This is a good book about an interesting and important subject: waves of centralizing and decentralizing change over time and across four different country cases. The book analyzes Argentina, Brazil, Chile, and Uruguay from the nineteenth century to the present. More specifically, Kent Eaton examines 30 periods of movement toward greater or lesser centralization—12 in Argentina, 6 in Brazil, 5 in Chile, and 7 in Uruguay.

Eaton focuses mainly on change over time within these four countries and secondarily on differences across the four countries. He analyzes both the causes and consequences of centralization and decentralization. As many countries around the world have embarked on decentralization, this issue has become prominent. He eschews a parsimonious theory and emphasizes case specificity, path dependence (pp. 22–24), and causal complexity, yet he still advances one major generalizable argument. Democratization consistently leads to decentralization when subnational actors are important within their parties (Argentina and Brazil), but not when they are insignificant as party actors (Chile and Uruguay). This argument has multiple virtues: It is a testable generalization; the case studies support the conclusion; and it adds nuance to a deductively appealing simpler argument that democratization should consistently (rather than only when subnational actors are important in their parties) favor decentralization. Eaton carefully considers alternative explanations for understanding decentralization/centralization (pp. 9–19, 52–54).

The author also develops some interesting arguments on how presidentialism affects the bargaining game involved in designing subnational institutions. Presidents usually prefer greater centralization; during periods of competitive politics, legislatures often push for more decentralization.

Eaton persuasively argues that centralizing and decentralizing reforms often have unintended consequences. For example, in the 1990s, democratic leaders in Chile and Uruguay undertook what they intended to be modest decentralizing reforms, but these reforms have had more significant consequences than politicians envisioned. In Argentina, subnational politicians secured major decentralizing concessions in the 1980s, but these concessions fueled the country's economic meltdown in 2001.

The author's reasons for his country case selection are compelling. He argues that Argentina, Brazil, Chile, and Uruguay had the most significant record of relatively free and fair subnational elections of any countries in Latin America prior to the post-1978 wave of democratization. At the same time, these four countries have very different “scores” on the dependent variable, that is, the centralization or decentralization of subnational institutions. For this reason, they escape the methodological problems associated with lack of variance in the dependent variable. During democratic periods, Argentina and Brazil have historically had federal systems with powerful governors and significant resources at the subnational level. In contrast, Chile and Uruguay have unitary systems that were historically quite centralized.

Eaton argues that the actors that have pushed for decentralization—when such a push has occurred—have been different in the four countries. In Argentina and Brazil, subnational politicians, especially governors, have led the drive for decentralization. In Chile and Uruguay, in contrast, national-level politicians have occasionally fostered modest decentralizing initiatives in order to achieve some of their own national-level objectives. Subnational actors have not been major players in decisions related to decentralization.

Eaton presents several persuasive arguments for looking at a broad historical sweep. First, doing so increases the number of cases because each wave of centralization and decentralization is then a distinct case. As is well known, having more cases is useful for explanatory purposes. Second, the past often illuminates key points about the present. For example, the lengthy history of past movements toward recentralization suggests that the decentralizing trends of the past two decades in much of Latin America could be reversed. Examination of the past enables us to understand what is distinctive about the present wave of decentralization. Third, he argues that bargaining over subnational institutions is a path-dependent game; scholars cannot understand the present game without analyzing the past. Finally, including historical cases allows social scientists to assess whether some arguments presented as generalizable are time-bound. At a time when there is some tendency toward a growing gap between political science and history, it is refreshing to read a book by a political scientist that evinces awareness of history.

This work adds much new information and insight to the analysis of decentralization and centralization, and its many virtues overshadow its shortcomings. My two criticisms revolve around differences in social science methodology more than deficiencies in the argument. First, the book would have benefited from a sharper operationalization of the dependent variable. Eaton sometimes implies (for example, p. 48) that the dependent variable is so multifaceted that it is not useful to rely on a few simple operational measures thereof. Yet it is helpful to have some consistent and systematized measures of the dependent variable even when it is complex and multidimensional. Second, perhaps to a fault, he avoids formulaic grand causal arguments that purport to hold up over time and across cases. Politics Beyond the Capital is admirably fine-tuned to differences among cases and to change over time, but it offers less in the way of generalizations and theory building than some works in political science.