and the corresponding subgame perfect strategies are:
ρ(µL, ·, ht-1) =0, and:
Summarizing this discussion, we have the following:
Proposition 5.5: Assume θ > µ. Let µ** and s> µ** be given by (5.40) and (5.44). Then, the subgame perfect equilibrium that is best from the viewpoint of the elites, {r, p}, of the game G∞(β) is such that:
(1)
(2)
(3)
The important point that emerges from Proposition 5.5 is that there is now a larger set of parameter values that allows the elites to avoid revolution. In other words, in societies with µ such that µ** ≤ µ < µ*, there will be equilibrium revolutions if we do not allow the elites to make incentive-compatible promises of redistribution in future low-revolution threat periods; however, these revolutions can be avoided once we allow such promises. Moreover, even when µ ≥ µ*, the elites can achieve a better outcome for themselves by smoothing taxes because of the possibility of using incentive-compatible promises.
Nevertheless, it is important to emphasize that the elites still have limited abilities to make credible promises. Only promises of redistribution at the tax rate τL that satisfy Vr (N, µL, [τL, τH]) ≥ (N, µL) are incentive-compatible. This implies that in societies with µ < µ**, the same considerations as in Proposition 5.4 apply and credible redistribution is not enough to convince the citizens to live under nondemocracy, and they will prefer alternative routes. Here, the only option open to them is a revolution. In Chapter 6, we see how the elites can try to convince them not to undertake a revolution by offering a change in political institutions to make future redistribution more credible. Democratization gives the citizens political power, thereby making much higher levels of future redistribution credible.
8. Conclusion
In this chapter, we developed our basic model of nondemocratic politics and introduced the fundamental issue of political commitment that underlies much of our approach. We studied how, in the face of the threat of collective action and revolution, a nondemocratic regime would want to make concessions to avoid being expropriated. Nevertheless, because revolutionary threats are intrinsically transitory, the promise of concessions may not be sufficiently credible. When the revolution threat dissipates, the regime may renege on its promises, as we illustrated with some historical examples. If citizens anticipate that the nondemocratic regime will renege on its promises, the regime may be swept away by a revolution.
We illustrated these ideas first in a static extensive-form game in which we introduced an exogenous probability that a promise by the regime will be kept. Although this model is useful and tractable, the exogenous probability of reneging is too reduced-form. For this reason, we also developed a richer dynamic model in which the regime can make promises for today but cannot make promises for the future. We showed how the qualitative results of the dynamic model are identical to those of the static model.
Nevertheless, the options we have allowed so far are restrictive: for example, might such a regime not have other instruments it could use other than policy concessions, such as income redistribution? The answer to this is yes, and in Chapter 6 we argue that democratization precisely arises as a credible concession by the elites to stop a revolution. By democratizing, the elites allow the citizens to set the tax rate not only today but also in the future, and this makes their concessions credible. However, even there our discussion are not complete. Rather than make any type of concession, the elites might try to use repression to avoid a revolution or having to democratize. Therefore, we also discuss the issue of the interaction among concessions, democratization, and repression in Chapter 6. Chapter 6 also discusses in more detail the conceptual foundations of our approach to democratization - in particular, stressing why institutional change can help solve commitment problems.
PART THREE.
THE CREATION AND CONSOLIDATION OF DEMOCRACY
6
Democratization
1. Introduction
In this chapter, we motivate and develop our basic model of democratization. Individuals have preferences over different political institutions because they anticipate the different actions that political actors will take under these institutions - thus, the resulting different policies and social choices. In this sense, our analysis builds on our modeling of democracy and nondemocracy in the previous two chapters.
2. The Role of Political Institutions
Why do we need to talk about institutions at all? Why not simply say that the elites and the citizens have preferences over different policies, and political conflict between them results in a set of policies favoring one group or the other? We argue that there is more to the conflict between various social groups. Conflict over policies is static-it is about what happens today. Rational actors also care about the future. This is where political institutions - which are durable and, consequently, have the capacity to influence political actions and political equilibria in the future - come in. Therefore, we need to think seriously about political institutions in a dynamic setting; via this process, we can develop a theory of the emergence and, later, consolidation of democracy. Crucial to this is a notion of what political institutions do.
We emphasize that political institutions regulate the allocation of de jure political power. Political power is a measure of how influential a particular group (or individual) is in the political arena when there is conflict over which policy should be implemented. If the elites are the rich and if they are more powerful, we expect lower taxes, lower redistribution, and generally a range of policies favoring the rich rather than the poor. Political power is, therefore, inherent in every discussion of aggregating conflicting preferences. Various models of democracy aggregate those preferences differently and, therefore, as discussed in Chapter 4, they naturally allocate different amounts of political power to different groups. Nevertheless, critical to our approach is the assumption that, typically, the majority of citizens has more political power in a democratic rather than a nondemocratic society.
Forward-looking rational economic actors care not only about economic allocations and therefore policies today but also about the economic allocations and policies in the future. Therefore, political power is valuable and all groups would like to somehow ensure greater political power for themselves in the future. Political institutions can influence the allocation of de jure political power in the future by virtue of being durable. Our approach to institutions is based on the assumption that policies, even though they can sometimes be difficult to reverse, are generally easier to reverse than institutions. Therefore, democratization enables the citizens not only to be more powerful today but also in the future relative to an alternative regime that is nondemocratic. Hence, democratization is a way of transferring political power to the majority of citizens. If the citizens can secure democracy today, they will increase their de jure political power in the future because as long as democracy survives, they will have more say in the determination of economic and social policies.
How do the citizens ensure that society becomes democratic? They can do so only if they have sufficient political power. Clearly, starting from a situation of nondemocracy, the citizens are excluded from voting or, at the very least, their preferences matter only little. So how could they have political power? The answer is that political power is not only vested in the formal rules, it can also take the form of de facto political power. The citizens could have political power in nondemocracy if they pose a credible threat of revolution or significant social unrest that damages the economic and social interests of the elites who control de jure political power. Throughout this book, when we discuss political power, it includes both the power that comes from political institutions and the ability of the citizens to challenge the system or the ability of the elites to undertake a coup - that is, de facto ways of obtaining power. In other words, for our purposes, political power is anything that enables a social group to come close to its preferred policies, de j
ure or de facto.
However, our story is not yet complete. So far, what we have argued can be summarized diagrammatically as follows:
Groups that have political power today can introduce - or force others to introduce - political institutions that favor them. These political institutions persist and regulate the allocation of political power in the future. Therefore, democratization enables the citizens to increase their political power in the future. But, why do the citizens need political institutions to ensure their political power tomorrow? After all, they have political power today.
In our theory, political institutions are particularly useful when de facto political power is transitory in the sense that who has more de facto political power today is generally different from who will have it tomorrow. This transitory nature might result from a variety of economic, social, and political shocks to the system. Indeed, we saw in Chapter 5 that the empirical literature on the collective-action problem emphasizes that even when the problem can be solved, such solutions tend to be transitory. The fact that a group has solved the collective-action problem today does not guarantee that it will manage to solve the problem tomorrow. There may be a recession today, creating political instability, giving an advantage to whichever group wants to use de facto means to influence political outcomes; however, recessions are often transitory; tomorrow there may be a boom.
Now imagine a situation in which the citizens have de facto political power today, but they expect not to have similar political power tomorrow. In this situation, they would demand a set of institutions that will lock in their political power. This is precisely what democratic institutions may do. The de facto power of the citizens that comes from an unusual event, such as a political crisis or the end of a war, becomes institutionalized and translated to future political power by the introduction of relatively free and fair elections in which the votes of all the citizens count, not just the elites.
There is one final step in our argument. Democratization is a move from nondemocracy to democracy and, in nondemocracy, the elites make the decisions. Therefore, democratization happens when the elites decide to “extend the franchise” and include wider segments of society in the decision-making process. This is not only a theoretical statement. As the discussion in the Introduction and Chapter 2 illustrated, almost all major moves toward democracy in nineteenth-century Europe and nineteenth- and twentieth-century America were extensions of the franchise by the existing political system to previously excluded segments of society.
Why would they do so? The answer is that the temporary de facto political power of the citizens in nondemocracy comes from actions they can take that are costly to the system, such as revolution or significant social unrest. The elites would like to prevent this and they are willing to make concessions in order to do so. But, the citizens care not only about allocations and policies today but also about those in the future. Therefore, typically, the elites have to make promises about future as well as current concessions. However, when the revolution threat subsides - the crucial transitory nature of de facto political power! - these promises may be broken. Consequently, the elites would like to make credible commitments. This is where the commitment value of institutions is relevant. Democratization ensures a credible transfer of political power to the majority of citizens, increasing the likelihood that the promises of the elites will be honored. Therefore, democratization occurs when the elites would like to make a credible commitment to future policies and they can do so only by relinquishing (part of their) political power, the de jure part, to the citizens.
Diagrammatically, our theory can be summarized as follows:
3. Preferences over Political Institutions
In this section, we illustrate the basic conflict over political institutions, specifically democracy versus nondemocracy. With this purpose, let us return to the basic two-class model discussed in Chapter 4. Total population is normalized to 1, a fraction 1 - δ > ½ of the agents are poor with income yp, and the remaining fraction δ are rich with income yr > yp. The rich are the elites and the poor are the citizens. Mean income is denoted byand, as before, we use the notation θ to parameterize inequality. The incomes of poor citizens and rich elites are given by (4.7), and the preferred tax rate of a poor citizen satisfies (4.11).
We also use the notation for indirect utility introduced in Chapter 4: V(yi| τ) denotes the utility of an agent with income level yi when policy is given by τ. Now, denne Vp(D) ≡ V(yp| τp) as the indirect utility of a poor agent when the tax rate is equal to τp. Equivalently, in democracy, all the citizens have the same political preferences and they will vote for τp, so the equilibrium tax rate in democracy is τp. Therefore, V(yp | τp) is also the indirect utility of a citizen in democracy, Vp (D) (D is for democracy). Similarly, Vr (D) ≡ V (yr| τp) is the indirect utility of a member of the elite in democracy. In nondemocracy, the most preferred tax rate of a member of the elite, τr = 0, will result; therefore, Vp(N) ≡ V(yp|τr) is the indirect utility of a citizen in nondemocracy (N is for nondemocracy) in which the equilibrium tax rate is τr = 0. Finally, Vr(N) ≡ V(yr | τr) is the indirect utility of an elite agent in nondemocracy.
We have that:
(6.1)
In other words, the citizens obtain higher utility and income in democracy, whereas the elites obtain higher income in nondemocracy. An immediate implication of this observation is that there is conflict over political institutions - that is, over whether the society should be democratic or nondemocratic. In democracy, the citizens get relatively higher benefits; the elites benefit in nondemocracy.
4. Political Power and Institutions
4.1 Institutions versus Policies
What is the difference between institutions and policies? In both political science and other social sciences, there is an implicit understanding that institutions and policies are significantly different objects. For example, few people would think that tax policy is an “institution,” whereas whether there is a constitution or the society is democratic is generally seen as relating to institutions. So, what is the difference?
The Nobel-prize-winning economic historian Douglass North defines institutions as “the rules of the game in a society or, more formally... the humanly devised constraints that shape human interaction” (1990, p. 3). This definition of institutions is useful when we think of the broad set of institutions, encompassing many diverse social and political aspects underlying economic decisions, and the organization of economic and social activity. However, for our purposes, it might also be too broad. For us, the main difference between policies and institutions is their “durability” and the ability of institutions to influence the allocation of political power in the future. Policies are much easier to reverse, whereas institutions are more durable. Moreover, institutions determine how the political preferences of various groups are aggregated into social choices. Therefore, introducing a set of institutions today influences how powerful different social groups will be not only today but also tomorrow.
Their durability and ability to influence the allocation of power in the future make institutions valuable as a commitment device. Recall that the commitment problem in politics, discussed in Chapter 5, arises because the group in power, the elites, make promises for the future but honoring these promises later is not in their interests. They would rather renege and revert to a different course of action or choose different policies. We refer to this as a commitment problem because the group in power cannot credibly promise certain policies. The commitment problem is intimately linked to the fact that political power will be in the hands of a particular group in the future, and they can use this political power to revert to different policies instead of those they promised. This account also suggests that institutions could be useful as a commitment device because they influence the future allocation of political power. Stated simply, if a particular group wants to make a commitment to a course of action, what better way to make this credible than give more power to th
e party that wants to see this course of action implemented? In other words, the commitment problem emerges because there was a “decoupling” between those who had political power and those who benefited from the promised policies. Change the identity of who has political power and promises become credible.
We are not the first to emphasize the commitment value of institutions. Although this theme appears in many writings and is implicit in others (e.g., the literature on structure-induced equilibrium; see Shepsle 1979; Romer and Rosenthal 1978; and Shepsle and Weingast 1984), it is probably most clearly associated with the seminal paper by North and Weingast (1989). They argued that the establishment of the constitutional regime in Britain after the Glorious Revolution of 1688 provided commitment that the Crown would not repudiate its debt, thereby increasing its borrowing capacity. This led to fundamental changes in financial institutions and provided part of the preconditions for the Industrial Revolution.
What does this institutional change correspond to in practice? How is it achieving this commitment? Thinking about these questions clarifies the role of institutions in this specific context and, more generally, their role in our approach to political institutions and democratization. The first important feature is that institutions are durable. After the Glorious Revolution, the ruler could not revert back to the days when he had been able to arbitrarily manipulate debt and tax policy without the agreement of Parliament. The Glorious Revolution introduced regular parliaments (previously they had to be “called” by the king) and gave Parliament control over fiscal matters. Second, these institutions constrain the behavior of the ruler. It is this feature of the institutions that makes them a credible commitment to repay the debt.
Economic Origins of Dictatorship and Democracy Page 26