Barbara Geddes, Joseph Wright, and Erica Frantz’s book How Dictatorships Work is the latest addition to two decades of influential research on authoritarian regimes: how they come to power, how their institutions work, and how they fall. Readers familiar with the research agenda that Geddes launched with her essay “What Do We Know About Democratization After Twenty Years?” (Annual Review of Political Science 2, 1999) might be tempted to approach this book as a synthesis of that agenda. That would be a mistake, for the book does not summarize that body of research, but rather extends and modernizes it. Rather than piecing together various bits of the logic of authoritarianism, as do various articles by these three authors and others, a book length treatment affords Geddes, Wright, and Frantz the ability to weave a narrative that reflects the life course of an authoritarian regime, from its birth and consolidation to its operation and eventual collapse. How Dictatorships Work is clearly written in an engaging style, and the concluding chapter offers policy recommendations that might find a sympathetic ear in Washington, DC, and among the international donor community.

Abstracting away the details contained across the chapters, there are two broad themes that run throughout the book, the first one explicit, the second implicit. The explicit theme is that the conditions under which authoritarian regimes are born determine how they work. The authors focus on what they term the “seizure group,” which is “the small group that literally ousts the incumbent and takes over in order to initiate dictatorship, as well as their organized support base” (p. 3). The nature of the group that seizes power shapes the strategic dilemmas that aspiring dictators have to solve. For example, if the seizure group is riven by factions (say, among ethnic groups or clans), then this gives dictators more discretion to amass personal power than they would have in the context of a unified seizure group (see Chapter 4).

The implicit theme is that authoritarian politics is the politics of survival. The politics of an institution such as the military, in other words, is about the fact that militaries contain specialists in violence who have a particular capacity for overthrowing dictators. The politics of a legislature is about how that legislature promotes the dictator’s interests. And so forth. This tight focus on survival is clarifying, in that it provides a common optic for understanding the various things that authoritarian regimes do. It is also common among related “big picture” treatments of authoritarian regimes (see, e.g., Daron Acemoglu and James A. Robinson, Economic Origins of Dictatorship and Democracy, 2006; Bruce Bueno de Mesquita, Alastair Smith, Randolph Siverson, and James Morrow, The Logic of Political Survival, 2003; Jennifer Gandhi, Political Institutions under Dictatorship, 2008; and Milan Svolik, The Politics of Authoritarian Rule, 2012). It might strike some readers as excessively spare, however. Readers who are interested in how authoritarian regimes create development policy, how they propagandize, or how they build national identity will have few tools from this book to work with, aside from the entry point that however those things happen, the dictator’s calculus of political survival will matter. Still, that is not such a bad way to start, even if it does not get us very far on its own.

Turning to the core empirical chapters, what distinguishes this book from the other books on authoritarian politics cited is its breadth. So the seizure group is the main conceptual move, and the logic of survival is always at play, but arguments do not cumulate as tightly as they do in other treatments. There is not a single, underlying theoretical model; instead, Geddes, Wright, and Frantz build their case through an accumulation of smaller verbal theories and lots of evidence. In exchange for parsimony, the authors have opted for explanatory capacity, and by collecting data that no one else has on all sorts of variables that other authors have not yet measured, they are able to test more claims and fill out more features of authoritarian rule. Other researchers interested in these variables will surely have an opportunity to build on the findings in this book.

At the same time, the individual chapters can be unsatisfying. Verbal theories are presented in terms of tendencies, and supported by evidence about how things usually work. But some regimes are exceptional, and those exceptions fall under the authors’ radar. For example, dictators tend to create parties in order to marginalize the military (Chapter 5), but other pathways to party creation (in those cases where parties are created when the theory indicates that they are not needed) are not explored.

Taken together, How Dictatorships Work is most effective in discussing dictatorships that come to power through coups and dictators who personalize their rule. That the former is the most common way that dictatorships come to power is helpful—the book explains a lot. That personalization has never been so carefully and consistently measured across regimes and across time gives the authors a window into the dynamics of authoritarian rule that few other treatments have. But other cases that do not comport to the general tendencies found in the data are not so easy to understand through the authors’ framework, and some of these are interesting cases. The book has surprisingly little to say about the USSR, or about politics in countries ruled by communist parties or even mass-mobilizational parties like UMNO in Malaysia, the PAP in Singapore, the KMT (Kuomintang) in Taiwan, or the CCM in Tanzania. It is better at dictators like Kwame Nkrumah, Augusto Pinochet, and Islam Karimov; a peek at the index reveals more mentions of Mobutu Sese Seko and Gamel Abdel Nasser than of the Soviet Union, and most discussions of China focus on personalism over time rather than the structure and organization of the Chinese Communist Party.

A final point concerns changes over world-historical time. At various points the authors present findings over time (postseizure party creation from 1945 to 2010 [p. 114], elections and coup attempts from 1950 to 2010 [p. 180], personalism and democratization from 1950 to 2010 [p. 212]). Yet they do not focus on how changes in the international system might condition their theoretical approaches to the ways in which regimes work, opting instead to interpret differences across time periods as affecting the stock of regimes that emerge; that is, the post–World War II period is a period that generated many new regimes that did not yet have parties (p. 114). If “how dictatorships work” is timeless, this is fine. If dictatorships work differently depending on the world around them, then there is more to explore. Happily, Geddes, Wright, and Frantz have provided us with new data and a coherent theoretical approach from which to begin that exploration.