ECONOMY

714 the role of the state in development
G’s incentives to adhere to the equilibrium choice of strategies derive from the
revenues he can secure from taxation. To induce G to refrain from predation, the
level of revenues he earns from the provision of protection needs to be high enough
that G finds it optimal, given the private agents’ choice of strategies, to refrain from
confiscating the agents’ wealth if they pay taxes. But the revenues must also be
sufficiently low that private agents prefer to purchase the services of G rather than
to incur the costs of providing their own security. The level of revenues must satisfy
both G’s and the citizens’ participation constraints, where the latter is defined by a
capacity to revert to the private provision of security and to the taking up of arms.
G’s willingness to adhere to the equilibrium path also depends upon his payoffs in
the punishment phase. The magnitude of these payoffs constitute the shadow of the
future, to use Axelrod’s phrasing (Axelrod 1984). This shadow will be the dimmer and
the punishment phase therefore less of a deterrent to predation should G be assured of
a prosperous future even should the state break down. Should, for example, G be able
to retain control over oil fields, gold mines, or other assets, then he need not fear the
loss of taxes, should he abandon his role as a guardian. The shadow will dim as well
should G more heavily discount the future and thus the losses during the punishment
phase. Should his future become less certain or his level of impatience rise, then G
will come to weigh the prospects of immediate gain more highly than future pain.
He will find the immediate rewards (even given the attendant punishment) from
predation more attractive than the steady but moderate future flow of earnings from
the provision of protective services. Both the magnitude of these rewards in the
punishment phase and the degree to which the payoffs are discounted thus determine
the strength of G’s incentives to behave like the government of a state rather than like
awarlord. 3
I have focused on the incentives facing G. But note too the behavior of the private
citizens. If public revenues decline, then, understanding G’s incentives, the citizens
might fear their government’s behavior. They might expect the specialist in violence
to begin to behave as a predator, using his power to extract resources from the private
economy. So too if G’s hold on power becomes less secure: the citizens might then
fear that their government—now facing greater prospects of the loss of power—
would now begin to despoil rather than to protect the private economy. And should
a major new source of wealth arise—a resource boom, say; or discoveries of oil
or mineral deposits—comprehending the incentives that shape the choices of the
specialist in violence, the citizens might anticipate a change in the conduct of their
government. They might anticipate that the government would forswear costly efforts
on their behalf and turn instead to consuming the bounty created by the bonanza. In
anticipation of the transformation of the specialist’s role—his change from protector
to predator—the citizenry would itself then alter its behavior: it would return to the
private provision of security.
The conditions that yield the possibility of the state also imply that the state is
developmental. That they do so provides closure to my argument.
³ For proofs of these arguments, see Bates, Greif, and Singh 2002.