman, Paul (1980) “Scale Ec<strong>on</strong>omies, Product Differentiati<strong>on</strong>, <strong>and</strong> the Pattern of Trade”<br />
American Ec<strong>on</strong>omic Review, 70(5), pp. 950-959.<br />
Krugman, Paul (1981) “Intraindustrial Specialisati<strong>on</strong> <strong>and</strong> the Gains from Trade” Journal of Political<br />
Ec<strong>on</strong>omy, 89(5), pp. 959-973.<br />
Krugman, Paul (1991) “Increasing Returns <strong>and</strong> Ec<strong>on</strong>omic Geography” Journal of Political<br />
Ec<strong>on</strong>omy, 99(3), pp. 483-499.<br />
Martin, R<strong>on</strong> <strong>and</strong> Peter Sunley (1996) “Paul Krugman’s Geographical Ec<strong>on</strong>omics <strong>and</strong> Its<br />
Implicati<strong>on</strong>s for Regi<strong>on</strong>al <strong>Development</strong> Theory. A Critical Assessment” Ec<strong>on</strong>omic Geography,<br />
72(3), pp. 259-292.<br />
Midelfart-Knarvik K.H., H.G. Overman, S.J. Redding <strong>and</strong> A.J. Venables, (2000) The Locati<strong>on</strong> of<br />
European Industry, European Commissi<strong>on</strong>, DG ECFIN, Ec<strong>on</strong>omic Papers No. 142, April.<br />
Neary, Peter (2001) “Of Hypes <strong>and</strong> Hyperbolas: Introducing the New Ec<strong>on</strong>omic Geography”<br />
Journal of Ec<strong>on</strong>omic Literature, 39, June, pp. 536-561.<br />
Overman, Henry G., Stephen Redding <strong>and</strong> Anth<strong>on</strong>y J. Venables (2001) “The Ec<strong>on</strong>omic<br />
Geography of Trade, Producti<strong>on</strong> <strong>and</strong> Income. A Survey of Empirics” L<strong>on</strong>d<strong>on</strong> School of<br />
Ec<strong>on</strong>omics, Centre for Ec<strong>on</strong>omic Performance, September.<br />
Robert-Nicoud, Frédéric (2002) “The structure of the simple ‘New Ec<strong>on</strong>omic Geography’ models”<br />
L<strong>on</strong>d<strong>on</strong> School of Ec<strong>on</strong>omic, Centre for Ec<strong>on</strong>omic Performance, August.<br />
Samuels<strong>on</strong>, Paul A. (1954) “The Transfer Problem <strong>and</strong> Transport Costs II: Analysis of Effects of<br />
Trade Impediments” The Ec<strong>on</strong>omic Journal, 64, pp. 264-289.<br />
4.4 Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics<br />
Baumol,William (1993) Entrepreneurship, Management <strong>and</strong> the Structure of Payoff, Cambridge,Ma.<br />
The MIT Press Chapts. 1-7<br />
Bergek, Anna <strong>and</strong> Staffan Jacobss<strong>on</strong> (2002) The emergence of a growth industry: A comparative<br />
analysis of the German, Dutch <strong>and</strong> Swedish wind turbine industry, in Stan Metcalfe <strong>and</strong> Uwe<br />
Cantner: Change, transformati<strong>on</strong> <strong>and</strong> development, Springer pp. 197-228.<br />
Denicolò, Vincenzo <strong>and</strong> Zanchettin, Piercarlo (2002) “Competiti<strong>on</strong> <strong>and</strong> growth in a neo-<br />
Schumpeterian model”, mimeo Università di Bologna.<br />
75<br />
293
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Dosi, G (1988) “Sources, procedures, <strong>and</strong> microec<strong>on</strong>omic effects of innovati<strong>on</strong>”, Journal of<br />
Ec<strong>on</strong>omic Literature, Vol 26, No 3, p 1120-1171.<br />
Edquist, Charles (2001) The Systems of Innovati<strong>on</strong> Approach <strong>and</strong> Innovati<strong>on</strong> Policy: An account<br />
of the state of the art, lead paper presented at the DRUID C<strong>on</strong>ference, Aalborg, June 12-15,<br />
2001.<br />
Etro, Federico (2008), “Growth Leaders”, Journal of Macroec<strong>on</strong>omics 30, 1148-1172<br />
Fagerberg, J (2002) “A layman’s guide to evoluti<strong>on</strong>ary ec<strong>on</strong>omics”, mimeo, September 2002.<br />
Foster, J (2001) “Competiti<strong>on</strong>, competitive selecti<strong>on</strong> <strong>and</strong> ec<strong>on</strong>omic evoluti<strong>on</strong>”, in Garrouste, P <strong>and</strong><br />
Ioannides, S (eds) Evoluti<strong>on</strong> <strong>and</strong> Path Dependence in Ec<strong>on</strong>omic Ideas, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Freeman (1987) <strong>Technology</strong>, policy <strong>and</strong> ec<strong>on</strong>omic performance: less<strong>on</strong>s from Japan,<br />
L<strong>on</strong>d<strong>on</strong>: Pinter.<br />
Freeman (1994) Critical survey: The ec<strong>on</strong>omics of technical change, Cambridge Journal of<br />
Ec<strong>on</strong>omic, 18, pp. 463-514.<br />
Hayek, F. A. "The Use of Knowledge in Society" American Ec<strong>on</strong>omic Review, 35,<br />
(September 1945), p 519-530.<br />
Hauknes, Johan (1999) “Innovati<strong>on</strong> systems <strong>and</strong> capabilities” STEP Working Paper, Oslo.<br />
Levin, R.C. – Klevorick, A.K. – Nels<strong>on</strong>, R.R. – Winter, S.G. (1987) Appropriating the returns from<br />
industrial research <strong>and</strong> development, Brookings Papers <strong>on</strong> Ec<strong>on</strong>omic Activity, 3, pp. 783-820.<br />
Liu, Xieling – White, Steven (2001) Comparing innovati<strong>on</strong> systems: A framework <strong>and</strong> applicati<strong>on</strong><br />
to China`s transiti<strong>on</strong>al c<strong>on</strong>text, Research Policy, 30, pp. 1091-1114.<br />
Lundvall, Bengt-Åke (1992) Nati<strong>on</strong>al systems of innovati<strong>on</strong>: Towards a theory of innovati<strong>on</strong> <strong>and</strong><br />
interactive learning, L<strong>on</strong>d<strong>on</strong>: Pinter.<br />
Metcalfe, J S (1995), “The Ec<strong>on</strong>omic Foundati<strong>on</strong>s of <strong>Technology</strong> Policy: Equilibrium <strong>and</strong><br />
Evoluti<strong>on</strong>ary Perspectives”, in Paul St<strong>on</strong>eman ed., H<strong>and</strong>book of the Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong><br />
Technological Change, Oxford UK <strong>and</strong> Cambridge USA: Blackwell.<br />
Metcalfe, J S (1998) Evoluti<strong>on</strong>ary Ec<strong>on</strong>omics <strong>and</strong> Creative Destructi<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge.<br />
Nels<strong>on</strong>, Richard R. (1996) The sources of ec<strong>on</strong>omic growth, Harvard University Press, Chapter<br />
10: Nati<strong>on</strong>al innovati<strong>on</strong> systems: A retrospective <strong>on</strong> a study, pp. 274-301.<br />
Nels<strong>on</strong>, R <strong>and</strong> Winter, G (2002) “Evoluti<strong>on</strong>ary theorizing in ec<strong>on</strong>omics”, Journal of Ec<strong>on</strong>omic<br />
Perspectives, Vol 16, No 2, p 23-46.<br />
Nels<strong>on</strong>, R <strong>and</strong> Winter, S (1982) An Evoluti<strong>on</strong>ary Theory of Ec<strong>on</strong>omic Change, L<strong>on</strong>d<strong>on</strong>: Belknap<br />
Press.<br />
76<br />
294
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Pavitt, K (2002) “Knowledge about knowledge since Nels<strong>on</strong> & Winter: a mixed record”, SPRU<br />
Electr<strong>on</strong>ic Working Paper Series, No. 83.<br />
Saviotti, Pier Paolo (2000) Networks, nati<strong>on</strong>al innovati<strong>on</strong> systems <strong>and</strong> self-organisati<strong>on</strong>,<br />
University of Pierre Mendes-France, mimeo.<br />
Segerstrom, P., 2007. Intel ec<strong>on</strong>omics. Internati<strong>on</strong>al Ec<strong>on</strong>omic Review 48 (1), 247–280<br />
St<strong>on</strong>eman, Paul (ed) (1998) H<strong>and</strong>book of the Ec<strong>on</strong>omics of Innovati<strong>on</strong> <strong>and</strong> Technological<br />
Change, Oxford: Blackwell Publisher.<br />
Winter, S (1987) “Knowledge <strong>and</strong> competence as strategic assets” in Teece, D (ed) The<br />
Competitive Challenge, New York: Harper <strong>and</strong> Row.<br />
Zeira, J.(2003) Innovati<strong>on</strong>s, Patent Races <strong>and</strong> Endogenous Growth, mimeo. Harvard University.<br />
4.5 Cliometrics<br />
Atkens<strong>on</strong>, Andrew <strong>and</strong> Patrick Kehoe (2001) “The Transiti<strong>on</strong> to a New Ec<strong>on</strong>omy After the<br />
Sec<strong>on</strong>d Industrial Revoluti<strong>on</strong>”, NBER Working Paper, No. 8676, December<br />
Coase, R<strong>on</strong>ald (1960): The Problem of Social Cost, Journal of Law <strong>and</strong> Ec<strong>on</strong>omics, 3(1), pp. 1-<br />
44<br />
Coleman, D.C. <strong>and</strong> Christine MacLeod (1986) “Attitude to New Techniques: British Businessmen”<br />
1800-1950, Ec<strong>on</strong>omic History Review, 39(4), pp. 588-611<br />
C<strong>on</strong>rad, Alfred H. <strong>and</strong> John R. Meyer (1957) “Ec<strong>on</strong>omic Theory, Statistical Inference, <strong>and</strong><br />
Ec<strong>on</strong>omic History” Journal of Ec<strong>on</strong>omic History, 17(4), pp. 524-44<br />
C<strong>on</strong>rad, Alfred H. <strong>and</strong> John R. Meyer (1958) “The Ec<strong>on</strong>omics of Slavery in the Ante Bellum<br />
South” Journal of Political Ec<strong>on</strong>omy, 66(2), pp. 95-130.<br />
Crafts, N.F.R (1987) “Cliometrics, 1971-1986: a survey” Journal of Applied Ec<strong>on</strong>ometrics, 2(3),<br />
pp. 171-192<br />
Crafts, Nicholas (2002) “Productivity growth in the industrial revoluti<strong>on</strong>: A new growth accounting<br />
perspective”, mimeo, L<strong>on</strong>d<strong>on</strong> School of Ec<strong>on</strong>omics, January.<br />
Crafts, Nicholas F.R <strong>and</strong> C.K. Harley (1992) “Output growth <strong>and</strong> British industrial revoluti<strong>on</strong>: a<br />
restatement of the Crafts-Harley view” Ec<strong>on</strong>omic History Review, 45(4), pp. 703-730<br />
Crafts, Nicholas F.R <strong>and</strong> Terence C. Mills (1997) “Endogenous innovati<strong>on</strong>, trend growth, <strong>and</strong> the<br />
British industrial revoluti<strong>on</strong>: reply to Greasley <strong>and</strong> Oxley” Journal of Ec<strong>on</strong>omic History, 57(4), pp.<br />
950-956.<br />
Crafts, Nicholas F.R. (1995a) “Exogenous or Endogenous Growth? The Industrial Revoluti<strong>on</strong><br />
Rec<strong>on</strong>sidered” Journal of Ec<strong>on</strong>omic History, 55(4), pp. 745-772<br />
Crafts, Nicholas F.R. (1995b). “Macroinventi<strong>on</strong>s, Ec<strong>on</strong>omic Growth, <strong>and</strong> `Industrial Revoluti<strong>on</strong>` in<br />
Britain <strong>and</strong> France” Ec<strong>on</strong>omic History Review, 48(3), pp. 591-598.<br />
77<br />
295
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Crafts, Nicholas F.R. (1996) “The First Industrial Revoluti<strong>on</strong>: A Guided Tour for Growth<br />
Ec<strong>on</strong>omists” American Ec<strong>on</strong>omic Review, 86(2), pp. 197-201.<br />
Fischlow, Albert (1965) American Railroads <strong>and</strong> the Transformati<strong>on</strong> of the Ante-Bellum Ec<strong>on</strong>omy,<br />
Harvard University Press, Chapter 2.<br />
Fogel, Robert W. (1962): “A quantitative approach to the study of railroads in American ec<strong>on</strong>omic<br />
growth: a report of some preliminary findings” Journal of Ec<strong>on</strong>omic History, 22(2), pp. 163-197.<br />
Fogel, Robert W. (1964): Railroads <strong>and</strong> the American Ec<strong>on</strong>omic Growth: <str<strong>on</strong>g>Essays</str<strong>on</strong>g> in Ec<strong>on</strong>ometric<br />
History, Baltimore: Johns Hopkins University Press.<br />
Fogel, Robert W. (1965) “The reunificati<strong>on</strong> of ec<strong>on</strong>omic history with ec<strong>on</strong>omic theory” American<br />
Ec<strong>on</strong>omic Review, 55 (1/2), pp. 92-98.<br />
Fogel, Robert W. (1966a) “The new ec<strong>on</strong>omic history: its findings <strong>and</strong> methods” The Ec<strong>on</strong>omic<br />
History Review, 19 (3), pp. 642-656.<br />
Fogel, Robert W. (1966b) “Railroads as an analogy to the space effort: some ec<strong>on</strong>omic aspects”<br />
Ec<strong>on</strong>omic Journal, 76(301), pp. 16-43.<br />
Goldin, Claudia (1995) “Cliometrics <strong>and</strong> the Nobel” Journal of Ec<strong>on</strong>omic Perspectives, 9 (2), pp.<br />
191-208.<br />
Greasley, David <strong>and</strong> Les Oxley (1994) “Rehabilitati<strong>on</strong> sustained: the industrial revoluti<strong>on</strong> as a<br />
macroec<strong>on</strong>omic epoch” Ec<strong>on</strong>omic History Review, 47(4), pp. 760-768.<br />
Greasley, David <strong>and</strong> Les Oxley (1997a) “Endogenous growth or “Big Bang”: two views of the first<br />
industrial revoluti<strong>on</strong>” Journal of Ec<strong>on</strong>omic History, 57(4), pp. 935-949.<br />
Greasley, David <strong>and</strong> Les Oxley (1997b) “Endogenous growth, trend output, <strong>and</strong> the industrial<br />
revoluti<strong>on</strong>: reply to Crafts <strong>and</strong> Mills” Journal of Ec<strong>on</strong>omic History, 57(4), pp. 957-960.<br />
Gregers<strong>on</strong>, Brigitte <strong>and</strong> Björn Johns<strong>on</strong>, (1998) Nati<strong>on</strong>al Systems of Innovati<strong>on</strong> as a Framework<br />
for Innovati<strong>on</strong> Policy, The United Nati<strong>on</strong>s University, April.<br />
Heckman, James J. (1997) “The value of quantitative evidence <strong>on</strong> the effect of the past <strong>and</strong> the<br />
present” American Ec<strong>on</strong>omic Review, 87(2), pp. 404-408.<br />
J<strong>on</strong>es, Charles I. (1999) “Was an industrial revoluti<strong>on</strong> inevitable? Ec<strong>on</strong>omic growth over the very<br />
l<strong>on</strong>g run” NBER Working Paper, No. 7375, October.<br />
Khan, Zorina B. (1995) “Property rights <strong>and</strong> patent litigati<strong>on</strong> in early nineteenth-century America”<br />
Journal of Ec<strong>on</strong>omic History, 55(1), pp. 58-97.<br />
L<strong>and</strong>es, David (1994) “What room for accident in history? Explaining big changes by small<br />
events” Ec<strong>on</strong>omic History Review, 47(4), pp. 637-656.<br />
L<strong>and</strong>es,David (1998), The Wealth <strong>and</strong> Poverty of Nati<strong>on</strong>s, W.W.Nort<strong>on</strong>, New York<br />
78<br />
296
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Lorenz, Edward H. (1991) “An evoluti<strong>on</strong>ary explanati<strong>on</strong> for competitive decline: the British<br />
shipbuilding industry, 1890-1970” Journal of Ec<strong>on</strong>omic History, 51(4), pp. 911-935.<br />
MacLeod, Christine (1992) “Strategies for innovati<strong>on</strong>: the diffusi<strong>on</strong> of new technology in<br />
nineteenth-century British industry” Ec<strong>on</strong>omic History Review, 45(2), pp. 285-307.<br />
Maddis<strong>on</strong>, Angus (2001) The World Ec<strong>on</strong>omy: A Millennial Perspective, OECD, <strong>Development</strong><br />
Centre Studies.<br />
McCloskey, D<strong>on</strong>ald N. (1978) “The achievements of the cliometric school” Journal of Ec<strong>on</strong>omic<br />
History, 38(1), pp. 13-28<br />
Mokyr (1992) “Technological inertia in ec<strong>on</strong>omic history” Journal of Ec<strong>on</strong>omic History, 52(2), pp.<br />
325-338.<br />
Mokyr, Joel (1993) Introducti<strong>on</strong>: The New Ec<strong>on</strong>omic History <strong>and</strong> the Industrial Revoluti<strong>on</strong>, in Joel<br />
Mokyr (ed.): The British Industrial Revoluti<strong>on</strong>: An Ec<strong>on</strong>omic perspective, Oxford.<br />
Mokyr, Joel (1997) Inventi<strong>on</strong> <strong>and</strong> Rebelli<strong>on</strong>: Why Do innovati<strong>on</strong> Occur at All? An Evoluti<strong>on</strong>ary<br />
Approach, paper presented at the C<strong>on</strong>ference <strong>on</strong> Minorities <strong>and</strong> Ec<strong>on</strong>omic growth, Bar Ilan<br />
University, June 2-5 1997.<br />
Mokyr, Joel (1999a) Innovati<strong>on</strong> <strong>and</strong> Rebelli<strong>on</strong>: Why do Innovati<strong>on</strong>s Occur at all? An Evoluti<strong>on</strong>ary<br />
Approach, in Elise Brezis <strong>and</strong> Peter Temin, eds., Minorities <strong>and</strong> Ec<strong>on</strong>omic Growth. Amsterdam:<br />
Elsevier Publishers, pp. 179-203.<br />
Mokyr, Joel (1999b) Knowledge, <strong>Technology</strong>, <strong>and</strong> Industrial growth During the Industrial<br />
Revoluti<strong>on</strong>, Northwestern University, mimeo.<br />
North, Douglass C. (1968) “Sources of productivity change <strong>on</strong> ocean shipping, 1600-1850”<br />
Journal of Political Ec<strong>on</strong>omy, 76 (5), pp. 953-970.<br />
North, Douglass C. (1990) Instituti<strong>on</strong>s, instituti<strong>on</strong>al change, <strong>and</strong> ec<strong>on</strong>omic performance,<br />
Cambridge University Press.<br />
North, Douglass C. (1991) “Instituti<strong>on</strong>s” Journal of Ec<strong>on</strong>omic Perspectives, 5(1), pp. 97-112.<br />
North, Douglass C. (1994) “Ec<strong>on</strong>omic performance through time” American Ec<strong>on</strong>omic Review,<br />
84(3), pp. 359-368.<br />
North, Douglass C. (1997) “Cliometrics – 40 years later” American Ec<strong>on</strong>omic Review, 87(2), pp.<br />
412-414.<br />
North, Douglass (2005), Underst<strong>and</strong>ing the Process of Ec<strong>on</strong>omic Change, Princet<strong>on</strong> Univ. Press<br />
O’Rourke, Kevin H. <strong>and</strong> Jeffrey G. Williams<strong>on</strong>, (2002) “From Malthus to Ohlin, trade, growth <strong>and</strong><br />
distributi<strong>on</strong> since 1500” NBER Working Paper, No. 8955, May<br />
Ols<strong>on</strong>, Mancur (1996) “Big bills left <strong>on</strong> the sidewalk: why some nati<strong>on</strong>s are rich, <strong>and</strong> other poor”<br />
Journal of Ec<strong>on</strong>omic Perspectives, 10(2), pp. 3-24.<br />
Ols<strong>on</strong>,Mancur (1982),The Rise <strong>and</strong> Decline of Nati<strong>on</strong>s, Yale Univ. Press<br />
79<br />
297
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Pearce, Robert (2002): Nati<strong>on</strong>al systems of innovati<strong>on</strong> <strong>and</strong> the internati<strong>on</strong>al technology strategies<br />
of multinati<strong>on</strong>als, University of Reading.<br />
Phillips, William H. (1992) “Patent growth in the old domini<strong>on</strong>: the impact of railroad integrati<strong>on</strong><br />
before 1880” Journal of Ec<strong>on</strong>omic History, 52(2), pp. 389-400.<br />
Posner, Richard (1980) “A theory of primitive society, with special reference to the law” Journal of<br />
Law <strong>and</strong> Ec<strong>on</strong>omics, 23, pp. 1-54.<br />
Schoser, Christof (1999) “The instituti<strong>on</strong>s defining nati<strong>on</strong>al systems of innovati<strong>on</strong>: a new<br />
tax<strong>on</strong>omy to analyse the Impact of globalisati<strong>on</strong>”, University of Freiburg, Discussi<strong>on</strong> Paper.<br />
Shiue, Carol H (2002) “Transport costs <strong>and</strong> the geography of arbitrage in eighteenth-century<br />
China” American Ec<strong>on</strong>omic Review, 92 (5), pp. 1406-1419.<br />
Smith, Chauncey (1890) “A century of patent law” Quarterly Journal of Ec<strong>on</strong>omics, 5(1), pp. 44-<br />
69.<br />
Sokoloff, Kenneth L. (1988) “Incentive activity in early industrial America: evidence from patent<br />
records, 1790-1846” Journal of Ec<strong>on</strong>omic History, 48(4), pp. 813-850.<br />
Sullivan, Richard J. (1994) “Estimates of the value of patent rights in Great Britain <strong>and</strong> Irel<strong>and</strong>,<br />
1852-1876” Ec<strong>on</strong>omica, 61(241), pp. 37-58.<br />
Triplett, Jack E. (1999) “The Solow productivity paradox: what do computers do to productivity?”<br />
Canadian Journal of Ec<strong>on</strong>omics, 32(2), pp. 309-334.<br />
Wallis, John J. – North, Douglass C. (1986): Measuring the Transacti<strong>on</strong> Sector in the American<br />
Ec<strong>on</strong>omy, in Engerman, S.L. – Gallman, R.E. (eds), L<strong>on</strong>g-term factors in American ec<strong>on</strong>omic<br />
growth, University of Chicago Press, pp. 95-148.<br />
4.6 Strategic Management <strong>and</strong> Networks<br />
Audretsch, D, Baumol, W <strong>and</strong> Burke, A (2001) “Competiti<strong>on</strong> policy in dynamic markets”,<br />
Internati<strong>on</strong>al Journal of Industrial Organizati<strong>on</strong>, Vol 19, p 665-693.<br />
B<strong>on</strong>accorsi, Andrea <strong>and</strong> Paola Giuri (2000) “Industry life cycle <strong>and</strong> the evoluti<strong>on</strong> of an industry<br />
network” LEM Working Paper Series, No. 4, December<br />
Ellig, Jerry <strong>and</strong> Daniel Lin (2001): A tax<strong>on</strong>omy of dynamic competiti<strong>on</strong> theories, in Jerry Ellig (ed):<br />
Dynamic Competiti<strong>on</strong> <strong>and</strong> Public Policy: <strong>Technology</strong>, Innovati<strong>on</strong> <strong>and</strong> Antitrust Issues, Cambridge<br />
University Press.<br />
Evans, D <strong>and</strong> R. Schmalensee (2001) “Some ec<strong>on</strong>omic aspects of antitrust analysis in<br />
dynamically competitive industries“ NBER working paper, No.8268.<br />
Farrel, Joseph <strong>and</strong> Garth Sal<strong>on</strong>er, “Coordinati<strong>on</strong> through committees <strong>and</strong> markets”, RAND<br />
Journal of Ec<strong>on</strong>omics, Vol 16 (Spring), p70-83<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W. (2003), Ec<strong>on</strong>omies of Network Industries, L<strong>on</strong>d<strong>on</strong>: Routledge, 2003<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>,H.W. (2006), Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>: Routledge,<br />
2006<br />
80<br />
298
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)<br />
HANS W. GOTTINGER<br />
Grant, Robert M. (2002): C<strong>on</strong>temporary Strategy Analysis, 4th editi<strong>on</strong>, Blackwell Publishing<br />
Katz, Michael L. <strong>and</strong> Shapiro, Carl (1986) “Product compatibility choice in a market with<br />
technological progress”, Oxford Ec<strong>on</strong>omic Papers, Special Issue <strong>on</strong> the New Industrial<br />
Ec<strong>on</strong>omics, November.<br />
Katz, Michael L. <strong>and</strong> Shapiro, Carl (1994) “Systems competiti<strong>on</strong> <strong>and</strong> network effects”, Journal of<br />
Ec<strong>on</strong>omic Perspectives, Vol 8, No 2 (Spring), p93-115<br />
Klein, Jeremy (2002): Bey<strong>on</strong>d Competitive Advantage, Strategic Change, 11(6), pp. 317-326.<br />
Klepper, Steven (1996) “Entry, exit, growth <strong>and</strong> innovati<strong>on</strong> over the product life cycle” American<br />
Ec<strong>on</strong>omic Review, Vol. 86, No 3, pp. 562-583.<br />
Klepper, Steven <strong>and</strong> Kenneth Sim<strong>on</strong>s (2001), “Industry shakeouts <strong>and</strong> technological change”<br />
mimeo.<br />
Jovanovic, Boyan (1982) “Selecti<strong>on</strong> <strong>and</strong> the evoluti<strong>on</strong> of industry” Ec<strong>on</strong>ometrica, Vol. 50 No 3, pp.<br />
649-670.<br />
Le<strong>on</strong>ard-Bart<strong>on</strong>, D (1992) “Core capabilities <strong>and</strong> core rigidities: a paradox in managing new<br />
product development” Strategic Management Journal, Summer, pp. 111-125.<br />
Liebowitz, S.J. <strong>and</strong> Stephen E. Margolis (1994) “Network externality: an uncomm<strong>on</strong> tragedy”<br />
Journal of Ec<strong>on</strong>omic Perspectives, p133-150.<br />
Liebowitz, S.J. <strong>and</strong> Stephen E. Margolis, (1995) “Are network externalities a new source of<br />
market failure?” Research in Law <strong>and</strong> Ec<strong>on</strong>omics¸ Vol 17, p.41-46<br />
March, J.G. (1991) “Explorati<strong>on</strong> <strong>and</strong> exploitati<strong>on</strong> in organizati<strong>on</strong>al learning” Organizati<strong>on</strong>al<br />
Science, 2, pp. 71-87.<br />
Porter, Michael (1990): The Competitive Advantage of Nati<strong>on</strong>s, The Free Press, New York<br />
Porter, Michael E (1985): Competitive Advantage, The Free Press, New York<br />
Posner, R (2000) “Antitrust in the new ec<strong>on</strong>omy“ John M. Olin Law <strong>and</strong> Ec<strong>on</strong>omics Working<br />
Paper, No 106, University of Chicago.<br />
Shapiro, C <strong>and</strong> H.R. Varian (1999) Informati<strong>on</strong> Rules: A Strategic Guide to the Network<br />
Ec<strong>on</strong>omy, Harvard Business School Press, Bost<strong>on</strong>.<br />
Sim<strong>on</strong>s, Kenneth (2001) “Product market characteristics <strong>and</strong> the industry life cycle”, mimeo.<br />
Tushman, M L <strong>and</strong> E. Romanelli (1994) “Organizati<strong>on</strong>al transformati<strong>on</strong> as punctuated equilibrium:<br />
an empirical test” Academy of Management Journal, 37, pp. 1141-1166.<br />
81<br />
299
HANS W. GOTTINGER<br />
3.1 MICROECONOMIC FOUNDATIONS OF ECONOMIC GROWTH (REVIEW)S<br />
Wecker, Claudia (2003) “Market performance <strong>and</strong> competiti<strong>on</strong>: a product life cycle model”,<br />
Technovati<strong>on</strong>, Vol 23, No 4, pp. 281-290.<br />
Comprehensive treatments of Reviewer(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>)<br />
Ec<strong>on</strong>omies of Network Industries, Routledge : L<strong>on</strong>d<strong>on</strong>, New York 2003<br />
Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hyper-Competiti<strong>on</strong> Routledge, L<strong>on</strong>d<strong>on</strong>, New York 2006<br />
Strategic Ec<strong>on</strong>omics of Network Industries, NovaScience, New York 2009<br />
Strategic Alliances in Biotech <strong>and</strong> Pharmaceuticals, NovaScience, New York 2010 (with F.Floether<br />
<strong>and</strong> C.Umali)<br />
Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch-Up, NovaScience, New York 2012 (with M. Goosen, <strong>and</strong><br />
M.Takashima)<br />
82<br />
300
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
301
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
302
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
303
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
304
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
305
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
306
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
307
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
308
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
309
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
310
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
311
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
312
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
HANS W. GOTTINGER<br />
313
HANS W. GOTTINGER<br />
3.2 ECONOMIC GROWTH, CATCHING-UP, FALLING BEHIND AND GETTING AHEAD<br />
314
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
CATCH‐UP AND CONVERGENCE:<br />
MECHANISM DESIGN FOR<br />
ECONOMIC DEVELOPMENT*<br />
*Paper presented at the 1st Internati<strong>on</strong>al C<strong>on</strong>gress Jean Baptiste Say, Research Network <strong>on</strong> Innovati<strong>on</strong>,<br />
Boulogne‐sur‐Mer, France, August 27‐30, 2014<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
STRATEC Munich, Germany<br />
gottingerhans@gmail.com<br />
Abstract<br />
This paper identifies factors that promote l<strong>on</strong>g‐run sustainable ec<strong>on</strong>omic growth <strong>and</strong> industrial catchup<br />
in achieving superior performance. An informati<strong>on</strong>ally efficient mechanism design of the Hayek‐<br />
Hurwicz type would c<strong>on</strong>tain as essential elements allocative efficiency with a limited state role,<br />
fostering of private entrepreneurship for high intensity dynamic competiti<strong>on</strong> through industry<br />
specific ‘technological racing’ within a supporting market structure, free trade <strong>and</strong> intellectual<br />
property rights (IPRs), flexible labour markets, low taxes <strong>on</strong> labour, capital income <strong>and</strong> profits.<br />
Emphasis <strong>on</strong> high value‐added network industries directed toward increasing returns. If resource<br />
endowment am<strong>on</strong>g countries are similar these factors would eventually lead to c<strong>on</strong>vergence through<br />
catch‐up performance, otherwise they would lead to divergence.<br />
Key words<br />
Catch‐Up; C<strong>on</strong>vergence; Mechanism Design; Ec<strong>on</strong>omic <strong>Development</strong><br />
Taking a philosophical point of view, this may be seen as the mechanics for the<br />
implementati<strong>on</strong> of Adam Smith’s invisible h<strong>and</strong>: despite private informati<strong>on</strong> <strong>and</strong> pure selfish<br />
behavior, social welfare is achieved. All the field of Mechanism Design is just a generalizati<strong>on</strong><br />
of this possibility. – Noam Nisan (2007)<br />
INTRODUCTION<br />
What are the major factors for ec<strong>on</strong>omies to succeed <strong>on</strong> catching‐up in terms of GDP<br />
or GDP per capita (of purchasing power parity PPP) as an indicator of prosperity?<br />
As has been broadly covered in the ec<strong>on</strong>omic growth <strong>and</strong> development literature<br />
(<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> & Goosen, 2012), there have been c<strong>on</strong>vergence theories <strong>on</strong> advanced<br />
(developed) ec<strong>on</strong>omies that have been partially verified for some OECD ec<strong>on</strong>omies,<br />
67<br />
315
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
but equally there have been observati<strong>on</strong>s of divergence am<strong>on</strong>g some developed <strong>and</strong><br />
developing (transiti<strong>on</strong>al) ec<strong>on</strong>omies that appear to show a growing gap. <strong>Technology</strong><br />
adopti<strong>on</strong> within regi<strong>on</strong>s of a nati<strong>on</strong>al ec<strong>on</strong>omy (say , China) were a major factor of<br />
less developed regi<strong>on</strong>s to catch up. Catch‐up metrics could also be identified in<br />
disaggregated form next to aggregates such as GDP (total), GDP per head, Human<br />
<strong>Development</strong> Index (HDI), for example, industry/technology index, infrastructure<br />
indicator (transportati<strong>on</strong>, communicati<strong>on</strong> networks), R&D expenditure<br />
(government, private industry) <strong>and</strong> competitiveness (global market share of key<br />
industries). A process of catching up induced by industrial races may tend to<br />
c<strong>on</strong>verge over time within a bloc of similar countries if technological <strong>and</strong><br />
educati<strong>on</strong>al endowment is similar as covered by Abramovitz’ (1986) famous ‘social<br />
capabilities’, as key industries engage in more incremental <strong>and</strong> complementary<br />
innovati<strong>on</strong> that through internati<strong>on</strong>al trade <strong>and</strong> foreign direct investment (FDI)<br />
spread to emerging industries in likewise developing ec<strong>on</strong>omies. In the post World<br />
War II history what was the mechanism that induced Japan to be <strong>on</strong> a path of catchup<br />
growth in the 1960s <strong>and</strong> 1970s in terms of per capita income growth? As<br />
suggested in the Asian miracle (Krugman, 1994), <strong>on</strong>e factor was input growth in key<br />
value‐added industries through capital expansi<strong>on</strong>, the other cumulative<br />
technological advancement through largely incremental innovati<strong>on</strong> leading to<br />
superiority in some industries. But as so<strong>on</strong> as the technological fr<strong>on</strong>tier was close to<br />
be reached it became increasingly difficult to dominate the market. Also in a wide<br />
array of high‐tech markets it became increasingly decisive to have integrative<br />
technologies to catch complementary <strong>and</strong> increasing returns markets.<br />
The paper shows that catch‐up processes should be primarily understood as<br />
technological races, establishing new industries that allow free flow of informati<strong>on</strong><br />
through entrepreneurial activity <strong>and</strong> innovati<strong>on</strong>. Under these circumstances, it is<br />
more likely that an ec<strong>on</strong>omic mechanism design generating more informati<strong>on</strong>,<br />
choices, ec<strong>on</strong>omic freedom <strong>and</strong> market transparency supported by democratic<br />
instituti<strong>on</strong>s, will have a better ec<strong>on</strong>omic performance record, in a sustainably l<strong>on</strong>g<br />
run, than any other that fails <strong>on</strong> the informati<strong>on</strong> distributi<strong>on</strong> <strong>and</strong> strategic incentive<br />
side, which would coincide with various types of socialist systems.<br />
The c<strong>on</strong>tent of the paper is as follows. Secti<strong>on</strong> 2 gives a brief historical outline <strong>on</strong> the<br />
interrelati<strong>on</strong> between ec<strong>on</strong>omic development <strong>and</strong> catching‐up paths. Secti<strong>on</strong> 3<br />
sketches the basic structure of algorithmic mechanism design (AMD) as it pertains to<br />
computati<strong>on</strong>al ec<strong>on</strong>omic systems, in particular to superior performance of<br />
decentralized (distributed) dynamic systems. Secti<strong>on</strong> 4 shows the c<strong>on</strong>necti<strong>on</strong><br />
between industrial competiti<strong>on</strong> <strong>and</strong> ist aggregati<strong>on</strong> to macro competiti<strong>on</strong> <strong>on</strong> a<br />
nati<strong>on</strong>al or regi<strong>on</strong>al level – in view of technological racing. Secti<strong>on</strong> 5 identifies<br />
technological fr<strong>on</strong>tiers <strong>on</strong> the firm level (FTF) as embedded <strong>on</strong> the industry level<br />
(ITF), amenable to a statistical profiling of technological evoluti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong>. It<br />
shows some rules of technological race behaviour resulting in statistical indicators<br />
68 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
316
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
<strong>on</strong> an industry level. Sec. 6 looks at industrial (in)efficiencies in catch‐up countries as<br />
benchmarked against the industrial leader. C<strong>on</strong>clusi<strong>on</strong>s follow <strong>and</strong> open problems<br />
are discussed in the last secti<strong>on</strong>.<br />
A BRIEF HISTORICAL REVIEW<br />
In a historical c<strong>on</strong>text, a catch‐up process becomes multidimensi<strong>on</strong>al, emerging with<br />
Engl<strong>and</strong>, moving to the US <strong>and</strong> a few European ec<strong>on</strong>omies, Germany, France, later<br />
to Japan <strong>and</strong> the newly industrialized ec<strong>on</strong>omies (NIEs) in Asia overlapping the<br />
BRICs. The argument is that due to expansi<strong>on</strong> of trade through globalizati<strong>on</strong> the<br />
catch‐up process is multidimensi<strong>on</strong>al <strong>and</strong> multispeed (Wan, 2004). Even for a few<br />
increasing in overall speed the benchmark of catch‐up keeps shifting from <strong>on</strong>e or a<br />
few to many more. For example, in the past WW II period, industrial <strong>and</strong> ec<strong>on</strong>omic<br />
growth of Japan was driven toward the US ec<strong>on</strong>omy less so toward Europe; the<br />
pattern of growth <strong>and</strong> catch‐up of China hits many more emerging ec<strong>on</strong>omies with<br />
significant potentials. We have modelled this by a more complex differential game as<br />
in <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Goosen (2012).<br />
As a precursor of ec<strong>on</strong>omic mechanism design, from an ec<strong>on</strong>omic history<br />
perspective, we first identify Gershenkr<strong>on</strong> (1962) who emphasized state acti<strong>on</strong> <strong>on</strong><br />
industrializati<strong>on</strong> setting free guided entrepreneurial activities through targeted<br />
industrial policy (al<strong>on</strong>g the line of Japan <strong>and</strong> later South Korea, Taiwan) (Lee, 2013).<br />
This compares to Abramovitz’ (1986) <strong>and</strong> Abramovitz <strong>and</strong> David’s (1992) catching<br />
up process where self‐reinforcing, industry specific, competitive market forces,<br />
rather than state acti<strong>on</strong>, would initiate <strong>and</strong> sustain a technological race resulting in<br />
leadership positi<strong>on</strong>s across a range of industries. Here the roots of growth rest in the<br />
microec<strong>on</strong>omic industry structure <strong>and</strong> new industry creating high technology<br />
entrepreneurship as c<strong>on</strong>cepti<strong>on</strong>ally <strong>and</strong> empirically explored by Scherer (1992, 1999).<br />
The World Bank in the past takes a middle ground, propagating the state’s role to<br />
develop social capital which then by itself creates social capability to induce an<br />
upward potential through market forces. A more activistic role has recently been<br />
favoured by Justus Lin (2013), the former Chinese Chief Ec<strong>on</strong>omist of the World<br />
Bank. Amsden’s development statism is a further extensi<strong>on</strong> of Gershenkr<strong>on</strong>,<br />
catching‐up as a process of learning how to compete through ’Late Industrializati<strong>on</strong>’<br />
(Amsden, 1989). Other than the main catchup processes listed here have been<br />
reviewed by Burkett <strong>and</strong> Hart‐L<strong>and</strong>sberg (2003), <strong>and</strong> Fagerberg <strong>and</strong> Godinho (2004).<br />
In ’Achieving Rapid Growth’ Sachs <strong>and</strong> Warner (1996), in comparing growing<br />
middle income countries identify major factors such as allocative efficiency (with<br />
government interacti<strong>on</strong> low, markedly less regulati<strong>on</strong>), high degree of competiti<strong>on</strong>,<br />
free trade. flexible labour markets, low taxes <strong>on</strong> labour, capital income <strong>and</strong> profits.<br />
While c<strong>on</strong>vergence am<strong>on</strong>g developing ec<strong>on</strong>omies may be facilitated by becoming<br />
69<br />
317
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
more alike in terms of sources of growth such as technological progress, innovati<strong>on</strong>,<br />
levels of physical capital, labour productivity, quality of human capital, extent of<br />
trade openess, industrial structure <strong>and</strong> instituti<strong>on</strong>al framework, <strong>on</strong>e clear<br />
distinguishing factor may involve increasing returns mechanisms (IRMs). IRMs<br />
eliminate any kind of c<strong>on</strong>vergence <strong>and</strong> allow for <strong>on</strong>going quasi‐linear growth.<br />
Anecdotal observati<strong>on</strong>s as put forward by Easterly <strong>and</strong> Levine (2002) can be<br />
summarized as stylized facts <strong>on</strong> growth mechanisms.<br />
(1) Factor accumulati<strong>on</strong> does not account for growth differences but total factor<br />
productivity does for a substantial amount;<br />
(2) There are hugely growing differences in GDP per capita. On a global scale<br />
divergence <strong>and</strong> not c<strong>on</strong>diti<strong>on</strong>al c<strong>on</strong>vergence is the major c<strong>on</strong>cern in<br />
development policy; <strong>and</strong><br />
(3) Am<strong>on</strong>g developing ec<strong>on</strong>omies growth is not persistent over time.<br />
If movable through market forces all factors of producti<strong>on</strong> flow to the same places<br />
suggesting important externalities.<br />
A SIMPLE MECHANISM DESIGN FOR CATCHUP AND<br />
DEVELOPMENT<br />
A simple mechanism design emanates from the following paradigmatic situati<strong>on</strong>:<br />
Let there be k agents in an (Internet) ec<strong>on</strong>omy that collectively generate dem<strong>and</strong><br />
competing for resources from a supplier. The supplier herself announces prices<br />
entering a bulletin board accessible to all agents (as a sort of transparent market<br />
instituti<strong>on</strong>). In a simple form of a trading process we could exhibit a “tat<strong>on</strong>nement<br />
process” <strong>on</strong> a graph where the agents set up a dem<strong>and</strong> to the suppliers who<br />
advertise prices <strong>on</strong> a bulletin board which are c<strong>on</strong>verted to new prices in interacti<strong>on</strong><br />
with the agents.<br />
The tat<strong>on</strong>nement process in ec<strong>on</strong>omics is a simple form of an algorithmic<br />
mechanism design (AMD) (Nisan & R<strong>on</strong>en, 2001), which in modern computer<br />
science (CS) emerges as an offspring to algorithmic game theory (Nisan et al, 2007).<br />
Algorithmic ingredients apply to rati<strong>on</strong>al <strong>and</strong> selfish agents having well defined<br />
utility functi<strong>on</strong>s repesenting preferences over possible outputs of the algorithm. A<br />
payment ingredient motivates the agents. Mechanism Design Theory (MDT) aims to<br />
show how privately known preferences of the entire populati<strong>on</strong> can be aggregated<br />
towards a ’social choice’ that drives the mechanism of the whole ec<strong>on</strong>omy.<br />
(i) Each agent or group of agents have some some private input represented<br />
by its type. Its type is embedded in public knowledge or resource<br />
endowment as the ’social envir<strong>on</strong>ment’;<br />
(ii) There is a producti<strong>on</strong> functi<strong>on</strong> (or output specificati<strong>on</strong>) that associates<br />
each type t = t 1 , ..., t n with a set of socially allowable outputs (producti<strong>on</strong>s)<br />
oO; <strong>and</strong><br />
70 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
318
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
(iii)<br />
Each agent’s preferences are codified as real‐valued utility functi<strong>on</strong>s noted<br />
by u i (t i ,o).<br />
The universal functi<strong>on</strong> is specified in linear terms as u i = r i + v i (t i ,o) where r i 0 is the<br />
initial endowment (or resource), or the agent’s wealth, which the agent attempts to<br />
optimize.<br />
If we assume that the mechanism is truthful to the extent that the agents report their<br />
real type then truth‐telling is the <strong>on</strong>ly dominant strategy. This is c<strong>on</strong>sidered a<br />
truthful implementati<strong>on</strong>. The societal objective functi<strong>on</strong> is simply the aggregati<strong>on</strong> of<br />
all agents valuati<strong>on</strong>s. A ’maximizing mechanism design process’ (MDP) is called<br />
utilitarian or ’Hayekian’ if its objective functi<strong>on</strong> is the sum of all agents’ utilities.<br />
In the c<strong>on</strong>text of the internet ec<strong>on</strong>omy an MDP would enable users of network<br />
applicati<strong>on</strong>s to present their ’quality of service’ dem<strong>and</strong>s via utility functi<strong>on</strong>s<br />
defining the system performance requirements (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2013). The resource<br />
allocati<strong>on</strong> process involves ec<strong>on</strong>omic actors to perform ec<strong>on</strong>omic optimizati<strong>on</strong> given<br />
scheduling policies, load balancing <strong>and</strong> service provisi<strong>on</strong>ing.<br />
Distributed algorithmic mechanism design (DMD) for internet resource allocati<strong>on</strong> in<br />
distributed systems is akin to an equilibrium c<strong>on</strong>verging market based ec<strong>on</strong>omy<br />
where selfish agents maximize utility <strong>and</strong> firms seek to maximize profits <strong>and</strong> the<br />
state keeps an ec<strong>on</strong>omic order providing basic public goods <strong>and</strong> public safety<br />
(Feigenbaum et al, 2007). A distributed algorithmic mechanism design thus c<strong>on</strong>sists<br />
of three comp<strong>on</strong>ents: a feasible strategy space at the network nodes for each agent<br />
(or aut<strong>on</strong>omous system), an aggregated outcome functi<strong>on</strong> computed by the<br />
mechanism <strong>and</strong> a set of multi‐agent prescribed strategies induced by the<br />
mechanism.<br />
For such DMDP in place an internet ec<strong>on</strong>omy can be shown computati<strong>on</strong>ally <strong>and</strong><br />
informati<strong>on</strong>ally efficient in the sense of Hurwicz <strong>and</strong> Reiter (2006), corroborated<br />
from a CS view by C<strong>on</strong>itzer <strong>and</strong> S<strong>and</strong>holm (2002). Furthermore, for efficient macromanagement<br />
it would satisfy ’moral hazard’ <strong>and</strong> incentive based c<strong>on</strong>cerns, <strong>and</strong> in<br />
view of ’informati<strong>on</strong>al c<strong>on</strong>straints’, e.g. adverse selecti<strong>on</strong>, it may also be superior in<br />
performance since in a market based decentralized capitalistic system due to keen<br />
competiti<strong>on</strong> am<strong>on</strong>g operating managers ‐ there will be more of higher type<br />
performing managers keeping truthful payoff‐relevant informati<strong>on</strong> than in a<br />
comparative socialist planning system with plenty of lower type performing<br />
managers providing lower quality or less payoff relevant informati<strong>on</strong>.<br />
A distributed algorithmic mechanism design (Feigenbaum et al, 2007) being<br />
computati<strong>on</strong>ally efficient in a large decentralized internet ec<strong>on</strong>omy is a powerful<br />
paradigm to substantiate claims by Hayek (1945) that an industrialized ec<strong>on</strong>omy<br />
71<br />
319
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
based <strong>on</strong> market principles has an overall better performance than socialist type<br />
ec<strong>on</strong>omies of a similar nature <strong>and</strong> scale. It is a paradigm that even c<strong>on</strong>temporary<br />
theorists of MDT seem to have partially missed (Myers<strong>on</strong>, 2008) <strong>and</strong> that puts<br />
historically the socialist planning debate in a new light which ir<strong>on</strong>ically, by some<br />
proposals, has been c<strong>on</strong>ducted <strong>on</strong> the basis of computati<strong>on</strong>al feasibility <strong>and</strong><br />
superiority.<br />
Such a Hayekian MDP also extends to a dynamic real ec<strong>on</strong>omy which also invokes<br />
highly desirable properties of incentive structures (Myers<strong>on</strong>, 2008) <strong>and</strong> knowledge<br />
creati<strong>on</strong> through hi‐tech entrepreneurship. This suggests that an MDP of the<br />
Hayekian‐Hurwicz type should be more likely to generate a l<strong>on</strong>g‐run sustainable<br />
growth <strong>and</strong> development process with comparative greater welfare benefits than<br />
what any socialist type planning could achieve. This is in compliance with Hayekian<br />
development ideas as put forward recently by Easterly (2013).<br />
The focus of this paper would be to explore growth <strong>and</strong> development generating<br />
structures <strong>and</strong> factors that are compatible with a Hayek‐Hurwicz design scheme for<br />
the developing world.<br />
INDUSTRIAL AND MACRO COMPETITION<br />
The striking pattern that emerges in the innovative activities of firms is their rivalries<br />
for a technological leadership positi<strong>on</strong> in situati<strong>on</strong>s that are best described as races<br />
or hypercompetiti<strong>on</strong> (Harris & Vickers, 1987; <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006b). A race is an<br />
interactive pattern characterized by firms or nati<strong>on</strong>s c<strong>on</strong>stantly trying to get ahead of<br />
their rivals, or trying not to fall too far behind. In high‐technology industries, where<br />
customers are willing to pay a premium for advanced technology, leadership<br />
translates into increasing returns in the market through positive network<br />
externalities. Abramovitz (1986), in explaining the catch‐up hypothesis, lays stress<br />
<strong>on</strong> a country’s social capability in terms of years of educati<strong>on</strong> as a proxy of technical<br />
competence <strong>and</strong> its instituti<strong>on</strong>s. Competing behaviour is also a dynamic story of<br />
how technology unfolds in an industry. In c<strong>on</strong>trast to any existing way of looking at<br />
the evoluti<strong>on</strong> of technology, racing behaviour, though in character more ’a<br />
productivity race than a runner’s race (Abramovitz & David, 1997), recognizes the<br />
fundamental importance of strategic interacti<strong>on</strong>s between competing firms. Thus<br />
firms take their rivals’ acti<strong>on</strong>s into account when formulating their own decisi<strong>on</strong>s.<br />
The importance of this characterizati<strong>on</strong> is at least twofold. At <strong>on</strong>e level, racing<br />
behaviour has implicati<strong>on</strong>s for appreciating technology strategy at the level of the<br />
individual firm; at the other level, for underst<strong>and</strong>ing the impact of policies that aim<br />
to spur technological innovati<strong>on</strong> in an industry or country.<br />
On a nati<strong>on</strong>al scale, simple catch‐up hypotheses have put emphasis <strong>on</strong> the great<br />
potential of adopting unexploited technology in the early stage <strong>and</strong> the increase of<br />
self‐limiting power in the later stage. However, the actual growth path of the<br />
72 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
320
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
technological trajectory of a specific ec<strong>on</strong>omy may be overwhelmingly c<strong>on</strong>strained<br />
by social capability. The capability endogenously changes as states of the ec<strong>on</strong>omy<br />
<strong>and</strong> technology evolve. The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of<br />
advanced technology or the possibility of leapfrogging is mainly attributable to how<br />
the social capability evolves. In other words, which effects become more influential:<br />
growing resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of<br />
vested interests <strong>and</strong> established positi<strong>on</strong>s?<br />
Observati<strong>on</strong>s <strong>on</strong> industrial patterns in Europe, the US or Asia point to which type of<br />
racing behaviour is prevalent in global high‐ technology industries. The pattern<br />
evolving from such c<strong>on</strong>duct could be benchmarked against the fr<strong>on</strong>tier pursuit type<br />
of the global technological leaders. Another observati<strong>on</strong> relates to policy inferences<br />
<strong>on</strong> market structure, entrepreneurship, innovati<strong>on</strong> activity, industrial policy <strong>and</strong><br />
regulatory frameworks in promoting <strong>and</strong> hindering industry fr<strong>on</strong>tier races in a<br />
global industrial c<strong>on</strong>text. Does lagging behind <strong>on</strong>e’s closest technological rivals’<br />
cause a firm to increase its innovative effort? The term ‘race’ suggests that no single<br />
firm would want to fall too far behind, <strong>and</strong> that every<strong>on</strong>e would like to get ahead. If<br />
a company tries to innovate more when it is behind than when it is ahead, then<br />
‘catch‐up’ behaviour will be the dominant effect. Once a firm gets far enough ahead<br />
of its rivals, then the latter will step up their efforts to get closer. The leading<br />
company will slow down its innovative efforts until its competitors have drawn<br />
uncomfortably close or have surpassed it. Of course, the process of getting closer to<br />
may be much easier than surpassing the rival This process repeats itself every time a<br />
firm gets far enough ahead of its rivals. Of course, catch‐up may <strong>on</strong>ly c<strong>on</strong>sistently<br />
apply to the next rivals but will not impact the leader. This is called ’persistent<br />
leadership’. On a nati<strong>on</strong>al level catchup processes like this may not lead to<br />
c<strong>on</strong>vergence.<br />
An alternative behaviour pattern would corresp<strong>on</strong>d to a business increasing its<br />
innovative effort if it gets far enough ahead, thus making catch‐up by the lagging<br />
companies increasingly difficult. For any of these businesses there appears to be a<br />
clear link to market <strong>and</strong> industry structure, as termed ‘intensity of rivalry’.<br />
We investigated two different kinds of races: <strong>on</strong>e that is a fr<strong>on</strong>tier race between itself<br />
<strong>and</strong> the technological leader at any point in time (‘fr<strong>on</strong>tier‐ sticking’ behaviour), or it<br />
might try to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there<br />
are disproporti<strong>on</strong>ately large payoffs to being in the technical lead (relative to the<br />
payoffs that a firm can realize if it is simply close enough to the technical fr<strong>on</strong>tier),<br />
then <strong>on</strong>e would expect that leapfrogging behaviour would occur more frequently<br />
than fr<strong>on</strong>tier‐sticking behaviour. Alternatively, racing toward the fr<strong>on</strong>tier creates the<br />
reputati<strong>on</strong> of being an innovati<strong>on</strong> leader hoping to maintain <strong>and</strong> increase market<br />
73<br />
321
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
share in the future. All attempts to leapfrog the current technological leader might<br />
not be successful since many lagging firms might be attempting to leapfrog the<br />
leader simultaneously <strong>and</strong> the leader might be trying to get further ahead<br />
simultaneously. Corresp<strong>on</strong>dingly, <strong>on</strong>e could distinguish between attempted<br />
leapfrogging <strong>and</strong> realized leapfrogging.<br />
Am<strong>on</strong>g the key issues to be addressed is the apparent inability of technologyoriented<br />
corporati<strong>on</strong>s to maintain leadership in fields that they pi<strong>on</strong>eered. There is a<br />
presumpti<strong>on</strong> that firms fail to remain competitive because of agency problems or<br />
other suboptimal managerial behaviour within these organizati<strong>on</strong>s. An alternative<br />
explanati<strong>on</strong> is that technologically trailing firms, in symmetric competitive<br />
situati<strong>on</strong>s, will devote greater effort to innovati<strong>on</strong>, so that a failure of technological<br />
leaders to maintain their positi<strong>on</strong> is an appropriate resp<strong>on</strong>se to the competitive<br />
envir<strong>on</strong>ment. In asymmetric situati<strong>on</strong>s, with entrants challenging incumbents,<br />
research does dem<strong>on</strong>strate that startup firms show a str<strong>on</strong>ger endeavor to close up<br />
to or leapfrog the competitors. Such issues highlight the dynamics of the race within<br />
the given market structure in any of the areas c<strong>on</strong>cerned.<br />
We observe two different kinds of market asymmetries with bearing <strong>on</strong> racing<br />
behaviour: risk‐driven <strong>and</strong> resource‐based. When the incumbents’ profits are large<br />
enough <strong>and</strong> do not vary much with the product characteristics, the entrant is likely<br />
to choose the faster opti<strong>on</strong> in each stage as l<strong>on</strong>g as he has not fallen behind in the<br />
c<strong>on</strong>test. In view of resource‐based asymmetries, as a firm’s stage resource<br />
endowment increases, it could use the additi<strong>on</strong>al resources to either choose more<br />
aggressive targets or to attempt to finish the stage so<strong>on</strong>er, or both. Previous work<br />
has suggested that a firm that surges ahead of its rival increases its investment in<br />
R&D <strong>and</strong> speeds up, while a lagging firm reduces its investment <strong>and</strong> slows down.<br />
C<strong>on</strong>sequently, preceding effort suggests that the lead c<strong>on</strong>tinues to increase.<br />
However, based <strong>on</strong> related work for the US <strong>and</strong> Japanese telecommunicati<strong>on</strong>s<br />
industry when duopolistic <strong>and</strong> m<strong>on</strong>opolistic competiti<strong>on</strong> <strong>and</strong> product system<br />
complexity for new products are accounted for, the speeding up of a leading firm<br />
occurs <strong>on</strong>ly under rare circumstances. For example, a company getting far enough<br />
ahead such that the (temporary) m<strong>on</strong>opoly term dominates its payoff expressi<strong>on</strong>,<br />
will always choose the fast strategy, while a company that gets far enough behind<br />
will always choose the aggressive approach. Under these c<strong>on</strong>diti<strong>on</strong>s, the lead is<br />
likely to c<strong>on</strong>tinue to increase. If, <strong>on</strong> the other h<strong>and</strong>, both m<strong>on</strong>opoly <strong>and</strong> duopoly<br />
profits increase substantially with increased aggressiveness then even large leads<br />
can vanish with significant probability.<br />
Overall, this characterizati<strong>on</strong> highlights two forces that influence a firm’s choices in<br />
the various stages: proximity to the finish line <strong>and</strong> distance between the firms. This<br />
probability of reaping m<strong>on</strong>opoly profits is higher the farther ahead a firm is of its<br />
rival <strong>and</strong> even more so the closer the firm is to the finish line. If the lead company is<br />
74 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
322
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
far from the finish line, even a sizeable lead may not translate into the dominance of<br />
the m<strong>on</strong>opoly profit term, since there is plenty of time for the lead situati<strong>on</strong> to be<br />
reversed, <strong>and</strong> failure to finish first remains a probable outcome. In c<strong>on</strong>trast, the<br />
probability that the lagging company will get to be a m<strong>on</strong>opolist becomes smaller as<br />
it falls behind the leader. This raises the following questi<strong>on</strong>: what kind of acti<strong>on</strong>s<br />
cause a firm to get ahead? Intuitively, <strong>on</strong>e would expect that a firm that is ahead of<br />
its rival at any time t, in the sense of having completed more stages by time t, is<br />
likely to have chosen the faster strategy more often. We will c<strong>on</strong>struct numerical<br />
estimates of the probability that a leading firm is more likely to have chosen a<br />
strategy faster to verify this intuiti<strong>on</strong>.<br />
Moving away from the firm‐led race patterns revolving in a particular industry to a<br />
clustering of racing <strong>on</strong> an industry level is putting industry in different geoec<strong>on</strong>omics<br />
z<strong>on</strong>es against each other <strong>and</strong> becoming dominant in strategic<br />
product/process technologies. Here racing patterns am<strong>on</strong>g industries in a relatively<br />
free‐trade envir<strong>on</strong>ment could lead to competitive advantages <strong>and</strong> more wealth<br />
creating <strong>and</strong> accumulating dominance in key product/process technologies in <strong>on</strong>e<br />
regi<strong>on</strong> at the expense of others. There appears to be a link that individual races <strong>on</strong><br />
the firm level induce similar races <strong>on</strong> the industry level <strong>and</strong> will be a c<strong>on</strong>tributing<br />
factor to the globalizati<strong>on</strong> of network industries.<br />
Thus similar catch‐up processes are taking place between leaders <strong>and</strong> followers<br />
within a group of industrialized countries in pursuit of higher levels of productivity.<br />
Supposing that the level of labour productivity were governed entirely by the level<br />
of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the differentials in<br />
productivities am<strong>on</strong>g countries are caused by the ‘technological age’ of the stock<br />
relative to its ‘chr<strong>on</strong>ological age’. The technological age of capital is the age of<br />
expertise at the time of investment plus years elapsing from that time. Since a<br />
leading state may be supposed to be furnished with the capital stock embodying, in<br />
each vintage, technology which was ‘at the very fr<strong>on</strong>tier’ at the time of investment,<br />
the technological age of the stock is, so to speak, the same as its chr<strong>on</strong>ological age.<br />
While a leader is restricted in increasing its productivity by the advance of new<br />
technology, trailing countries have the potential to make a larger leap as they are<br />
provided with the privilege of exploiting the backlog in additi<strong>on</strong> of the newly<br />
developed technology. Hence, followers being behind with a larger gap in<br />
technology will have a str<strong>on</strong>ger potential for growth in productivity. The potential,<br />
however, will be reduced as the catch‐up process goes <strong>on</strong> because the unexploited<br />
stock of technology becomes smaller <strong>and</strong> smaller. However, as new technologies<br />
arise <strong>and</strong> are rapidly adopted in a Schumpeterian process of ‘creative destructi<strong>on</strong>’,<br />
their network effects induce rapid accelerating <strong>and</strong> cumulative growth potentials<br />
75<br />
323
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
which are catalyzed through industry competiti<strong>on</strong>. In the absence of such a process<br />
we can explain the tendency to c<strong>on</strong>vergence of productivity levels of follower<br />
countries. Historically, however, it fails to answer alleged puzzles as to why a<br />
country, such as the United States, has preserved the st<strong>and</strong>ing of the technological<br />
leader for a l<strong>on</strong>g time since taking over leadership from Britain in around the end of<br />
the nineteenth century <strong>and</strong> why the shifts have taken place in the ranks of follower<br />
states in their relative levels of productivity (i.e. technological gaps between them<br />
<strong>and</strong> the leader). Abramovitz (1986) poses some extensi<strong>on</strong>s <strong>and</strong> qualificati<strong>on</strong>s <strong>on</strong> this<br />
simple catch‐up hypothesis in an attempt to explain these facts. Am<strong>on</strong>g other factors<br />
than technological backwardness, he lays stress <strong>on</strong> a country’s social capability in<br />
terms of years of educati<strong>on</strong> as a proxy of technical competence <strong>and</strong> its political,<br />
commercial, industrial, <strong>and</strong> financial instituti<strong>on</strong>s. To become effective social<br />
capability may also include or exp<strong>and</strong> to ’deep craft’, a ’set of knowings’ <strong>on</strong><br />
technological performance <strong>and</strong> industrial techniques (Arthur, 2009). The social<br />
capacity of a state may become str<strong>on</strong>ger or weaker as technological gaps close or<br />
grow <strong>and</strong> thus Abramovitz argues that the actual catch‐up process does not provide<br />
itself to simple formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what another<br />
ec<strong>on</strong>omist, Ols<strong>on</strong> (1996), expresses to be ‘public policies <strong>and</strong> instituti<strong>on</strong>s’ as his<br />
explanati<strong>on</strong> of the great differences in per capita income across countries, stating<br />
that any poorer states that implement relatively good ec<strong>on</strong>omic policies <strong>and</strong><br />
instituti<strong>on</strong>s enjoy rapid catch‐up growth.<br />
The suggesti<strong>on</strong> should be taken seriously when we wish to underst<strong>and</strong> the<br />
technological catching‐up to American leadership by Japan, in particular during the<br />
post‐war period, <strong>and</strong> explore the possibility of a shift in st<strong>and</strong>ing between these two<br />
countries. This c<strong>on</strong>siderati<strong>on</strong> will directly bear <strong>on</strong> the future trend of the state of the<br />
art which exerts a crucial influence <strong>on</strong> the development of the world ec<strong>on</strong>omy (Juma<br />
& Clark, 2002; Fagerberg & Godinho, 2004). These explanati<strong>on</strong>s notwithst<strong>and</strong>ing, we<br />
venture as a major factor for divergent growth processes the level of intensity of the<br />
racing process within the most prevalent value‐added industries with cross‐secti<strong>on</strong>al<br />
spillovers. These are the communicati<strong>on</strong>s <strong>and</strong> informati<strong>on</strong> industries which have<br />
been shaped <strong>and</strong> led by leading American firms <strong>and</strong> where the rewards benefited<br />
their industries <strong>and</strong> country. Although European <strong>and</strong> Japanese companies were part<br />
of the race they were left behind in core markets reaping lesser benefits. (Since ICT<br />
investment relative to GDP is <strong>on</strong>ly less than half in states such as Japan, Germany<br />
<strong>and</strong> France compared to the US, 2% vs. more than 4% in 1999, this does not bode<br />
well for a rapid catch‐up in those countries <strong>and</strong> a fortiori, for the EU as a whole).<br />
Steering or guiding the process of racing through the pursuit of industrial policies<br />
aiming to increase competitive advantage of respective industries, as having been<br />
practiced in Japan, would stimulate catch‐up races but appears to be less effective in<br />
promoting fr<strong>on</strong>tier racing. Another profound reas<strong>on</strong> lies in the phenomen<strong>on</strong> of<br />
network externalities affecting ICT industries. That is, racing ahead of rivals in<br />
76 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
324
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
respective industries may create external ec<strong>on</strong>omies to the effect that such ec<strong>on</strong>omies<br />
within dominant industries tend to improve their internati<strong>on</strong>al market positi<strong>on</strong> <strong>and</strong><br />
therefore pull ahead in competitiveness vis‐à‐vis their (trading) partners. As<br />
Krugman (1991) observed: ‘It is probably true that external ec<strong>on</strong>omies are a more<br />
important determinant of internati<strong>on</strong>al trade in high technology sectors than<br />
elsewhere’. The point is that racing behaviour in leading high‐growth network<br />
industries by generating fr<strong>on</strong>tier positi<strong>on</strong>s, create critical cluster <strong>and</strong> network<br />
externalities pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong> create competitive<br />
advantages elsewhere, as supported by the increasing returns debate (Arthur, 1996).<br />
In this sense we can speak of positive externalities endogenizing growth of these<br />
ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage. All these characteristics lay<br />
the foundati<strong>on</strong>s of the ‘Network Ec<strong>on</strong>omy’.<br />
The Network Ec<strong>on</strong>omy is formed through an ever‐emerging <strong>and</strong> interacting set of<br />
increasing returns industries; it is about high‐intensity, technology driven‐racing,<br />
dynamic entrepreneurship, <strong>and</strong> focused risk‐taking through (free) venture capital<br />
markets endogenized by societal <strong>and</strong> instituti<strong>on</strong>al support. With the excepti<strong>on</strong> of<br />
pockets of activity in some parts of Europe (the UK <strong>and</strong> Sc<strong>and</strong>inavia), <strong>and</strong> in specific<br />
areas such as mobile communicati<strong>on</strong>s, these ingredients for the Network Ec<strong>on</strong>omy<br />
are <strong>on</strong>ly in the early stage of emerging in C<strong>on</strong>tinental Europe, <strong>and</strong> the political<br />
mindset in support of the Network Ec<strong>on</strong>omy is anything but prevalent. As l<strong>on</strong>g as<br />
we do not see a significant shift toward movements in this directi<strong>on</strong>, Europe will not<br />
see the full benefits of the Network Ec<strong>on</strong>omy within a Global Ec<strong>on</strong>omy.<br />
Racing behaviour <strong>on</strong> technological positi<strong>on</strong>s am<strong>on</strong>g firms in high‐ technology<br />
industries, as exemplified by the globally operating telecommunicati<strong>on</strong>s <strong>and</strong><br />
computer industries, produce spillover benefits in terms of increasing returns <strong>and</strong><br />
widespread productivity gains. Due to relentless competiti<strong>on</strong> am<strong>on</strong>g technological<br />
leaders the network effects result in significant advantages in the value added to this<br />
industry c<strong>on</strong>tributing to faster growth of GDP, <strong>and</strong> through a flexible labour market,<br />
also to employment growth. This c<strong>on</strong>stitutes a new paradigm in ec<strong>on</strong>omic thinking<br />
through network ec<strong>on</strong>omies <strong>and</strong> is a major gauge to compare the wealth‐creating<br />
power of the US ec<strong>on</strong>omy over the past decade against the European <strong>and</strong> advanced<br />
Asian ec<strong>on</strong>omies. It is interesting to speculate <strong>on</strong> the implicati<strong>on</strong>s of the way<br />
companies in major high‐technology markets, such as telecommunicati<strong>on</strong>s, split<br />
clearly into the two major technology races, with <strong>on</strong>e group of firms clearly lagging<br />
the other.<br />
The trajectories of technological evoluti<strong>on</strong> certainly seem to suggest that firms from<br />
<strong>on</strong>e fr<strong>on</strong>tier cannot simply jump to another trajectory. Witness, in this regard, the<br />
gradual process necessary for a firm in the catch‐up race to approach those in the<br />
77<br />
325
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
fr<strong>on</strong>tier race. There appears to be a fr<strong>on</strong>tier ‘lock‐in’, in that <strong>on</strong>ce a company is part<br />
of a race, the group of rivals within that same race are the <strong>on</strong>es whose acti<strong>on</strong>s<br />
influence that company’s strategy the most. Advancing technological capability is a<br />
cumulative process. The ability to advance to a given level of technical capability<br />
appears to be a functi<strong>on</strong> of existing technical capability. Given this path dependence,<br />
the questi<strong>on</strong> remains: why do some firms apparently choose a path of technological<br />
evoluti<strong>on</strong> that is less rapid than others? Two sets of possible explanati<strong>on</strong>s could be<br />
derived from our case analysis, which need not be mutually exclusive. The first<br />
explanati<strong>on</strong> lingers primarily <strong>on</strong> the expensive nature of R&D in industries like<br />
telecommunicati<strong>on</strong>s <strong>and</strong> computers which rely <strong>on</strong> novel discovery for their<br />
advancement. Firms choosing the catch‐up race will gain access to a particular<br />
technical level later than those choosing the fr<strong>on</strong>tier, but will do so at a lower cost.<br />
TECHNOLOGICAL FRONTIERS<br />
The evoluti<strong>on</strong> of a cross secti<strong>on</strong> of high technology industries reflects repetitive<br />
strategic interacti<strong>on</strong>s between companies in a c<strong>on</strong>tinuous quest to dominate the<br />
industry or at least to improve its competitive positi<strong>on</strong> through company level <strong>and</strong><br />
industry level technological evoluti<strong>on</strong>. We can observe several racing patterns across<br />
industries, each of which is the result of a subset of firms jockeying for a positi<strong>on</strong><br />
either as a race leader or for a positi<strong>on</strong> near the leader c<strong>on</strong>stituting a leadership club.<br />
The identificati<strong>on</strong> <strong>and</strong> interpretati<strong>on</strong> of the races relies <strong>on</strong> the fact that different firms<br />
take very different technological paths to target a superior performance level with<br />
the reward of increasing market shares, maintaining higher productivity <strong>and</strong><br />
profitability. In a Schumpeterian framework such races cannot be interpreted in a<br />
free‐riding situati<strong>on</strong> where <strong>on</strong>e firm exp<strong>and</strong>s resources in advancing the state of<br />
technology <strong>and</strong> the others follow closely behind. Such spillover interpretati<strong>on</strong>s are<br />
suspect when products are in the domain of high complexity, of high risk in<br />
succeeding, <strong>and</strong> different firms typically adopt different procedural <strong>and</strong><br />
architectural approaches.<br />
The logic underlying this evoluti<strong>on</strong> holds in any industry in which two broad sets of<br />
c<strong>on</strong>diti<strong>on</strong>s are satisfied. First, it pays for a firm to have a technological lead over its<br />
rival; it also boosts its market image <strong>and</strong> enhances its reputati<strong>on</strong>al capital. Sec<strong>on</strong>d,<br />
for various levels of technological complexity am<strong>on</strong>g the products introduced by<br />
various firms, technological complexity can be represented by a multi‐criteria<br />
performance measure, that is, by a vector‐valued distance measure. The collecti<strong>on</strong> of<br />
performance indicators, parameters, being c<strong>on</strong>nected with each other for individual<br />
companies form an envelope that shapes a ‘technological fr<strong>on</strong>tier’. The technological<br />
fr<strong>on</strong>tier is in fact a reas<strong>on</strong>able indicator of the evolving state of knowledge (technical<br />
expertise) in the industry. At any point in time the industry technology fr<strong>on</strong>tier (ITF)<br />
indicates the degree of technical sophisticati<strong>on</strong> of the most advanced products<br />
carried by companies in that industry in view of comparable performance st<strong>and</strong>ards.<br />
78 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
326
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Firm level technology fr<strong>on</strong>tiers (FTF) are c<strong>on</strong>structed analogously <strong>and</strong> indicate, at<br />
any point in time, the extent of technical sophisticati<strong>on</strong> achieved by the firm until<br />
that point in time. The evoluti<strong>on</strong> of company <strong>and</strong> industry level fr<strong>on</strong>tiers is highly<br />
interactive. Groups of company fr<strong>on</strong>tiers are seen to co‐evolve in a manner that<br />
suggests that the respective firms are racing to catch up with, <strong>and</strong> get ahead of each<br />
other.<br />
A data set could focus <strong>on</strong> a given set of products (systems) by major European,<br />
American or Asian enterprises in those industries for a sufficiently representative<br />
period of market evoluti<strong>on</strong>. In principle, we can identify at least two races in<br />
progress in the industries throughout a given period of durati<strong>on</strong>. One comprises the<br />
world fr<strong>on</strong>tier race in each of those industries, the other, for example, the European<br />
fr<strong>on</strong>tier race which technically would c<strong>on</strong>stitute a subfr<strong>on</strong>tier to the worldwide race.<br />
The aggregate technology fr<strong>on</strong>tier of the firms in a particular race (that is, ITF) is<br />
c<strong>on</strong>structed in a manner similar to the individual FTFs. Essentially, the maximal<br />
envelope of the FTFs in a particular race c<strong>on</strong>stitutes the ITF for that race. The ITF<br />
indicates, as a functi<strong>on</strong> of calendar time, the best achievable performance by any<br />
firm in the race at a given date.<br />
A statistical profiling of technological evoluti<strong>on</strong> <strong>and</strong> innovati<strong>on</strong> relates to<br />
competitive racing am<strong>on</strong>g rival companies. Am<strong>on</strong>g the (n<strong>on</strong>‐exclusive) performance<br />
criteria to be assessed are: (1) frequency of fr<strong>on</strong>tier pushing; (2) technological<br />
dominati<strong>on</strong> period; (3) innovati<strong>on</strong>s vs. imitati<strong>on</strong>s in the race; (4) innovati<strong>on</strong><br />
frequency when behind or ahead; (5) nature of jumps, leapfrogging or fr<strong>on</strong>tiersticking;<br />
(6) inter‐jump times <strong>and</strong> jump sizes; (7) race closeness measures; (8) interfr<strong>on</strong>tier<br />
distance; (9) market leading through ‘market making’ innovati<strong>on</strong>s; <strong>and</strong> (10)<br />
leadership in ‘innovati<strong>on</strong> markets’.<br />
A race may or may not have different firms in the leadership positi<strong>on</strong> at different<br />
times. It may be a tighter race at some times than at others, <strong>and</strong> in general, may<br />
exhibit a variety of forms of interesting behaviour. While analysis of racing<br />
behaviour is left to various interpretati<strong>on</strong>s, it is appropriate to ask why the firms are<br />
motivated to keep <strong>on</strong> racing at all. As access to superior technology exp<strong>and</strong>s the<br />
scope of opportunities available to the firms, the technology can be applied in a<br />
range of markets. However, leading edge technology is acquired at a cost. It seems<br />
unlikely that all the companies would find it profitable to compete to be at the<br />
leading edge all the time. Also not every firm has access to equal capabilities in<br />
leveraging a given level of technological resources. Firms may, for example, be<br />
expected to differ in their access to complementary assets that allows them to<br />
appropriately reap the benefits from their innovati<strong>on</strong>. It is reas<strong>on</strong>able to assume that<br />
whatever the level of competence of a company in exploiting its resources it will be<br />
79<br />
327
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
better off the more advanced the technology. Based <strong>on</strong> this procedure an analysis<br />
will show how dynamic competiti<strong>on</strong> evolved in the past.<br />
Unlike other (statistical) indicators (such as patent statistics) referring to the degree<br />
of competitiveness am<strong>on</strong>g industries, regi<strong>on</strong>s <strong>and</strong> countries c<strong>on</strong>cerned, the proposed<br />
measures cover behavioral dynamic movements in respective industries, <strong>and</strong><br />
therefore are able to lend intrinsic predictive value to crucial ec<strong>on</strong>omic variables<br />
relating to ec<strong>on</strong>omic growth <strong>and</strong> wealth creati<strong>on</strong>. The results are likely to provide<br />
strategic support for industrial <strong>and</strong> technology policy in a regi<strong>on</strong>al or nati<strong>on</strong>al<br />
c<strong>on</strong>text <strong>and</strong> will enable policy makers to identify strengths <strong>and</strong> weaknesses of<br />
relevant players <strong>and</strong> their envir<strong>on</strong>ments in those markets. While this process looks<br />
like a micro representati<strong>on</strong> of dynamic technological evoluti<strong>on</strong> driving companies<br />
<strong>and</strong> industries into leadership positi<strong>on</strong>s, we may c<strong>on</strong>strue an analogous process that<br />
drives a regi<strong>on</strong> or a nati<strong>on</strong> into advancement <strong>on</strong> a macro scale in order to achieve a<br />
higher level pecking order am<strong>on</strong>g its peers. This may allow using the micro<br />
foundati<strong>on</strong>s of racing as a basis for identifying clubs of nati<strong>on</strong>s or regi<strong>on</strong>s am<strong>on</strong>g<br />
them to achieve higher levels <strong>and</strong> rates of growth.<br />
Catch‐up or Leapfrogging<br />
It was Schumpeter (1947) who observed that it is the expectati<strong>on</strong> of supernormal<br />
profits from a temporary m<strong>on</strong>opoly positi<strong>on</strong> following an innovati<strong>on</strong> that is the chief<br />
driver of R & D investment. Al<strong>on</strong>g this line, the simplest technology race model can<br />
be explained as follows: A number of firms invest in R&D. Their investment results<br />
in an innovati<strong>on</strong> with the time spent in R&D subject to some varying level of<br />
uncertainty. However, a greater investment reduces the expected time to completi<strong>on</strong><br />
of R&D. The model investigates how many firms will choose to enter such a c<strong>on</strong>test,<br />
<strong>and</strong> how much they will invest. However, despite some extensive theoretical<br />
examinati<strong>on</strong> of technological races there have been very few empirical studies <strong>on</strong><br />
this subject (Lerner, 1997) <strong>and</strong> virtually n<strong>on</strong>e in the c<strong>on</strong>text of major global<br />
industries, <strong>and</strong> <strong>on</strong> a comparative basis.<br />
Technological fr<strong>on</strong>tiers at the firm <strong>and</strong> industry race levels offer a powerful tool<br />
through which to view evolving technologies within an industry. By providing a<br />
benchmarking roadmap that shows where an individual firm is relative to the other<br />
firms in the industry, they highlight the importance of strategic interacti<strong>on</strong>s in the<br />
firm’s technology decisi<strong>on</strong>s. From the interactive process of racing could emerge<br />
various behavioural patterns. Does lagging behind <strong>on</strong>e’s closest technological rivals<br />
cause a firm to increase its innovative effort? The term ‘race’ suggests that no single<br />
company would want to fall too far behind, <strong>and</strong> that every<strong>on</strong>e would like to get<br />
ahead. If a firm tries to innovate more when it is behind than when it is ahead, then<br />
‘catch‐up’ behaviour will be the dominant effect. Once a firm gets ahead of its rivals<br />
noticeably, then rivals will step up their efforts to catch up. The leader will slow<br />
down its innovative efforts until its rivals have drawn uncomfortably close or have<br />
80 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
328
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
surpassed it. This process repeats itself every time a company gets far enough ahead<br />
of its rivals. An alternative behaviour pattern would corresp<strong>on</strong>d to a firm increasing<br />
its innovative effort if it gets far enough ahead, thus making catch‐up by the lagging<br />
firms increasingly difficult. This looks like the ‘Intel Model’ where <strong>on</strong>ly the paranoid<br />
survives (Grove, 1992). For any of these forms there appears to be a clear link to<br />
market <strong>and</strong> industry structure, as termed ‘intensity of rivalry’ by Kamien <strong>and</strong><br />
Schwarz (1982).<br />
We group two different kinds of races: <strong>on</strong>e that is a fr<strong>on</strong>tier race am<strong>on</strong>g leaders <strong>and</strong><br />
would‐be leaders (first league) <strong>and</strong> another that is a catch‐up race am<strong>on</strong>g laggards<br />
<strong>and</strong> imitators (sec<strong>on</strong>d league). Though both leagues may play their own game, in a<br />
free market c<strong>on</strong>test, it would be possible that a member of the sec<strong>on</strong>d league may<br />
penetrate into the first, as <strong>on</strong>e in the first league may fall back into the sec<strong>on</strong>d.<br />
Another aspect of innovati<strong>on</strong> speed has been addressed by Kessler <strong>and</strong> Bierly (2002).<br />
As a general rule they found that the speed to racing ahead may be less significant<br />
the more “radical” (drastic) the innovati<strong>on</strong> appears to be <strong>and</strong> the more likely it leads<br />
to a dominant design. These two forms have been applied empirically to the<br />
development of the early Japanese computer industry (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006a), that is, a<br />
fr<strong>on</strong>tier racing model regarding the struggle for technological leadership in the<br />
global industry between IBM <strong>and</strong> ‘Japan Inc.’ guided by MITI (now METI), <strong>and</strong> a<br />
catch‐up racing model relating to competiti<strong>on</strong> am<strong>on</strong>g the leading Japanese<br />
mainframe manufacturers as laggards.<br />
It is also interesting to distinguish between two sub‐categories of catch‐up<br />
behaviour. A lagging firm might simply try to close the gap between itself <strong>and</strong> the<br />
technological leader at any point in time (‘fr<strong>on</strong>tier‐sticking’ behaviour), or it might<br />
try to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there are<br />
disproporti<strong>on</strong>ately large payoffs to being in the technical lead (relative to the payoffs<br />
that a firm can realize if it is simply close enough to the technical fr<strong>on</strong>tier), then <strong>on</strong>e<br />
would expect that leapfrogging behaviour would occur more frequently than<br />
fr<strong>on</strong>tier‐sticking behaviour (Owen & Ulph, 1994). Alternatively, racing toward the<br />
fr<strong>on</strong>tier creates the ‘reputati<strong>on</strong>’ of being an innovati<strong>on</strong> leader facilitating to maintain<br />
<strong>and</strong> increase market share in the future (Albach, 1997). All attempts to leapfrog the<br />
current technological leader might not be successful since many lagging firms might<br />
be attempting to leapfrog the leader simultaneously <strong>and</strong> the leader might be trying<br />
to get further ahead simultaneously. Corresp<strong>on</strong>dingly, <strong>on</strong>e should distinguish<br />
between attempted leapfrogging <strong>and</strong> realized leapfrogging. This phenomen<strong>on</strong><br />
(though dependent <strong>on</strong> industry structure) appears as the predominant behaviour<br />
pattern in the US <strong>and</strong> Japan fr<strong>on</strong>tier races (Brezis et al, 1991). Albach (1993) cites<br />
studies for Germany that show otherwise.<br />
81<br />
329
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
Leapfrogging behaviour influenced by the expected size of payoffs as suggested by<br />
Owen <strong>and</strong> Ulph (1994) might be revised in compliance with the characteristics of<br />
industrial structure of the local (regi<strong>on</strong>al) markets, the amount of R&D efforts for<br />
leapfrogging <strong>and</strong> the extent of globalizati<strong>on</strong> of the industry. Even in the case where<br />
the payoffs of being in the technological lead are expected to be disproporti<strong>on</strong>ately<br />
large, the lagging companies might be satisfied to remain close enough to the leader<br />
so as to gain or maintain a share in the local market. This could occur when the<br />
amount of R&D efforts (expenditures) required for leapfrogging would be too large<br />
for a lagging firm to be viable in the industry <strong>and</strong> when the local market has not<br />
been open enough for global competiti<strong>on</strong>: the local market might be protected for<br />
the lagging local companies under the auspices of measures of regulati<strong>on</strong> by the<br />
government (e.g. government purchasing, c<strong>on</strong>trols <strong>on</strong> foreign capital) <strong>and</strong> the<br />
c<strong>on</strong>diti<strong>on</strong>s preferable for these firms (e.g. language, marketing practices).<br />
When the industrial structure is composed of multi‐product companies, as for<br />
example it used to be in the Japanese computer industry, sub‐fr<strong>on</strong>tier firms may<br />
derive spill over benefits in developing new products in other technologically<br />
related fields (e.g. communicati<strong>on</strong>s equipment, c<strong>on</strong>sumer electr<strong>on</strong>ic products). These<br />
companies may prefer an R&D strategy just to keep up with the technological<br />
fr<strong>on</strong>tier level (catch‐up) through realizing a greater profit stream over a whole range<br />
of products.<br />
What are the implicati<strong>on</strong>s of the way firms split cleanly into the two technology<br />
races, with <strong>on</strong>e group clearly lagging the other technologically? The trajectories of<br />
technological evoluti<strong>on</strong> certainly seem to suggest that firms from <strong>on</strong>e fr<strong>on</strong>tier cannot<br />
simply jump to another trajectory. Witness, in this regards the gradual process<br />
necessary for the companies in the Japanese fr<strong>on</strong>tier to catch up with those at the<br />
global fr<strong>on</strong>tier. There appears to be a fr<strong>on</strong>t line ‘lock‐in’ in that <strong>on</strong>ce a firm is part of<br />
a race, the group of rivals within that same race are the <strong>on</strong>es whose acti<strong>on</strong>s influence<br />
the firm’s strategy the most.<br />
Advancing technological capability is a cumulative process. The ability to advance to<br />
a given level of technical capability appears to be a functi<strong>on</strong> of existing technical<br />
potential. Given this ‘path dependence’, the questi<strong>on</strong> remains: why do some firms<br />
apparently choose a path of technological evoluti<strong>on</strong> that is less rapid than others<br />
are? We propose two sets of possible explanati<strong>on</strong>s, which need not to be mutually<br />
exclusive. The first explanati<strong>on</strong> hinges primarily <strong>on</strong> the expensive nature of R & D in<br />
industries like the computer industry, which rely <strong>on</strong> novel scientific discovery for<br />
their advancement. Firms choosing the subfr<strong>on</strong>tier will gain access to a particular<br />
technical level later than those choosing the fr<strong>on</strong>tier, but will do so at a lower cost.<br />
Expending fewer resources <strong>on</strong> R & D ensures a slower rate of technical evoluti<strong>on</strong>.<br />
The sec<strong>on</strong>d explanati<strong>on</strong> relates mainly to technological spillovers. Following the<br />
success of the fr<strong>on</strong>tier firms in achieving a certain performance level, these become<br />
82 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
330
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
known to the subfr<strong>on</strong>tier firms. In fact, leading edge research in the computer<br />
industry is usually reported in patent applicati<strong>on</strong>s <strong>and</strong> scientific journals <strong>and</strong> is<br />
widely disseminated throughout the industry. The hypothesis is that partial<br />
spillover of knowledge occurs to the subfr<strong>on</strong>tier firms, whose task is then simplified<br />
to some extent. Notice that the subfr<strong>on</strong>tier firms still need to race to be technological<br />
leaders, as evidenced by the analysis above. This implies that the spillovers are<br />
nowhere near perfect. Company specific learning is still the norm. However, it is<br />
possible that knowing something about what research avenues have proved<br />
successful (for the fr<strong>on</strong>tier firms) could greatly ease the task for the firms that follow<br />
<strong>and</strong> try to match the technical level of the fr<strong>on</strong>tier company.<br />
Statistical Metrics of Industrial Racing Patterns<br />
Statistically descriptive measures of racing behaviour can be established that reflect<br />
the richness of the dynamics of ec<strong>on</strong>omic growth am<strong>on</strong>g competing nati<strong>on</strong>s. The<br />
point of departure for a statistical analysis of industrial racing patterns is the<br />
aggregate technological fr<strong>on</strong>tier represented by the nati<strong>on</strong>al producti<strong>on</strong> functi<strong>on</strong> as a<br />
reas<strong>on</strong>able indicator of the evolving state of knowledge (technical expertise) in a<br />
nati<strong>on</strong> or regi<strong>on</strong> which is the weighted aggregate of all industries or activities that<br />
themselves are represented by industry technology fr<strong>on</strong>tier (ITF). Firm level<br />
technology fr<strong>on</strong>tiers (FTF) are c<strong>on</strong>structed analogously <strong>and</strong> indicate, at any point in<br />
time, the weighted c<strong>on</strong>tributi<strong>on</strong> of that firm to the industry <strong>on</strong> st<strong>and</strong>ard industry<br />
classificati<strong>on</strong>.<br />
In this c<strong>on</strong>text we define ‘race’ as a c<strong>on</strong>tinual c<strong>on</strong>test for technological superiority<br />
am<strong>on</strong>g nati<strong>on</strong>s or regi<strong>on</strong>s with key industries. Under this c<strong>on</strong>ceptualisati<strong>on</strong> a race is<br />
characterised by a number of countries whose ITF’s remain ‘close’ together over a<br />
period (T) of, say, 25 to 50 years. The distinctive element is that countries engaging<br />
in a competiti<strong>on</strong> have ITF’s substantially closer together than those of any company<br />
not in the race. A statistical analysis should reflect that a race, as defined, may or<br />
may not have different countries in the leadership positi<strong>on</strong> at different times. It may<br />
be a tighter c<strong>on</strong>test at some times than at others, <strong>and</strong> in general, may exhibit a<br />
variety of forms of industrial behaviour. We look for clusters of firms who’s ITFs<br />
remain close enough throughout the durati<strong>on</strong> (formal measures of closeness are<br />
defined <strong>and</strong> measured). We identify races to take place at any level of industrial<br />
performance between the very top <strong>and</strong> the very bottom throughout 50 years<br />
durati<strong>on</strong> that is racing from the bottom to racing to the top.<br />
One comprises the world fr<strong>on</strong>tier race in each of those industries, the other a<br />
subfr<strong>on</strong>tier race (say, North America, Europe, East Asia, China, India, Latin<br />
America, Africa) which technically would c<strong>on</strong>stitute a subfr<strong>on</strong>tier to the world,<br />
allowing under the best of circumstances for the subfr<strong>on</strong>tier to be the fr<strong>on</strong>tier. Since<br />
83<br />
331
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
the level <strong>and</strong> breadth of industrial activity is reflected as an indicator for ec<strong>on</strong>omic<br />
welfare, racing to the top would go parallel with ec<strong>on</strong>omic growth <strong>and</strong> welfare<br />
enhancing, whereas racing from the bottom would corresp<strong>on</strong>d to poverty reducti<strong>on</strong><br />
<strong>and</strong> avoiding stati<strong>on</strong>ary (under)development traps.<br />
Characterizati<strong>on</strong> of Statistical Indicators of Industrial Racing<br />
While a variety of situati<strong>on</strong>s are possible, the extremes are the following: (a) <strong>on</strong>e<br />
country may push the fr<strong>on</strong>tier at all times, with the others following closely behind,<br />
(b) some countries share more or less equally in the task of advancing the most value<br />
generating industry technology fr<strong>on</strong>tiers (ITFs). Depending <strong>on</strong> the situati<strong>on</strong> the most<br />
value generating industries may be high technology based increasing returns or<br />
network industries that are able to induce complementary emerging industries with<br />
high potentials. Extreme situati<strong>on</strong> (a) corresp<strong>on</strong>ds to the existence of a unique<br />
technological leader for a particular race, <strong>and</strong> a number of quick followers. Situati<strong>on</strong><br />
(b), <strong>on</strong> the other h<strong>and</strong>, corresp<strong>on</strong>ds to the existence of multiple technological leaders.<br />
Assessment of Fr<strong>on</strong>tier Pushing: .The relevant statistics for the races relate to counting<br />
the times the ITFs are pushed forward by countries or regi<strong>on</strong>s at large within a<br />
global or regi<strong>on</strong>al fr<strong>on</strong>tier. Fr<strong>on</strong>tier pushing can be triggered through industrial<br />
policy by governments or well fostered entrepreneurship in an advanced capitalistic<br />
system<br />
Dominati<strong>on</strong> Period Statistics: Accepting the view that a country/regi<strong>on</strong> has greater<br />
potential to earn income <strong>and</strong> build wealth from its technological positi<strong>on</strong> if it is<br />
ahead of its race suggests that it would be interesting to examine the durati<strong>on</strong> of<br />
time for which a country can expect to remain ahead <strong>on</strong>ce it finds itself pushing its<br />
ITF. We statistically define the ‘dominati<strong>on</strong> period’ to be the durati<strong>on</strong> of time for<br />
which a country leads its particular race. It is interesting to note that the mean<br />
dominati<strong>on</strong> period is virtually indistinguishable for the three races, <strong>and</strong> lies between<br />
three <strong>and</strong> four years. A difference of means test cannot reject the hypothesis that the<br />
mean years of dominati<strong>on</strong> tend to cluster but hardly c<strong>on</strong>verge. So countries in each<br />
of the races can expect to remain ahead approximately in proporti<strong>on</strong> to their<br />
technological capability <strong>and</strong> more than the amount of time after they have propelled<br />
themselves to the fr<strong>on</strong>t of their respective races. However, the dominati<strong>on</strong> period<br />
tends to be a more uncertain quantity in the world fr<strong>on</strong>tier race, to a lesser degree in<br />
the EU fr<strong>on</strong>tier race than in any smaller regi<strong>on</strong>al races (as evidenced by the lower<br />
dominati<strong>on</strong> period st<strong>and</strong>ard deviati<strong>on</strong>).<br />
Catch‐up Statistics: If key industries of a country push to innovate more when they<br />
are behind than when they are ahead, then ‘catch‐ up’ behaviour will be the<br />
dominant effect. For each country/regi<strong>on</strong>, these statistics compare the fracti<strong>on</strong> of the<br />
total innovati<strong>on</strong>s carried out by industries in that country (i.e. the fracti<strong>on</strong> of the total<br />
number of times that their ITFs advance) when it was engaging in its race when<br />
lagging, with the fracti<strong>on</strong> of times that the country actually led its race. In the<br />
84 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
332
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
absence of catch‐up behaviour, or behaviour leading to a country increasingly<br />
dominating its rivals, we would expect to see no difference in these fracti<strong>on</strong>s. Then<br />
the fracti<strong>on</strong> of time that a country is ahead of its race could be an unbiased estimator<br />
of the fracti<strong>on</strong> of innovati<strong>on</strong>s in its key industries that it engages in when it is ahead.<br />
Relevant data, however, suggest that this is usually not the case. They appear to<br />
show that the fracti<strong>on</strong> of times a state leads its race at any development level in a<br />
group or club is larger than the fracti<strong>on</strong> of innovati<strong>on</strong>s that occur when the country<br />
is ahead, i.e. more innovati<strong>on</strong>s occur when the country is lagging than would be<br />
expected in the absence of catch‐up or increasing dominance behaviour. A major<br />
excepti<strong>on</strong> would arise if the country would act like an ‘Intel Ec<strong>on</strong>omy’, where<br />
unchallenged leadership in key industries creates incentives to increase the lead to<br />
its rivals. Catch‐up behaviour is supported by additi<strong>on</strong>al observati<strong>on</strong>s, as derivable<br />
from c<strong>on</strong>vergence <strong>and</strong> c<strong>on</strong>diti<strong>on</strong>al c<strong>on</strong>vergence in the ec<strong>on</strong>omic growth process that<br />
countries make larger jumps (i.e. the ITFs advance more) when they are behind than<br />
when they are leading the race<br />
Leapfrogging Statistics: From this, the distincti<strong>on</strong> emerges between two kinds of catchup.<br />
A lagging country might simply try to close the gap between itself <strong>and</strong> the<br />
technological leader at any point in time (fr<strong>on</strong>tier‐sticking behaviour), or it might try<br />
to actually usurp the positi<strong>on</strong> of the leader by ‘leapfrogging’ it. When there are<br />
disproporti<strong>on</strong>al larger incomes per head when being in the technical lead (relative to<br />
a situati<strong>on</strong> that a country can realize if it is simply close enough to the technological<br />
fr<strong>on</strong>tier), then <strong>on</strong>e would expect that leapfrogging behaviour would make it a more<br />
attractive incentive than fr<strong>on</strong>tier‐sticking behaviour.<br />
All attempts to leapfrog the current technological leader might not be successful<br />
since many lagging firms/industries might be attempting to leapfrog the leader<br />
simultaneously. Corresp<strong>on</strong>dingly, we observe both the attempted leapfroggings <strong>and</strong><br />
the realized leapfroggings. It appears likely that the leapfrogging phenomen<strong>on</strong><br />
would be more predominant in the premier league than in following up leagues.<br />
Interfr<strong>on</strong>tier Distance: How l<strong>on</strong>g does ‘knowledge’ take to spillover from fr<strong>on</strong>tier to<br />
subfr<strong>on</strong>tier industries? This requires investigating “interfr<strong>on</strong>tier distance”. One<br />
measure of how much subfr<strong>on</strong>tier industries’ technology lags the fr<strong>on</strong>tier industries’<br />
technology could be graphed as “subfr<strong>on</strong>tier lag” in terms of calendar time. At each<br />
point in time, this is simply the absolute difference in the subfr<strong>on</strong>tier performance<br />
<strong>and</strong> the fr<strong>on</strong>tier performance time. The graph would clearly indicate that this<br />
measure has been declining or increasing more or less m<strong>on</strong>ot<strong>on</strong>ically over the past<br />
50 years to the extent that the subfr<strong>on</strong>tier industries have been able/unable to catch<br />
up with the fr<strong>on</strong>tier industries. A complementary measure would be to assess the<br />
difficulty of bridging the lag. That is, how much l<strong>on</strong>ger does it take the subfr<strong>on</strong>tier to<br />
85<br />
333
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
reach a certain level of technical achievement after the fr<strong>on</strong>tier has reached that<br />
level? Thus it might very well turn out that the interfr<strong>on</strong>tier distance may be<br />
decreasing though the difficulty in bridging the gap is increasing.<br />
Race Closeness Measure (RCM): N<strong>on</strong>e of the previous analyses tell us how close any of<br />
the overall races are over a period of time. The races are all distant/close by<br />
c<strong>on</strong>structi<strong>on</strong>, however, some might be closer than others, We define ‘a measure of<br />
closeness’ of a race (RCM) at a particular time as follows:<br />
RCM (t) = 0 N Fi (t) – Fj (t) 2 /N (t) (1)<br />
where Fi (t) is country’s i ITF at time t, Fj(t) is country’s j comparable ITF at time<br />
t = max ITF(t) for each i, j <strong>and</strong> N(t) is the number of active key value‐generating<br />
industries at time t.<br />
The measure (Equati<strong>on</strong> 1) thus c<strong>on</strong>structed has a lowest value of 0, which<br />
corresp<strong>on</strong>ds to a ‘dead heat’ race. Higher values of the measure corresp<strong>on</strong>d to races<br />
that are less close. Unlike the earlier characteristics (dominati<strong>on</strong> period length,<br />
innovati<strong>on</strong> when ahead versus when behind, leapfrogging versus fr<strong>on</strong>tier‐sticking)<br />
which investigate the behaviour of a particular feature of the race <strong>and</strong> of a particular<br />
industry in relati<strong>on</strong> to the race fr<strong>on</strong>tier, the RCM is more of an aggregate statistic of<br />
how close the various racing parties are at a point in time. The closeness measure is<br />
simply an indicati<strong>on</strong> of parity, <strong>and</strong> not <strong>on</strong>e that says anything per se about the<br />
evoluti<strong>on</strong> of the technological fr<strong>on</strong>tier. To see this, note that if n<strong>on</strong>e of the fr<strong>on</strong>tiers<br />
were evolving, the closeness measure would be 0, as it would be if all the fr<strong>on</strong>tiers<br />
were advancing in perfect lock‐step with <strong>on</strong>e another.<br />
TABLE 1. ITF SHIFTS ACROSS AGGREGATED INDUSTRIES<br />
Aggregate Industries 1980‐ ITF (max = 100) 2010 GDP (%)<br />
US 80 85 70<br />
EU 60 75 60<br />
China 15 60 50<br />
USSR 30 35 40<br />
India 25 40 30<br />
Brazil 20 30 25<br />
Japan 70 70 65<br />
We talk about value‐added increasing returns industries over a period of 30 years.<br />
The industries comprise ICT, C<strong>on</strong>sumer Electr<strong>on</strong>ics, Chemicals <strong>and</strong> Materials,<br />
Automobiles, Pharma/Biotech, Machine Tools, Medical Instruments,<br />
Aerospace/Defense, Energy Technologies, <strong>and</strong> HT Transportati<strong>on</strong> Systems. Industry<br />
sectors can be assigned to various countries/regi<strong>on</strong>s such as US, EU, China, Russia,<br />
India, Brazil, Japan (Table 1). We benchmark the industry technology fr<strong>on</strong>tiers (ITFs)<br />
accordingly, that is, highest ‘state of knowledge’ at time t is 100 pc. The countries’<br />
rank to the max ITFs is assessed as the share of the max ITF. The assessment<br />
intervals are spaced in five year intervals starting in 1980 until 2010.<br />
86 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
334
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Efficiency<br />
Let’s explore the inefficiency of the follower nati<strong>on</strong>s; i.e., the negative effect <strong>on</strong> the<br />
potential technology gap stemming from inefficient social <strong>and</strong> instituti<strong>on</strong>al factors. A<br />
good example of cross industrial inefficiencies over a historically representative<br />
period (1810‐2000) is Russia that was hardly advancing ec<strong>on</strong>omically against<br />
underdeveloped benchmark countries <strong>and</strong> falling behind leading ec<strong>on</strong>omies,<br />
reinforced through the bolshevik revoluti<strong>on</strong> <strong>and</strong> its underperforming ec<strong>on</strong>omic<br />
mechanism design (Gaidar, 2012). Increasing efficiencies deblock catch‐up in lagging<br />
countries (Juma & Clark, 2002). Efficiency is found by dividing a nati<strong>on</strong>’s estimated<br />
fixed effect by the regi<strong>on</strong>al adopti<strong>on</strong> rate. As defined here, it is quite robust to<br />
different estimati<strong>on</strong>s <strong>and</strong> samples. The relative efficiencies of the nati<strong>on</strong>s within<br />
regi<strong>on</strong>s appear to c<strong>on</strong>form to comm<strong>on</strong> beliefs. For example, in Europe, the<br />
Netherl<strong>and</strong>s, Belgium <strong>and</strong> Switzerl<strong>and</strong> are the most efficient while Turkey, Portugal<br />
<strong>and</strong> Greece are the least efficient. In East Asia, H<strong>on</strong>g K<strong>on</strong>g is the most efficient while<br />
Ind<strong>on</strong>esia <strong>and</strong> Thail<strong>and</strong> are the least efficient. Finally, in Latin America, Mexico <strong>and</strong><br />
Argentina are at the top <strong>and</strong> H<strong>on</strong>duras <strong>and</strong> Bolivia at the bottom. Another way to<br />
discuss the findings is to c<strong>on</strong>sider the time required to catch‐up. Previously, Parente<br />
<strong>and</strong> Prescott (2004) showed that countries with lower levels of relative efficiency will<br />
adopt modern technologies at much later dates. C<strong>on</strong>versely, <strong>on</strong>e could argue that if<br />
those countries adopt modern technologies c<strong>on</strong>currently with their low level of<br />
relative efficiency then their rates of growth would stay at a subpar level of their<br />
potential.<br />
One major source of efficiency generati<strong>on</strong> for a country, according to Parente <strong>and</strong><br />
Prescott (2004), is bel<strong>on</strong>ging to a ‘free trade club’ that improves efficiency through<br />
greater industrial competiti<strong>on</strong>. We calculate the required time period until the<br />
nati<strong>on</strong>s reach their fr<strong>on</strong>tier when <strong>on</strong>ly the catch‐up term <strong>and</strong> inefficiency are allowed<br />
to vary across regi<strong>on</strong>s <strong>and</strong> countries. Two fr<strong>on</strong>tiers are c<strong>on</strong>sidered: nati<strong>on</strong>s’<br />
inefficiency fr<strong>on</strong>tier <strong>and</strong> the leader nati<strong>on</strong>’s fr<strong>on</strong>tier. The latter requires that the<br />
inefficiency levels fade away in time which we assume occurs at the rate of . The<br />
European countries, with the excepti<strong>on</strong> of Turkey, all seem to have reached their<br />
inefficiency reduced fr<strong>on</strong>tier. The same is true for most of the East Asian countries.<br />
Thus, these nati<strong>on</strong>s will not catch‐up with the US without higher accumulati<strong>on</strong> rates<br />
or improved efficiency. For Latin America, most countries are still catching‐up with<br />
their inefficiency fr<strong>on</strong>tier, so that if accumulati<strong>on</strong> rates were the same catch‐up<br />
would still take place through diffusi<strong>on</strong> of disembodied technology. Of course, if<br />
inefficiency levels remain then a follower could never completely catch‐up with the<br />
leader by taking advantage of the technology gap al<strong>on</strong>e. As an illustrative example,<br />
for the required time to catch‐up with the leader if inefficiency levels were<br />
improving at the rate much of Europe <strong>and</strong> Latin America could then approach the<br />
87<br />
335
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
fr<strong>on</strong>tier faster than East Asia <strong>on</strong> account of East Asia’s lower rate of technology<br />
adopti<strong>on</strong>. This begs the questi<strong>on</strong> of what determines these. (in)efficiencies?<br />
It is reas<strong>on</strong>able to expect a tradeoff between a general technology level (GTL) of a<br />
nati<strong>on</strong>’s leading industries <strong>and</strong> its instituti<strong>on</strong>al efficiencies (IE). Thus, using an<br />
aggregate score, (GTL, IE) , say, a country may be in the top rank of GTL but weak<br />
<strong>on</strong> IE which may be surpassed in growth by <strong>on</strong>e which is lower in GTL rank but<br />
str<strong>on</strong>g <strong>on</strong> IE.<br />
CONCLUSION AND FURTHER DISCUSSION<br />
Ec<strong>on</strong>omic growth over the l<strong>on</strong>g‐run can <strong>on</strong>ly be achieved in the course of a real,<br />
sustainable value‐creating process through industrial performance <strong>and</strong> open<br />
markets in which technology <strong>and</strong> innovati<strong>on</strong> are the key facilitators. Nati<strong>on</strong>s with<br />
their industries engage in rival c<strong>on</strong>tests in what we term industrial races within a<br />
given internati<strong>on</strong>al trade regime. This reflects a micro‐ec<strong>on</strong>omic based behavioral<br />
focus <strong>on</strong> ec<strong>on</strong>omic growth (positive or negative). It builds a deeper foundati<strong>on</strong> to<br />
explanati<strong>on</strong>s of ec<strong>on</strong>omic growth than c<strong>on</strong>venti<strong>on</strong>al macro‐ec<strong>on</strong>omic texts . It also<br />
uncovers the true sources of growth as a tool for growth diagnostics (Rodrik, 2007)<br />
allowing to embrace other observati<strong>on</strong>s <strong>on</strong> urban growth <strong>and</strong> n<strong>on</strong> primarily<br />
ec<strong>on</strong>omic factors. In an influential paper in Foreign Affairs entitled ‘Can India<br />
overtake China’ Huang <strong>and</strong> Khanna (2003) first looked at macro‐ec<strong>on</strong>omic factors,<br />
which favor China. They then c<strong>on</strong>sidered micro‐ec<strong>on</strong>omic structures <strong>and</strong> behaviors<br />
such as competent indigenous entrepreneurship, a sound capital market, an<br />
independent legal system, property rights <strong>and</strong> a grass roots approach to<br />
development. The latter all favor India in the l<strong>on</strong>g run, say over the next fifty years.<br />
In a widely covered empirical investigati<strong>on</strong> <strong>on</strong> global growth patterns we c<strong>on</strong>cur<br />
with Easterly <strong>and</strong> Levins’ (2002) finding that it is not factor accumulati<strong>on</strong>, per se, but<br />
total factor productivity that explains cross‐country differences in the level of GDP<br />
growth rates. This total productivity in turn is derived from technology (innovati<strong>on</strong>)<br />
transfer <strong>and</strong> diffusi<strong>on</strong>, its’ supporting instituti<strong>on</strong>al characteristics <strong>and</strong> cultural<br />
dependence. Of course, <strong>on</strong> a deeper level, c<strong>on</strong>siderati<strong>on</strong>s of merely formal<br />
instituti<strong>on</strong>s may not suffice for explanati<strong>on</strong>s but instead forms of ec<strong>on</strong>omic<br />
mechanism design may be called for that effectively deal with (enforce rules <strong>on</strong>)<br />
‘moral hazard’ <strong>and</strong> ‘adverse selecti<strong>on</strong>’ issues (Myers<strong>on</strong>, 2006). Ec<strong>on</strong>omic growth in a<br />
decentralized system would be fully supported by a a Hayek‐Hurwicz mechanism<br />
design.<br />
Observati<strong>on</strong>s <strong>on</strong> firm‐led racing patterns emerging in oligopolistic market structures<br />
of particular high tech industries, <strong>and</strong> the clustering of racing <strong>on</strong> an industry level<br />
are putting companies in different geo‐ec<strong>on</strong>omic z<strong>on</strong>es against each other, becoming<br />
dominant in strategic product/process technologies. Here racing patterns am<strong>on</strong>g<br />
industries in a relatively free trade envir<strong>on</strong>ment could lead to competitive<br />
88 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
336
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
advantages <strong>and</strong> more wealth creating <strong>and</strong> accumulating dominance in key product /<br />
process technologies in <strong>on</strong>e regi<strong>on</strong> at the expense of others. The questi<strong>on</strong> is whether<br />
individual c<strong>on</strong>tests <strong>on</strong> a firm level induce similar effects <strong>on</strong> an industry level <strong>and</strong> if<br />
so, what c<strong>on</strong>trolling effects may be rendered by regi<strong>on</strong>al or multilateral policies <strong>on</strong><br />
regulatory, trade <strong>and</strong> investment matters? The point is that racing behaviour in<br />
leading high technology industries by generating fr<strong>on</strong>tier positi<strong>on</strong>s create cluster<br />
<strong>and</strong> network externalities pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong><br />
creating competitive advantages elsewhere, as supported by the ‘increasing returns’<br />
debate. In this sense we can speak of positive externalities endogenizing growth of<br />
these ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage.<br />
We are about to show in the upcoming chapters how technological racing, rivalry<br />
<strong>and</strong> competiti<strong>on</strong> instigates a process of innovati<strong>on</strong>, industrial <strong>and</strong> market evoluti<strong>on</strong><br />
<strong>and</strong> how it extends to larger entities than firms <strong>and</strong> industries to regi<strong>on</strong>s <strong>and</strong><br />
nati<strong>on</strong>al ec<strong>on</strong>omies or ec<strong>on</strong>omy networks. It will show what drives ec<strong>on</strong>omic<br />
growth <strong>and</strong> globalizati<strong>on</strong>, which industries are most significantly affected <strong>and</strong> how<br />
technological racing results in value generati<strong>on</strong> in increasing returns <strong>and</strong> network<br />
industries. Furthermore, we c<strong>on</strong>sider how the emergence of selective managerial<br />
strategies is most likely to carry success in the pursuit of corporate <strong>and</strong> industrial<br />
policies.<br />
Welfare enhancing technology racing as a c<strong>on</strong>stituent element of the capitalist<br />
process reinforced by globalizati<strong>on</strong> provides social benefits far exceeding the costs.<br />
Even more important, any alternative path, other than the competitive, would likely<br />
be inferior given the costs in that it would generate a less valued <strong>and</strong> less welfare<br />
producing technology portfolio. That is, even if the competitive process is wasteful,<br />
(for example, in parallel or correlated technology development) its unique high<br />
value innovati<strong>on</strong> outcome far exceeds the benefits of any alternative path. There is<br />
historical, observati<strong>on</strong>al <strong>and</strong> analytical evidence given in <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> <strong>and</strong> Goosen<br />
(2012).<br />
On a nati<strong>on</strong>al scale simple catch‐up hypotheses have put emphasis <strong>on</strong> the great<br />
potential of adopting unexploited technology in the early stage <strong>and</strong> the increase of<br />
self‐limiting power in the later stage. However, an actual growth path of<br />
technological trajectory of a specific ec<strong>on</strong>omy may overwhelmingly be c<strong>on</strong>strained<br />
by social capability. The capability also endogenously changes as states of the<br />
ec<strong>on</strong>omy <strong>and</strong> technology evolve. The success of ec<strong>on</strong>omic growth due to diffusi<strong>on</strong> of<br />
advanced technology or the possibility of leapfrogging is mainly attributable to how<br />
the social capability evolves (i.e., which effects become more influential: growing<br />
resp<strong>on</strong>siveness to competiti<strong>on</strong> or growing obstacles to it <strong>on</strong> account of vested<br />
interests <strong>and</strong> established positi<strong>on</strong>s). Another observati<strong>on</strong> relates to policy inferences<br />
89<br />
337
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
<strong>on</strong> market structure, entrepreneurship, innovati<strong>on</strong> activity, industrial policy <strong>and</strong><br />
regulatory frameworks in promoting <strong>and</strong> hindering industry fr<strong>on</strong>tier races in a<br />
global industrial c<strong>on</strong>text. Does lagging behind <strong>on</strong>e’s closest technological rivals<br />
cause an industry to increase its innovative effort?<br />
On an industry level, am<strong>on</strong>g the key issues to be addressed is the apparent inability<br />
of technology oriented corporati<strong>on</strong>s to maintain leadership in fields that they<br />
pi<strong>on</strong>eered. There is a presumpti<strong>on</strong> that firms fail to remain competitive because of<br />
agency problems or other suboptimal managerial behaviour within these<br />
organizati<strong>on</strong>s. An alternative explanati<strong>on</strong> is that technologically trailing firms, in<br />
symmetric competitive situati<strong>on</strong>s, will devote greater effort to innovati<strong>on</strong>, so that a<br />
failure of technological leaders to maintain their positi<strong>on</strong> is an appropriate resp<strong>on</strong>se<br />
to the competitive envir<strong>on</strong>ment. In asymmetric situati<strong>on</strong>s, with entrants challenging<br />
incumbents, research does dem<strong>on</strong>strate that start‐up firms show a str<strong>on</strong>ger<br />
endeavour to close up to or leapfrog the competitors. Such issues highlight the<br />
dynamics of the race within the given market structure in any of the areas<br />
c<strong>on</strong>cerned.<br />
Catch‐up processes are taking place between leaders <strong>and</strong> followers within a group of<br />
industrialized countries in pursuit of higher levels of productivity <strong>and</strong> ec<strong>on</strong>omic<br />
growth. Supposing that the level of labour productivity were governed entirely by<br />
the level of technology embodied in capital stock, <strong>on</strong>e may c<strong>on</strong>sider that the<br />
differentials in productivities am<strong>on</strong>g countries are caused by the ‘technological age’<br />
of the stock used by a country relative to its ‘chr<strong>on</strong>ological age’. The technological<br />
age of capital is a period of technology at the time of investment plus years elapsing<br />
from that time. Since a leading country may be supposed to be furnished with the<br />
capital stock embodying, in each vintage, technology which was ‘at the very fr<strong>on</strong>tier’<br />
at the time of investment, ‘the technological age of the stock is, so to speak, the same<br />
as its chr<strong>on</strong>ological age’. While a leader is restricted in increasing its productivity by<br />
the advance of new technology, trailing countries ‘have the potential to make a<br />
larger leap’ as they are provided with the privilege of exploiting the backlog in<br />
additi<strong>on</strong> of the newly developed technology. Hence, followers being behind with a<br />
larger gap in technology will have a str<strong>on</strong>ger potential for growth in productivity.<br />
The potential, however, will be reduced as the catch‐up process goes <strong>on</strong> because the<br />
unexploited stock of technology becomes smaller <strong>and</strong> smaller. However, as new<br />
technologies arise <strong>and</strong> are rapidly adopted in a Schumpeterian process of ‘creative<br />
destructi<strong>on</strong>’, their network effects induce rapid accelerating <strong>and</strong> cumulative growth<br />
potentials being catalyzed through industry racing.<br />
In the absence of such a process, we can explain the tendency to c<strong>on</strong>vergence of<br />
productivity levels of follower countries. Historically, it fails to answer alleged<br />
puzzles of why a country, such as the United States, has preserved the st<strong>and</strong>ing of<br />
the technological leader for a l<strong>on</strong>g time since taking over leadership from Britain in<br />
90 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
338
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
around the end of the nineteenth century <strong>and</strong> why the shifts have taken place in the<br />
ranks of follower countries in their relative levels of productivity (i.e. technological<br />
gaps between them <strong>and</strong> the leader). Abramovitz (1986) poses some extensi<strong>on</strong>s <strong>and</strong><br />
qualificati<strong>on</strong>s <strong>on</strong> this simple catch‐up hypothesis in the attempt to explain these<br />
facts. Am<strong>on</strong>g other factors than technological backwardness, he lays stress <strong>on</strong> a<br />
country’s ‘social capability’ (i.e. years of educati<strong>on</strong> as a proxy of technical<br />
competence <strong>and</strong> its political, commercial, industrial, <strong>and</strong> financial instituti<strong>on</strong>s). The<br />
social capability of a country may become str<strong>on</strong>ger or weaker as technological gaps<br />
close <strong>and</strong> thus, he states, the actual catch‐up process does not lend itself to simple<br />
formulati<strong>on</strong>. This view has a comm<strong>on</strong> underst<strong>and</strong>ing to what another ec<strong>on</strong>omist,<br />
Ols<strong>on</strong> (1996), expresses to be public policies <strong>and</strong> instituti<strong>on</strong>s as his explanati<strong>on</strong> of the<br />
great differences in per capita income across countries, stating that any poorer<br />
countries that adopt relatively good ec<strong>on</strong>omic policies <strong>and</strong> instituti<strong>on</strong>s enjoy rapid<br />
catch‐up growth. The suggesti<strong>on</strong> should be taken seriously when we wish to<br />
underst<strong>and</strong> the technological catching‐up to American leadership by Japan, in<br />
particular, during the post‐war period <strong>and</strong> explore the possibility of a shift in<br />
st<strong>and</strong>ing between these two countries. This c<strong>on</strong>siderati<strong>on</strong> will directly bear <strong>on</strong> the<br />
future trend of the state of the art, which exerts a crucial influence <strong>on</strong> the<br />
development of the world ec<strong>on</strong>omy.<br />
These explanati<strong>on</strong>s notwithst<strong>and</strong>ing, we venture as a major factor for divergent<br />
growth processes the level of intensity of the racing process within the most<br />
prevalent value‐added industries with cross secti<strong>on</strong>al spillovers. These are the<br />
communicati<strong>on</strong>s <strong>and</strong> informati<strong>on</strong> industries, which have been shaped <strong>and</strong> led by<br />
leading American firms <strong>and</strong> where the rewards benefited their industries <strong>and</strong><br />
country. Though European <strong>and</strong> Japanese companies were part of the race they were<br />
left behind in core markets reaping lesser benefits. The IT investment relative to<br />
GDP, for example, used to be <strong>on</strong>ly less than half in countries such as Japan,<br />
Germany <strong>and</strong> France compared to the US. This does not bode well for a rapid catchup<br />
in those countries. Steering or guiding the process of racing through the pursuit<br />
of industrial policies aiming to increase competitive advantage of respective<br />
industries, as having been practised in Japan, would stimulate catch‐up races but<br />
appears to be less effective in promoting fr<strong>on</strong>tier racing. Another profound reas<strong>on</strong><br />
lies in the phenomen<strong>on</strong> of network externalities affecting IT industries. That is,<br />
racing ahead of rivals in respective industries may create external ec<strong>on</strong>omies to the<br />
effect that such ec<strong>on</strong>omies within dominant industries tend to improve their<br />
internati<strong>on</strong>al market positi<strong>on</strong> <strong>and</strong> therefore pull ahead in competitiveness vis‐a‐vis<br />
their (trading) partners.<br />
The point is that racing behaviour in leading high growth network industries by<br />
generating fr<strong>on</strong>tier positi<strong>on</strong>s create critical cluster <strong>and</strong> network externalities<br />
91<br />
339
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
pipelining through other sectors of the ec<strong>on</strong>omy <strong>and</strong> creating competitive<br />
advantages elsewhere, as supported by the increasing returns debate (Arthur, 1996).<br />
In this sense we can speak of positive externalities endogenizing growth of these<br />
ec<strong>on</strong>omies <strong>and</strong> c<strong>on</strong>tributing to competitive advantage.<br />
All these characteristics lay the foundati<strong>on</strong>s of the ‘Network Ec<strong>on</strong>omy’. The latter is<br />
formed through an ever emerging <strong>and</strong> interacting set of increasing returns<br />
industries, it is about high‐intensity, technology driven racing, dynamic<br />
entrepreneurship, focussed risk‐taking through (free) venture capital markets<br />
endogenized by societal <strong>and</strong> instituti<strong>on</strong>al support.<br />
Racing behaviour <strong>on</strong> technological positi<strong>on</strong>s am<strong>on</strong>g firms in high technology<br />
industries, as exemplified by the globally operating telecommunicati<strong>on</strong>s, <strong>and</strong><br />
computer industries, produce spillover benefits in terms of increasing returns <strong>and</strong><br />
widespread productivity gains. Due to relentless competiti<strong>on</strong> am<strong>on</strong>g technological<br />
leaders the network effects lead to significant advantages in the value added to this<br />
industry, c<strong>on</strong>tributing to faster growth of GDP, <strong>and</strong> through a flexible labour<br />
market, also to employment growth. This c<strong>on</strong>stitutes a new paradigm in ec<strong>on</strong>omic<br />
thinking through network ec<strong>on</strong>omies <strong>and</strong> is a major gauge to compare the wealth<br />
creating power of the US ec<strong>on</strong>omy against the European <strong>and</strong> advanced Asian<br />
ec<strong>on</strong>omies.<br />
REFERENCES<br />
Abramovitz, M. (1986). Catching up, forging ahead, <strong>and</strong> falling behind, Journal of<br />
Ec<strong>on</strong>omic History, 66, 385‐406.<br />
Abramovitz, M. & David, P. (1995). C<strong>on</strong>vergence <strong>and</strong> Deferred Catchup’, Chapter 1<br />
in Growth <strong>and</strong> <strong>Development</strong>: The Ec<strong>on</strong>omics of the 21 st Century, Ralph L<strong>and</strong>au,<br />
Timothy Taylor, <strong>and</strong> Gavin Wright, eds., Stanford CA: Stanford University Press.<br />
Albach, H. (1994). Informati<strong>on</strong>, Zeit und Wettbewerb (Informati<strong>on</strong>, Time <strong>and</strong><br />
Competiti<strong>on</strong>), J. H. v<strong>on</strong> Thuenen Lecture, in Jahrestagung des Vereins fuer<br />
Socialpolitik, Muenster 1993, Berlin: Duncker und Humblot, 113‐154.<br />
Albach, H. (1997). Global competiti<strong>on</strong> am<strong>on</strong>g the few, The Ehrenrooth Lectures,<br />
Swedish School of Ec<strong>on</strong>omics <strong>and</strong> Business Administrati<strong>on</strong>, Res. Reports 40,<br />
Helsingfors.<br />
Amsden, A. H. (1989). Asia’s Next Giant, New York: Oxford University Press.<br />
Arthur,W. B. (1996). Increasing returns <strong>and</strong> the new world of business, Harvard<br />
Business Review, July–August, 100‐109.<br />
Arthur, W.B. (2009). The Nature of <strong>Technology</strong>, New York: Free Press.<br />
92 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
340
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Brezis, E., Krugman, P. <strong>and</strong> Tsidd<strong>on</strong>, D. (1991). Leapfrogging: a theory of cycles in<br />
nati<strong>on</strong>al technological leadership, Nati<strong>on</strong>al Bureau of Ec<strong>on</strong>omic Research (NBER),<br />
Working Paper # 3886.<br />
Burkett, P. & Hart‐L<strong>and</strong>sberg, M. (2003). A Critique of “Catch‐Up“ Theories of<br />
<strong>Development</strong>, Journal of C<strong>on</strong>temporary Asia 33(2), 147‐171.<br />
C<strong>on</strong>itzer, V. & S<strong>and</strong>holm, T. (2002). Complexity of Mechanism Design, Computer<br />
Science Dept., Carnegie Mell<strong>on</strong> Univ., Pittburg,Pa., Disc.Paper, May 28, 1‐8.<br />
Easterly, W. & Levine, R. (2002). It’s not Factor Accumulati<strong>on</strong>: Stylized Facts <strong>and</strong><br />
Growth Models, Central Bank of Chile , Working Papers No. 164.<br />
Easterly, W. (2013). The Tyranny of Experts, Ec<strong>on</strong>omists, Dictators, <strong>and</strong> the<br />
Forgotten Rights of the Poor, New York: Basic Books.<br />
Fagerberg, J. & Godinho, M. M. (2004). Innovati<strong>on</strong> <strong>and</strong> Catching‐Up’, Chap. 20 <strong>on</strong><br />
H<strong>and</strong>book of Innovati<strong>on</strong>, Oxford: Blackwell.<br />
Feigenbaum, J., Schapiro, M. & Shenker, S. (2007). Distributed Algorithmic<br />
Mechanism Design, In: Nisan et al, Algorithmic Game Theory, Cambridge:<br />
Cambridge Univ. Press, 363‐384.<br />
Gaidar, Y. (2004). Modern Ec<strong>on</strong>omic Growth in Russia <strong>and</strong> Catch‐Up <strong>Development</strong>,<br />
Herald of Europe, 1‐11.<br />
Gaidar, Y. (2012), Russia, A L<strong>on</strong>g View. Cambridge, MA: MIT Press.<br />
Gerschenkr<strong>on</strong>, A. (1962). Ec<strong>on</strong>omic Backwardness in Historical Perspective,<br />
Cambridge, MA: Belknap Press.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2006a). Global Technological Races, Japan <strong>and</strong> the World Ec<strong>on</strong>omy<br />
18, 181‐193.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2006b). Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>:<br />
Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. & Goosen, M. (2012), Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch‐Up,<br />
New York: Nova Science.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. W. (2013). Quality of Service <strong>on</strong> Queueing Networks for the Internet,<br />
iBusiness 5, 1‐12.<br />
Grossman, G. & Shapiro, C. (1987). Dynamic R&D Competiti<strong>on</strong>, Ec<strong>on</strong>omic Journal,<br />
97, 372‐387.<br />
Grove, A. S. (1999). Only the Paranoid Survive, New York: Doubleday.<br />
Harris, C. & Vickers, J. (1987). Racing with uncertainty, Review of Ec<strong>on</strong>omic Studies,<br />
54, 305‐321.<br />
93<br />
341
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g><br />
Catch‐up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic <strong>Development</strong><br />
Hayek, F. A. (1945). The Use of Knowledge in Society, American Ec<strong>on</strong>omic Review<br />
35, 519‐530.<br />
Huang, Y. & Khanna, T. (2003). Can India overtake China?, Foreign Policy July‐Aug.<br />
Hurwicz & Reiter, S. (2006). Designing Ec<strong>on</strong>omic Mechanisms, Cambridge Univ.<br />
Press, Cambridge<br />
Juma, C. & Clark, N. (2002). <strong>Technology</strong> Catch‐Up: Opportunities <strong>and</strong> Challenges for<br />
Developing Countries, Supra Occ. Papers, Research Centre for the Social Sciences,<br />
Univ. of Edinburgh, UK.<br />
Kessler, E. H. & Bierly, P. E. (2002). Is faster really better? An empirical test of the<br />
implicati<strong>on</strong>s of innovati<strong>on</strong> speed, IEEE Transacti<strong>on</strong>s <strong>on</strong> Engineering Management,<br />
49, 2‐12.<br />
Kamien, M. I. & Schwarz, N. L. (1982). Market Structure <strong>and</strong> Innovati<strong>on</strong>, Cambridge:<br />
Cambridge University Press.<br />
Krugman, P. (1991). Increasing Returns <strong>and</strong> Ec<strong>on</strong>omic Geography, Journal of<br />
Political Ec<strong>on</strong>omy, 99(3), 483‐499.<br />
Krugman, P. (1994). The Myth of Asia’s Miracle, Foreign Affairs 73 ,63‐78.<br />
Lee, K. (2013). Schumpeterian Analysis of<br />
Cambridge University Press.<br />
Ec<strong>on</strong>omic Catchup. Cambridge:<br />
Lerner, J. (1997). An empirical explorati<strong>on</strong> of a technology race, R<strong>and</strong> Journal of<br />
Ec<strong>on</strong>omics, 28(2), 228‐224.<br />
Lin, J. Y. (2013). Against the C<strong>on</strong>sensus, Cambridge: Cambridge University Press.<br />
Myers<strong>on</strong>, R. B. (2008). Perspectives of Mechanism Design in Ec<strong>on</strong>omic Theory,<br />
American Ec<strong>on</strong>omic Review, 98(3), 586‐603.<br />
Nisan, N. & R<strong>on</strong>en, A. (2001). Algorithmic Mechanism Design, Games <strong>and</strong> Ec<strong>on</strong>omic<br />
Behavior 35, 166‐196.<br />
Nisan, N., Roughgarden, T., Tardos, E. & Vazirani, V. V. eds. (2007). Algorithmic<br />
Game Theory, Cambridge: Cambridge University Press.<br />
Olsen, M. (1996). Big Bills left <strong>on</strong> the Sidewalk: Why some nati<strong>on</strong>s are rich <strong>and</strong> others<br />
poor, Journal of Ec<strong>on</strong>omic Perspectives 10, 3‐24.<br />
Owen, R. & Ulph, D. (1994). Racing in two dimensi<strong>on</strong>s, Journal of Evoluti<strong>on</strong>ary<br />
Ec<strong>on</strong>omics, 4, 185‐206.<br />
Rodrik,D. (2007). One Ec<strong>on</strong>omics, Many Recipes: Globalizati<strong>on</strong>, Instituti<strong>on</strong>s <strong>and</strong><br />
Ec<strong>on</strong>omic Growth, Princet<strong>on</strong>: Princet<strong>on</strong> University Press.<br />
94 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL.2, ISSUE 3 – SEPTEMBER , 2014, PP. 67‐95<br />
342
3.3 CATCH-UP AND CONVERGENCE<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Sachs, J. D. & Warner, A. M. (1996). Achieving Rapid Growth in the Transiti<strong>on</strong><br />
Ec<strong>on</strong>omies of Central Europe, Harvard Institute for Internati<strong>on</strong>al <strong>Development</strong>,<br />
Discussi<strong>on</strong> Paper 544, July.<br />
Scherer, F. M. (1992). Internati<strong>on</strong>al High‐<strong>Technology</strong> Competiti<strong>on</strong>, Cambridge:<br />
Harvard University Press.<br />
Scherer, F. M. (1999). New Perspectives <strong>on</strong> Ec<strong>on</strong>omic Growth <strong>and</strong> Technological<br />
Innovati<strong>on</strong>, Washingt<strong>on</strong> D. C.: Brookings Instituti<strong>on</strong> Press.<br />
Schumpeter, J. (1947). Capitalism, Socialism <strong>and</strong> Democracy, New York: Harper,<br />
Rev. Editi<strong>on</strong>.<br />
Wan, H. Y. (2004). Ec<strong>on</strong>omic <strong>Development</strong> in a Globalized Envir<strong>on</strong>ment, East Asian<br />
Evidence, Dordrecht, Kluwer.<br />
95<br />
343
HANS W. GOTTINGER<br />
3.3 CATCH-UP AND CONVERGENCE<br />
344
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
ESCAPING FROM DEVELOPMENT<br />
TRAPS: INDUSTRIALIZATION AND<br />
RACING FROM THE BOTTOM<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> 1* , Celia Umali 2<br />
1<br />
STRATEC Munich, Germany<br />
2<br />
Faculty of Ec<strong>on</strong>omics, Nagasaki University, Nagasaki, Japan<br />
*umari@nagasaki‐u.ac.jp<br />
Abstract<br />
For development ec<strong>on</strong>omies, escaping costly development traps an industrializati<strong>on</strong> policy is likely to<br />
be gradual rather than take a “big push” form <strong>and</strong> becoming more balanced over time. Under certain<br />
c<strong>on</strong>diti<strong>on</strong>s the optimal industrializati<strong>on</strong> policy should be more unbalanced the weaker are the<br />
sectorial linkages, the str<strong>on</strong>ger are increasing returns, entrepreneurial resources, <strong>and</strong> the smaller are<br />
the domestic market size <strong>and</strong> the lesser the degree of dynamic competiti<strong>on</strong>. We show how to make<br />
tradeoffs at different levels of development <strong>and</strong> from the perspective of the industrializati<strong>on</strong> debate in<br />
a historical c<strong>on</strong>text of modern development policies.<br />
Key words<br />
<strong>Development</strong> traps; Tradeoffs; Industrializati<strong>on</strong>.<br />
INTRODUCTION<br />
The cumulative literature <strong>on</strong> industrializati<strong>on</strong> has formalized the l<strong>on</strong>g‐st<strong>and</strong>ing idea<br />
that development traps are the result of a failure of ec<strong>on</strong>omic organizati<strong>on</strong> rather<br />
than a lack of resources or other technological c<strong>on</strong>straints. The so‐called “big push”<br />
models of industrializati<strong>on</strong> have shown how, in the presence of increasing returns,<br />
there can exist preferable states to advance the ec<strong>on</strong>omic states of countries in<br />
c<strong>on</strong>test with other countries. Such a view not <strong>on</strong>ly provides an explanati<strong>on</strong> for the<br />
co‐existence of industrialized <strong>and</strong> less industrialized ec<strong>on</strong>omies, but also a rati<strong>on</strong>ale<br />
for government interventi<strong>on</strong> to coordinate investment in a “big‐push” toward<br />
industrializati<strong>on</strong>. Moreover, unlike competing theories, these models emphasize the<br />
temporary nature of any policy. Thus, industrializati<strong>on</strong> policy involves facilitating an<br />
adjustment from <strong>on</strong>e equilibrium to another rather than any change in the nature of<br />
the set of equilibria per se.<br />
5<br />
345
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
While recent formalizati<strong>on</strong> makes clear the possible role for the government in<br />
coordinating ec<strong>on</strong>omic activity, little has been said about the form such policy<br />
should take. Is there a c<strong>on</strong>ceptual model to analyze the questi<strong>on</strong>: what precise form<br />
should the “big push” take? It should be part of mechanism design for ec<strong>on</strong>omic<br />
development (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2014). It is argued that while many different<br />
industrializati<strong>on</strong> policies can be successful in generating escapes from development<br />
traps, the form of the policy that minimizes the costs of this transiti<strong>on</strong> depends <strong>on</strong><br />
the characteristics of the ec<strong>on</strong>omic situati<strong>on</strong> at h<strong>and</strong>. Factors such as the strength of<br />
the complementarities, externalities <strong>and</strong> increasing returns, am<strong>on</strong>g others, all play a<br />
role in influencing the nature of a “getting‐ahead” industrializati<strong>on</strong> policy. Such<br />
ideas were already present in the debates in development ec<strong>on</strong>omics in the 1940s<br />
<strong>and</strong> 1950s regarding the form of industrializati<strong>on</strong> policy. The models underlying<br />
these less formal debates inspired the recent more formal research but the policy<br />
elements of these have not been addressed, to date, in any substantive way.<br />
The paper proceeds as follows. We first give a brief history to recall different<br />
development strategies proliferating in the literature, in Secti<strong>on</strong> 2. Then we show<br />
how the increasing returns debate <strong>on</strong> industries impact structural change <strong>and</strong><br />
development paths, in Secti<strong>on</strong> 3. Secti<strong>on</strong> 4 gives the industrializati<strong>on</strong> policydevelopment<br />
c<strong>on</strong>text in an optimizati<strong>on</strong> framework. C<strong>on</strong>clusi<strong>on</strong>s follow in Secti<strong>on</strong> 5.<br />
A BRIEF HISTORY OF MAJOR DEVELOPMENT STRATEGIES<br />
Principal am<strong>on</strong>g the earlier policy debates was that surrounding the efficacy <strong>and</strong><br />
costs involved in the alternative strategies of “balanced” versus “unbalanced<br />
growth.” Rosenstein‐Rodan (1943, 1961) <strong>and</strong> Nurkse (1952, 1953) provided the<br />
rati<strong>on</strong>ale for the noti<strong>on</strong> that the adopti<strong>on</strong> of modern technologies must proceed<br />
across a wide range of industries more or less simultaneously. It was argued that the<br />
neglect of investment in a sector(s) could undermine any industrializati<strong>on</strong> strategy.<br />
Reacting to this policy prescripti<strong>on</strong> was the “unbalanced growth” school led by<br />
Hirschman (1958) <strong>and</strong> Streeten (1956). They saw the balanced strategy as far too<br />
costly. The advantages of multiple development may make interesting reading for<br />
ec<strong>on</strong>omists, but they are gloomy news indeed for underdeveloped countries. The<br />
initial resources for simultaneous developments <strong>on</strong> many fr<strong>on</strong>ts are generally<br />
lacking. By targeting many sectors, it was argued that scarce resources would be<br />
spread too thin‐ so thin, that industrializati<strong>on</strong> would be thwarted. It seemed more<br />
fruitful to target a small number of “leading sectors” (Rostow, 1960). Then those<br />
investments would “….call forth complementary investments in the next period with<br />
a will <strong>and</strong> logic of their own: they block out a part of the road that lies ahead <strong>and</strong><br />
virtually compel certain additi<strong>on</strong>al investment decisi<strong>on</strong>s” (Hirschman, 1958: 42).<br />
Thus, the existence of complementarity between investments (in particular those<br />
involving human capital) <strong>and</strong> increasing returns motivated an unbalanced approach<br />
(Easterly, 2002). Curiously, at the same time, “complementarity of industries<br />
6 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
346
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
provides the most important set of arguments in favor of a large‐scale planned<br />
industrializati<strong>on</strong>”(Rosenstein‐Rodan, 1943: 205). Further, <strong>on</strong>e of the first to preview<br />
the c<strong>on</strong>necti<strong>on</strong> between Big Push, Poverty Traps <strong>and</strong> Takeoffs was the essay by W.<br />
Easterly (2005) who integrated historical sources with present day modern<br />
development strategies. Both sides appeared to have agreed that a “big push” was<br />
warranted, but they disagreed as to its compositi<strong>on</strong>. Our purpose here is to use the<br />
guidelines provided by the more recent formalizati<strong>on</strong> of the “big push” theory of<br />
industrializati<strong>on</strong> to clarify the earlier debate of the appropriate degree of focus for<br />
industrializati<strong>on</strong> policy. After all, the more recent literature has stressed the roles of<br />
complementarities <strong>and</strong> increasing returns that both schools saw lying at the heart of<br />
their policy prescripti<strong>on</strong>s.<br />
The seminal article formalizing the “big push” theory of industrializati<strong>on</strong> is that of<br />
Murphy et al, (1989). In their model, firms choose between c<strong>on</strong>stant returns <strong>and</strong> an<br />
increasing returns technology based <strong>on</strong> their expectati<strong>on</strong>s of dem<strong>and</strong>. However,<br />
these choices spill over into aggregate dem<strong>and</strong> creating a strategic interacti<strong>on</strong> am<strong>on</strong>g<br />
sectors in their technology adopti<strong>on</strong> decisi<strong>on</strong>s. Thus, under certain c<strong>on</strong>diti<strong>on</strong>s, there<br />
exist two equilibria: with all firms choosing the c<strong>on</strong>stant returns or all choosing the<br />
increasing returns technology. Clearly, in the latter equilibrium, all households are<br />
better off.<br />
While the Murphy et al, (1989) model shows how increasing returns (<strong>and</strong> a wage<br />
effect) aggregate to strategic complementarity am<strong>on</strong>g sectors, it does not lend itself<br />
readily to the debate c<strong>on</strong>cerning the degree of balance in industrializati<strong>on</strong> policy.<br />
First, the static c<strong>on</strong>tent leaves open the questi<strong>on</strong> of whether the interventi<strong>on</strong> should<br />
take the form of anything more than indicative planning. Sec<strong>on</strong>d, the most<br />
comm<strong>on</strong>ly discussed policy instrument in the industrializati<strong>on</strong> debate is the<br />
subsidizati<strong>on</strong> of investments. However, in the Murphy et al, (1989) example, use of<br />
this instrument biases <strong>on</strong>e toward a more unbalanced policy. To see this, observe that<br />
it is the role of the government to facilitate a move to the industrializing equilibrium.<br />
This means that the government must subsidize a sufficient amount of investment to<br />
make it profitable for all sectors to adopt the modern technology.<br />
Given the binary choice set, there then exists some minimum critical mass of sectors<br />
that must be targeted to achieve a successful transiti<strong>on</strong>. A greater range of successful<br />
industrializati<strong>on</strong> policies might be more plausible, however, if firms had the choice<br />
of a wider variety of technology to choose from (<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, 2006; <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g> &<br />
Goosen, 2012). One might suppose that targeting a large number of sectors to<br />
modernize a little <strong>and</strong> targeting a small number of sectors for more radical<br />
modernizati<strong>on</strong> might both generate a big push. Thus, to c<strong>on</strong>sider the balanced<br />
approach properly, a greater technological choice space is required.<br />
7<br />
347
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
INCREASING RETURNS, STRUCTURAL CHANGE AND<br />
DEVELOPMENT PATHS<br />
What would be the choice variables available to the government provided it would<br />
be able to pick up what is likely to be increasing returns industries in the future?<br />
First, in each period, the government can choose the set of firms that it targets for<br />
structural change. Sec<strong>on</strong>d, for each targeted firm, the government can choose a<br />
target level for ‘increasing returns industry’ modernizati<strong>on</strong> in the period. Al<strong>on</strong>g this<br />
vein, the government could choose to target the same number of firms in each period<br />
but induce those firms to modernize gradually over time. Or in c<strong>on</strong>trast, the<br />
government chooses a single level of modernizati<strong>on</strong> to occur across all firms <strong>and</strong> all<br />
periods. It then targets a mass of firms each period for entry <strong>and</strong> modernizati<strong>on</strong>. This<br />
means that industrializati<strong>on</strong> policy is solely characterized by the critical mass of<br />
sectors targeted, <strong>and</strong> the target level of modernizati<strong>on</strong>. The level of modernizati<strong>on</strong><br />
could be sequentially exp<strong>and</strong>ed by infrastructural upgrading across the board to<br />
benefit all major sectors as suggested by the Chinese ec<strong>on</strong>omist Justin Yifu Lin<br />
(2013).<br />
Given a parameterized development path, the most significant parameter represents<br />
the strength of increasing returns in the technology adopted by industrial sectors,<br />
which generates a rati<strong>on</strong>ale for “big push” interventi<strong>on</strong>. A “big push” can be<br />
activated if the ec<strong>on</strong>omy is stuck in a “development trap” from which an escape<br />
could be made through sufficient coordinati<strong>on</strong> of decisi<strong>on</strong>s by input producers. For a<br />
developing ec<strong>on</strong>omy in its early phase a “poverty trap” is a special case of a<br />
“development trap” defined by Barro <strong>and</strong> Sala‐i‐Martin (1995: 49) as a stable steadystate<br />
with low levels of per capita output <strong>and</strong> capital stock. This is a trap because, if<br />
agents attempt to break out of it, the ec<strong>on</strong>omy has a tendency to return to the lowlevel<br />
steady‐state. Only by a very large change in their behavior, can the ec<strong>on</strong>omy<br />
break out of the poverty trap <strong>and</strong> move to the high‐income steady state. To evaluate<br />
the ec<strong>on</strong>omic characteristics, i.e., the strengths of complementarities <strong>and</strong> increasing<br />
returns, would affect the government’s policy choices <strong>and</strong> industrial policies (Gans,<br />
1994).<br />
Big Push theories of industrializati<strong>on</strong> could lead to ‘development traps’ if<br />
sequential industrializati<strong>on</strong> would add more diminishing returns than increasing<br />
returns industries which could be a result of government’s coordinati<strong>on</strong> failure.<br />
This would point to deficiencies in instituti<strong>on</strong>al quality as outlined by North (1990)<br />
impacting ec<strong>on</strong>omic performance. They could give explanati<strong>on</strong>s for decade l<strong>on</strong>g<br />
lackluster performance of Latin American ec<strong>on</strong>omies (Fukuyama, 2008). When a<br />
development trap is purely the result of coordinati<strong>on</strong> failure, to escape from the trap,<br />
would technically require the government to synchr<strong>on</strong>ize the expectati<strong>on</strong>s of<br />
individual agents (entrepreneurs) with targeting investment in industrializati<strong>on</strong><br />
activities. If a government were to announce that firms should modernize to a<br />
8 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
348
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
certain degree, even if this were believed perfectly by individuals <strong>and</strong> firms, each<br />
firm might still have an incentive to wait before investing. In that case, the optimistic<br />
expectati<strong>on</strong>s by the government would not be realized <strong>and</strong> the policy would be<br />
ineffective. Irreversibility <strong>and</strong> the time lag of producti<strong>on</strong> mean that history rather<br />
than expectati<strong>on</strong>s matter (Krugman, 1991). The previous level of industrializati<strong>on</strong><br />
determines what path the ec<strong>on</strong>omy will take in the future. This is why it is difficult<br />
to characterize the industrializing paths of the ec<strong>on</strong>omy. There is ec<strong>on</strong>ometric<br />
evidence that a c<strong>on</strong>tributing factor toward the emergence of development traps is the<br />
lack of surpassing some threshold of technological integrati<strong>on</strong> in the industrializing<br />
(manufacturing) sector (Ortiz et al, 2009).<br />
INDUSTRIALIZATION POLICIES AND DEVELOPMENT<br />
In the c<strong>on</strong>text of a big push development strategy the government faces a tradeoff<br />
between the number of sectors it targets <strong>and</strong> the degree to which it wishes them to<br />
modernize, that is, it chooses the critical mass of sectors that must be targeted at any<br />
point in time in order to generate an escape from a development trap <strong>and</strong> to achieve<br />
increasing returns. Let’s take a simple case where the industrializati<strong>on</strong> policy takes<br />
the form of a “big bang”, that is interventi<strong>on</strong> occurs for <strong>on</strong>e period <strong>on</strong>ly granting that<br />
the resources exist in that period to allow for such a policy. This means that the<br />
industrializati<strong>on</strong> policy is solely characterized by the critical mass of sectors targeted<br />
s * <strong>and</strong> the target level of modernizati<strong>on</strong> ƒ.<br />
Suppose naturally that individual transiti<strong>on</strong> costs are n<strong>on</strong>‐decreasing in ƒ, the<br />
optimal critical mass in terms of ƒ can be described by the path<br />
s * (ƒ¸) (ƒ + 1) ((1‐)(‐1)/ L)<br />
varieties of the industrial ec<strong>on</strong>omy.<br />
<br />
<br />
sI ) + sI with sI as the basic input<br />
Substituting this into the objective functi<strong>on</strong> with cost c (ƒ,1; sI, ), the ‘big bang’<br />
industrializati<strong>on</strong> policy problem becomes<br />
min (ƒ + 1) ((1‐)(‐1)/ L) sI ) c (ƒ,1; sI, )<br />
where use is made of the symmetry of the cost functi<strong>on</strong>s <strong>and</strong> the fact that s sI<br />
firms are targeted. could represent any given exogenous parameter, i.e. or<br />
L a given parameter linked to L <br />
In designing an optimal industrializati<strong>on</strong> policy it shows that a cost minimizing<br />
policy in the industry transiti<strong>on</strong> entails setting certain development model<br />
(exogeneous) parameters such as labour productivity improvements (), upstream<br />
firms discount future earnings () the fixed size of the labour force (L ), the number<br />
of basic industrial sector varieties (sI) , the product linkages between intermediate<br />
input producers (), <strong>and</strong> the use of the intermediate input composite (), the latter<br />
two showing a certain degree of interacti<strong>on</strong> referring to as the returns to<br />
9<br />
349
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
specializati<strong>on</strong> (Romer, 1986). Discussing these parameters qualitatively in terms of<br />
comparative statics would indicate industrial change. Raising any of these<br />
parameters , , L , <strong>and</strong> sI increases the marginal returns to upstream firms in both<br />
their entry <strong>and</strong> modernizati<strong>on</strong> decisi<strong>on</strong>s.<br />
Raising means that sunk costs are translated into labor improvements more<br />
effectively. Similarly, since the costs of modernizati<strong>on</strong> <strong>and</strong> entry are carried today<br />
<strong>and</strong> most of the returns occur in the future, the more likely they are to undertake<br />
those acti<strong>on</strong>s. A large market, a higher L , also raises the marginal return to entry<br />
<strong>and</strong> modernizati<strong>on</strong>. Finally, more industrial varieties mean that the past level of<br />
industrializati<strong>on</strong> is greater, thereby, reducing the marginal costs of inducing firms to<br />
adopt more modern technologies. Given this, the resp<strong>on</strong>siveness of firms to<br />
inducements by the government is enhanced when any of these parameters is raised.<br />
Therefore, the higher are these parameters, the fewer firms need to be targeted to<br />
facilitate an escape (from a development trap). Of these parameters has probably<br />
received the most discussi<strong>on</strong>. In many ways, this parameter represents the strength<br />
of increasing returns in the technology adopted by upstream producers. This is<br />
because higher levels of imply that, when they choose to modernize, upstream<br />
firms will choose technologies involving greater sunk (or fixed) costs. Therefore,<br />
while <strong>on</strong>e requires some degree of increasing returns or ec<strong>on</strong>omies of scale in<br />
producti<strong>on</strong> to generate a rati<strong>on</strong>ale for a “big bang” interventi<strong>on</strong>, the str<strong>on</strong>ger are<br />
those increasing returns to support a more unbalanced industrializati<strong>on</strong> policy. This<br />
relates back to arguments made <strong>on</strong> balanced vs. unbalanced growth. Of the three<br />
other parameters, <strong>on</strong>ly the discount rate seems to have been given a potential role<br />
in the past debate <strong>on</strong> industrializati<strong>on</strong> policy. Matsuyama (1992) interprets the<br />
discount rate as measure of effectiveness of entrepreneurship in coordinating<br />
investment, with a low discount rate indicating existence of greater entrepreneurial<br />
resources. If so, then the above result seems to imply that with a relative scarcity of<br />
entrepreneurial talent a more balanced approach should be followed.<br />
The comparative statics results for <strong>and</strong> require more restricti<strong>on</strong>s because each of<br />
these has two effects. On the <strong>on</strong>e h<strong>and</strong>, lowering <strong>and</strong> increasing raises the<br />
strength of strategic complementarities am<strong>on</strong>g upstream sectors. This tends to favor<br />
a more balanced growth approach. On the other h<strong>and</strong>, <strong>and</strong> each affect the<br />
marginal returns to entry <strong>and</strong> modernizati<strong>on</strong> of firms. The sec<strong>on</strong>d effect reinforces<br />
the first <strong>and</strong> leads to more balanced strategy that is, lowering <strong>and</strong> lifting <br />
increase the marginal returns to entry <strong>and</strong> modernizati<strong>on</strong>. A lower also implies<br />
str<strong>on</strong>ger technical complementarities. This effect is sometimes referred to as the<br />
returns to specializati<strong>on</strong> (Romer, 1987). The c<strong>on</strong>sequence is that a lower raises the<br />
marginal returns to employing greater variety of inputs in producti<strong>on</strong>. The higher is<br />
the weaker linkages am<strong>on</strong>g intermediate input sectors. C<strong>on</strong>versely, str<strong>on</strong>ger<br />
linkages between sectors raise the marginal return to targeting an additi<strong>on</strong>al sector<br />
for change supporting the arguments of the balanced growth strategy.<br />
10 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
350
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Looking at , it is a measure of the appropriability of the returns from supply as an<br />
additi<strong>on</strong>al intermediate input. As Romer (1994) discusses, the larger is , the greater<br />
is the surplus gained by intermediate input producers from the employment of their<br />
product in final goods producti<strong>on</strong>. Therefore, producers of inputs targeted in an<br />
industrializati<strong>on</strong> policy are more likely to react positively (in terms of adopting<br />
better technology) when the appropriable returns from the introducti<strong>on</strong> of their<br />
variety is larger. This effect would tend to favor a more unbalanced approach as <br />
increases.<br />
Summarizing, we have outlined the role of several parameters in influencing the<br />
kind <strong>and</strong> degree of balance in industrializati<strong>on</strong> policy. Factors addressed in the<br />
earlier literature such as strength of linkages, increasing returns <strong>and</strong> entrepreneurial<br />
resources all influence the compositi<strong>on</strong> of the ‘big push’. By c<strong>on</strong>sidering a ‘big bang’<br />
policy, some results are possible. For instance, str<strong>on</strong>g increasing returns in<br />
c<strong>on</strong>juncti<strong>on</strong> with weak sector linkages tend to favor a more unbalanced approach in<br />
order to minimize costs.<br />
CONCLUSIONS<br />
A major problem of the industrializati<strong>on</strong> debate is the timing of the industrializati<strong>on</strong><br />
policy <strong>and</strong> its degree of focus is complex <strong>and</strong> dependent <strong>on</strong> the characteristics of the<br />
case specific ec<strong>on</strong>omy. A ‘big push’ perspective <strong>on</strong> industrializati<strong>on</strong> does not imply<br />
that transiti<strong>on</strong> can be a simple matter of coordinating expectati<strong>on</strong>s via some kind of<br />
indicative planning. Nor does it mean that policy must be balanced <strong>and</strong> take a ‘big<br />
bang’ form in order to be successful. A wide variety of industrializati<strong>on</strong> policies can<br />
generate a ‘big push’ <strong>and</strong> the choice between them is therefore a matter of costs.<br />
In a dynamic model, however, this wide variety of industrializati<strong>on</strong> policies makes a<br />
characterizati<strong>on</strong> of the optimal policy quite difficult. To take advantage of full<br />
marginal modernizati<strong>on</strong> <strong>and</strong> entry costs, a gradual policy is always optimal.<br />
Moreover, in a policy of gradual entry, the number of sectors targeted in each period<br />
is rising over time. However, pairwise interacti<strong>on</strong>s between choice variables <strong>and</strong><br />
exogenous parameters tend to be qualitatively ambiguous in a dynamic setting. For<br />
instance, str<strong>on</strong>g increasing returns accompanied by weak sector linkages tend to<br />
favor a more unbalanced approach in order to minimize costs. The former effect<br />
favors the arguments of the balanced growth school, while the latter was part of the<br />
intuiti<strong>on</strong> of the unbalanced growth school.<br />
11<br />
351
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
<str<strong>on</strong>g>Hans</str<strong>on</strong>g> W. <str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, Celia Umali<br />
Escaping from <strong>Development</strong> Traps: Industrializati<strong>on</strong> <strong>and</strong> Racing from the Bottom<br />
REFERENCES<br />
Barro, R. A. & Sala‐i‐Martin, X. (1995). Ec<strong>on</strong>omic Growth, Cambridge, Ma.: MIT<br />
Press.<br />
Easterly, W. (2002). The Elusive Quest for Growth, Cambridge, Ma.: MIT Press.<br />
Easterly, W. (2005). Reliving the ‘50s’: the Big Push, Poverty Traps <strong>and</strong> Takeoffs in<br />
Ec<strong>on</strong>omic <strong>Development</strong>, Center for Global <strong>Development</strong>, Washingt<strong>on</strong>, D.C., WP No.<br />
65, Aug. 2005, 1‐37.<br />
Fukuyama, F. ed. (2008). Falling Behind, Explaining the <strong>Development</strong> Gap between<br />
Latin America <strong>and</strong> the United States, Oxford: Oxford University Press.<br />
Gans, J. (1994). Industrializati<strong>on</strong> with a Menu of Technologies: Complementarities,<br />
Market Structure <strong>and</strong> Multiple Equilibria, Ec<strong>on</strong>omics Department, Stanford<br />
University.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. (2006). Innovati<strong>on</strong>, <strong>Technology</strong> <strong>and</strong> Hypercompetiti<strong>on</strong>, L<strong>on</strong>d<strong>on</strong>:<br />
Routledge.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. & Goosen, M. (2012). Strategies of Ec<strong>on</strong>omic Growth <strong>and</strong> Catch‐Up,<br />
New York: Nova Science.<br />
<str<strong>on</strong>g>Gottinger</str<strong>on</strong>g>, H. (2014). Catch‐Up <strong>and</strong> C<strong>on</strong>vergence: Mechanism Design for Ec<strong>on</strong>omic<br />
<strong>Development</strong>, Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business, 2(3), 67‐95.<br />
Hirschman, A. O. (1958). The Strategy of Ec<strong>on</strong>omic <strong>Development</strong>, New Haven: Yale<br />
University Press.<br />
Krugman, P. R. (1991). History versus Expectati<strong>on</strong>s, Quarterly Journal of Ec<strong>on</strong>omics<br />
106(2), 651‐667.<br />
Lin, J. Y. (2013). Against the C<strong>on</strong>sensus, Cambridge: Cambridge University Press.<br />
Matsuyama, K. (1992). The Market Size, Entrepreneurship, <strong>and</strong> the Big Push,<br />
Journal of the Japanese <strong>and</strong> Internati<strong>on</strong>al Ec<strong>on</strong>omies, 6, 347‐364.<br />
Murphy, K. M., Shleifer, A. & Vishny, R. W. (1989). Industrializati<strong>on</strong> <strong>and</strong> the Big<br />
Push, Journal of Political Ec<strong>on</strong>omy, 97, 1003‐1102.<br />
North, D. C. (1990). Instituti<strong>on</strong>s, Instituti<strong>on</strong>al Change <strong>and</strong> Ec<strong>on</strong>omic Performance,<br />
Cambridge: Cambridge University Press.<br />
Nurkse, R. (1952). Some Internati<strong>on</strong>al Aspects of the Problem of Ec<strong>on</strong>omic<br />
<strong>Development</strong>, American Ec<strong>on</strong>omic Review, 42, 571‐583.<br />
Nurkse, R. (1953). Problems of Capital Formati<strong>on</strong> in Underdeveloped Countries,<br />
Oxford: Basil Blackwell.<br />
Ortiz, C. H., Castro, J. A. & Badillo, E. R. (2009). Industrializati<strong>on</strong> of Growth:<br />
Threshold Effects of <strong>Technology</strong> Integrati<strong>on</strong> , Documentos de Trabajo,<br />
12 JOURNAL OF APPLIED ECONOMICS AND BUSINESS, VOL. 3, ISSUE 1 ‐ MARCH , 2015, PP. 5‐13<br />
352
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
HANS W. GOTTINGER<br />
Journal of Applied Ec<strong>on</strong>omics <strong>and</strong> Business<br />
Departamento de Ec<strong>on</strong>omia, A.A. 25360, Cali, Columbia.<br />
Romer, P. M. (1986). Increasing Returns <strong>and</strong> L<strong>on</strong>g‐Run Growth, Journal of Political<br />
Ec<strong>on</strong>omy, 94, 1002‐1037.<br />
Romer, P. M. (1987). Growth Based <strong>on</strong> Increasing Returns Due to Specializati<strong>on</strong>,<br />
American Ec<strong>on</strong>omic Review, 77(2), 56‐62.<br />
Romer, P. M. (1994). New Goods, Old Theory, <strong>and</strong> the Welfare Costs of Trade<br />
Restricti<strong>on</strong>s, Journal of <strong>Development</strong> Ec<strong>on</strong>omics, 43, 5‐38.<br />
Rosenstein‐Rodan, P. N. (1943). Problems of Industrializati<strong>on</strong> of Eastern <strong>and</strong><br />
Southeastern Europe, Ec<strong>on</strong>omic Journal, 53, 202‐211.<br />
Rosenstein‐Rodan, P. N. (1961). Notes <strong>on</strong> the Theory of the “Big Push”, In: Ec<strong>on</strong>omic<br />
<strong>Development</strong> for Latin America, H. S. Ellis <strong>and</strong> H. C. Wallich (Eds), New York: St.<br />
Martin’s.<br />
Rostow, W. (1960). The Stages of Ec<strong>on</strong>omic Growth, Cambridge: Cambridge<br />
University Press.<br />
Streeten, P. (1956). Unbalanced Growth, Oxford Ec<strong>on</strong>omic Papers, 167‐190.<br />
13<br />
353
HANS W. GOTTINGER<br />
3.4 ESCAPING FROM DEVELOPMENT TRAPS<br />
354
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
355
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
356
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
357
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
358
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
359
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
360
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
361
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
362
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
363
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
364
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
365
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
366
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
367
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
368
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
369
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
370
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
371
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
372
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
373
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
374
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
HANS W. GOTTINGER<br />
375
HANS W. GOTTINGER<br />
3.5 SOME APPLICATIONS OF A RESULT IN CONTROL THEORY TO ECONOMIC PLANNING MODELS<br />
376
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
377
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
378
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
379
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
380
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
381
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
382
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
HANS W. GOTTINGER<br />
383
HANS W. GOTTINGER<br />
3.6 OPTIMAL ECONOMIC GROWTH WHEN CO2 CONSTRAINTS ARE CRITICAL<br />
384
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
385
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
386
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
387
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
388
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
389
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
390
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
391
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
392
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
393
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
394
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
395
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
396
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
397
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
398
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
399
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
400
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
401
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
402
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
403
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
404
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
405
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
406
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
407
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
408
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
409
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
410
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
411
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
412
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
HANS W. GOTTINGER<br />
<strong>Technology</strong> Diffusi<strong>on</strong> 119<br />
Mansfield, Edwin, "Technical Change <strong>and</strong> the Rate of Imitati<strong>on</strong>". Ec<strong>on</strong>ometrica, Val. 29,<br />
1961, pp. 741-766.<br />
-, Industrial Research <strong>and</strong> Technological Innovati<strong>on</strong>. New York 1968.<br />
Manski, Charles, Daniel McFadden, Structural Analysis of Discrete Data with Ec<strong>on</strong>ometric<br />
Applicati<strong>on</strong>s. Cambridge, Mass., 1981.<br />
McFadden, Daniel, "C<strong>on</strong>diti<strong>on</strong>al Logit Analysis of Qualitative Choice Behavior". In: Paul<br />
Zarembka (Ed.), Fr<strong>on</strong>tiers in Ec<strong>on</strong>ometrics. New York 1973, pp. 105-142.<br />
Metcalfe, J. Stanley, "Diffusi<strong>on</strong> of Innovati<strong>on</strong> in the Lancashire Textile". Manchester School<br />
of Ec<strong>on</strong>omics <strong>and</strong> Social Studies, Val. 10, 1970, pp. 35-50.<br />
Nabseth, Lars, George F. Ray, The Diffusi<strong>on</strong> of Industrial Processes: An Internati<strong>on</strong>al Study.<br />
Cambridge 1974.<br />
Nati<strong>on</strong>al Science Foundati<strong>on</strong>, <strong>Technology</strong> in Retrospect <strong>and</strong> Critical Events in Science<br />
(TRACES). Washingt<strong>on</strong>, D.C., 1968.<br />
Nels<strong>on</strong>, Richard (Ed.), Government <strong>and</strong> Technical Progress: A Cross lndustry Analysis. New<br />
York 1982.<br />
-, Sidney G. Winter, An Evoluti<strong>on</strong>ary Theory of Ec<strong>on</strong>omic Change. Cambridge, Mass., 1982.<br />
Oster, Shar<strong>on</strong>, ''The Diffusi<strong>on</strong> of Innovati<strong>on</strong> am<strong>on</strong>g Steel Firms". The Bell Journal of Ec<strong>on</strong>omics,<br />
Val. 13, 1982, pp. 45-56.<br />
Peterka, V., Macrodynamics of Technological Change: Market Penetrati<strong>on</strong> by New Technologies.<br />
1.1.A.S.A. RR-77-22. Laxenburg, Austria, November 1977.<br />
Reinganum, Jennifer F., "<strong>Technology</strong> Adopti<strong>on</strong> under Imperfect Informati<strong>on</strong>". The Bell Journal<br />
of Ec<strong>on</strong>omics, Val. 14, 1983, pp. 57-69.<br />
Rosenberg, Nathan, Inside the Black Box: <strong>Technology</strong> <strong>and</strong> Ec<strong>on</strong>omics. New York 1982.<br />
-, Ralph L<strong>and</strong>au (Eds.), The Positive Sum Strategy. Washingt<strong>on</strong>, D.C., 1986.<br />
Scherer, Frederic M., Innovati<strong>on</strong> <strong>and</strong> Growth: Schumpeterian Perspectives. Cambridge,<br />
t\1ass., 1984.<br />
Spinrad, Bernard 1., Market Substituti<strong>on</strong> Models <strong>and</strong> Ec<strong>on</strong>omic Parameter. 1.1.A.S.A.<br />
RR-80-28. Laxenburg, Austria, July 1980.<br />
Telser, Lester G., "A Theory of Innovati<strong>on</strong> <strong>and</strong> Its Effects". The Bell Journal of Ec<strong>on</strong>omics,<br />
Val. 13, 1982, pp. 60-92.<br />
* * *<br />
Zusammenfassung: ök<strong>on</strong>omische Wahl und Verbreitung der Technologie auf den<br />
Märkten für neue Produkte. - Dieser Aufsatz versucht modellhaft zu entwickeln und zu<br />
schätzen, wie sich Innovati<strong>on</strong>en unter den Produzenten im Falle v<strong>on</strong> Spitzentechnologien<br />
ausbreiten. Nach der Einführ'ung einer Innovati<strong>on</strong> soll es annahmegemäß Firmen geben, die<br />
die Innovati<strong>on</strong> übernehmen, und Firmen, die sie imitieren; außer diesen beiden Firmengruppen<br />
gibt es keine <strong>and</strong>eren Marktteilnehmer. Die grundlegende Annahme ist, daß es einen<br />
inneren Zusammenhang zwischen dem Muster der Diffusi<strong>on</strong> und der Dynamik des Marktes<br />
gibt. Es wird unterstellt, daß eine "repräsentantive" Firma in einem gegebenen Markt vor die<br />
Wahl gestellt ist, ob sie sich auf dem Markt betätigen will oder nicht bzw. ob sie aus ihm<br />
ausscheiden will oder nicht, und daß die Wahlmöglichkeiten v<strong>on</strong> wirtschaftlichen Parametern<br />
413
HANS W. GOTTINGER<br />
3.7 ECONOMIC CHOICE AND TECHNOLOGY DIFFUSION IN NEW PRODUCT MARKETS<br />
414