This sophisticated book addresses a problem that many political scientists may not have thought much about. A common, if not predominant, view of dictators is that they have the power to impose their will without obstruction by democratic institutions, such as the need to secure elections or the need to get legislative approval for their policies. Such a view also implies that dictators do not face a problem of needing to have agents comply with their policies.

Armando Razo's book takes a different and subtler approach. For Razo, the basic presumption is that dictatorships have a major problem that democracies do not have in their ability to foster growth and prosperity. Specifically, a dictatorial state that is strong enough to preserve law and order, arbitrate disputes over property rights, and enforce private contracts is strong enough to plunder its people. Since everyone in the society is presumed to know this, the dictator's problem is convincing economic agents that he will not use his power to take from them, thus reducing their incentive to be productive and thereby undermining economic growth and prosperity.

Instead of having it easier to produce conditions that are conducive to economic growth than democracies, with their checks and balances, dictatorships have a bigger challenge to bring about growth because of this credibility problem. How can dictators create credible commitments to protect property and investments when there are no institutions that would block predatory behavior? Social Foundations of Limited Dictatorship is an impressive effort to answer this important theoretical question.

Razo's book proficiently employs game theory and social network analysis to develop a case study of the regime of Porfirio Díaz, who was president of Mexico from 1876 to 1911. Díaz seized power in 1876 in a revolt “in the name of democracy against a government with evident dictatorial aspirations” (p. 60). By 1890, Díaz himself had consolidated one of the most enduring dictatorships in Latin America.

For Razo, there are two key differences between dictatorship and democracy. Dictators have more discretion in influencing the policymaking process, but they have more difficulties in enforcement because of the lack of institutional constraints and because of “limited or nonexistent public enforcement” (pp. 8–9). Dictators, then, face a “governability dilemma.” Democracies, according to the author, face a binary choice of offering universal protection and no protection at all. Dictators, in contrast, face many possibilities in between the two choices, protecting some but not all asset holders (p. 27).

Razo's theory owes a lot to the late Mancur Olson's stationary bandit theory of dictatorship. But contrary to Olson, Razo argues that “stationary bandits can credibly commit only to provide private rather than public benefits, so the commitments made by dictators will be qualitatively different from those of democracies” (p. 7). The credible commitments are made with private rather than public policies that are enforced by informal rather than formal institutions. These informal institutions are demonstrated with network analysis based on archival work.

Chapter 2 is the theoretical core of the book, in which assumptions and game trees are presented. Three propositions are developed, one for an isolated firm, one for two firms with one enforcer, and one for two firms and two enforcers. In each case, a unique subgame perfect Nash equilibrium is determined, based on parameters representing values for asset holders, enforcers, and the dictator.

Chapter 3 describes the historical and institutional context. Chapter 4 describes the “private policies” that Díaz used to promote economic investment. For the most part, these could be characterized as the awarding of monopolies, establishment of barriers to entry, and protection against competition, all of which produced rents to be shared by the dictator and enforcers. Chapter 5 describes the role of public officials as long-lived private enforcers in a system that prevented government predation. Chapter 6 describes the networks of private protection, and Chapter 7 concludes.

I do have some misgivings about the book. Firstly, Porfirian Mexico is presented as “a canonical case of growth under authoritarianism” (p. 13). But if the reader does not know this already, there is no direct evidence of economic growth in the Porfiriato relative to periods before or after in Mexico or relative to comparable experience in other countries, nor could I find references to work (such as that of Angus Maddison) where such a case might be documented.

Secondly, the case that dictators can have only policies that provide private rather than public benefits is asserted and illustrated instead of being convincingly made in a sustained argument that directly considers the alternative. The network analysis demonstrates how such informal institutions might work, but the argument that this is the only viable way for a dictator to solve the credibility problem is not convincingly made.

Razo's case is designed to explain how dictatorships might use elaborate networks of rent seeking and rent sharing to achieve what can come more naturally in a democracy. Presumably it can come more naturally in a democracy because a democracy can use public rather than private policies to provide public rather than private goods, and because enforcement is provided by public institutions. But I find it a caricature of democracy to assume that it provides only public goods, and to imply that rent seeking is not an important problem in such governments.

Razo does not address theories of economic growth and assumes that investment directly and unambiguously causes growth. William Easterly, Dani Rodrik, and others would say that it is not that simple. Moreover, if the Porfiriato is a canonical case of growth under dictatorship, then the protection, lack of competition, and entry barriers would make Porfirian Mexico less efficient than it might have been with more competition, less exclusion, and fewer rents. Porfirian Mexico may illustrate how dictatorships can “artificially” approximate what is more natural in a democracy, but it does not make a case for the superiority of authoritarian government for purposes of economic growth.

Of course, making such a case is not the author's intent. Indeed, his book is an explanation of how dictatorships can grow at all in spite of their disadvantages. Our theoretical understanding of the relative strengths and weaknesses of democracy and dictatorship would be enhanced if Razo were to make a direct comparison in which each system got comparable attention. At the same time, he is on to something important in this book, and I hope we can look forward to his applying the talent demonstrated here more broadly.