Gretchen Helmke’s book makes a significant contribution to the study of presidentialism in Latin America. It advances an elegant and persuasive theory of interbranch crises, backed by a wealth of empirical evidence. Institutions on the Edge tackles real-world problems: presidents being impeached, legislatures being closed or superseded by constitutional assemblies, and judiciaries being forcefully restructured for political reasons. Building on prior insights provided by game-theoretical models, Helmke offers a theoretically rich account of interbranch crises that speaks directly to current debates about the performance of presidentialism in Latin America.

Helmke starts by acknowledging that presidents have two types of power: formal institutional power derived from constitutional rules, and partisan power derived from their party’s level of support in congress. Her main argument is that the gap in the balance between the president’s institutional and partisan powers affects the likelihood of a presidential crisis, and changes in this gap cause legislative and judicial crises.

The book is organized into seven chapters. The first highlights the relevance of the topic as well as its core argument and its connection to prior literature. The last chapter summarizes the main findings and discusses some implications for future research. The intervening chapters develop the theoretical framework, specify the testable implications, and present the results of the empirical analyses.

The second chapter introduces the data set of interbranch crises in Latin America. Crises are defined as conflicts wherein one branch challenges the composition of another branch. They include instances in which one branch threatens another branch with removal, as well as events in which the goal is to change the median voter of the court or the legislature. To identify crises, the author applies a series of selection rules to news articles appearing in the publication Latin American Weekly Reports over the period from 1985 to 2008. The data set is not only well constructed to test the arguments advanced in this book but should also be useful to other scholars interested in political institutions in Latin America. Some may wonder whether the source used to identify crises may focus more on news from large rather than small Latin American countries, thereby missing some observations from the latter group. However, this is unlikely to be a problem here because the events of interest (institutional crises) are of the type of news included by this source in its weekly reports, regardless of the country of origin. This chapter presents an illuminating overview of the many crises that have occurred in the region, and provides needed information for researchers interested in extending this valuable data set.

The theoretical framework of the book is presented in Chapter 3. It begins with the assumption that presidents and legislatures bargain over the president’s authority to set policy. Institutional crises represent bargaining failures. The size of the rewards affects the threshold for conflict: the greater the presidents’ power, the greater the legislatures’ incentives to fight. Bargaining occurs under the shadow of power, in this case the capacity of the legislature to remove a president from office. This is driven by the president’s partisan powers, more specifically, the minority status of the president. Asymmetric information makes bargaining difficult: Political actors are uncertain about the costs of the crisis and each other’s tolerance for such costs.

Helmke hypothesizes that the likelihood of presidential crises increases with presidential powers, conditional on the president’s party being in the minority in the legislature. She also expects presidential crises to be more likely as the legislature’s costs for attacking the president decrease. These costs can be operationalized, for example, as mass protests against the government. Legislative crises—in which the aggressor is the president and the target is the legislature—are essentially the result of presidents waging a preventive war against the legislature to avoid the future possibility of removal.

The main hypotheses regarding presidential crises are tested in Chapter 4. The results confirm Helmke’s argument: Greater presidential powers increase the likelihood of a presidential crisis when the president’s party lacks a majority in congress, and protests make these crises more likely to occur. These results are robust to different specifications.

Legislative crises are the focus of Chapter 5. Such events involve, for instance, attempts at closing congress or replacing it with a constituent assembly. The results show that legislative crises increase with presidential powers when the president is in the minority, as previously theorized. Also important are findings showing that presidential popularity does not make these crises more likely, but prior instances of presidential crises do. The chapter concludes with an in-depth look at four well-known legislative crises.

The empirical analyses present convincing evidence that presidential powers matter and that their effect is conditional on partisan strength, but some may still wonder about the reasons that the formal powers of the president lead to interbranch crises. The argument is that the formal authority of the president (when in the minority) is meant to increase the incentives for presidents to go it alone or to rule around the legislature (pp. 75–76). Yet almost all measures of presidential strength derived from constitutional provisions are constructed on the basis of formal agenda-setting prerogatives that favor the president in his or her bargaining with congress; they do not primarily measure the ability of presidents to legislate unilaterally (e.g., via law decrees). Notwithstanding, the analyses provided in Chapters 4 and 5 show that presidential and legislative crises are indeed more likely to occur in contexts where constitutionally powerful presidents lack a partisan majority in congress.

Chapter 6 shifts the focus to judicial crises. The judiciary matters for interbranch bargaining because it can influence the parameters that affect the other branches’ likelihood of conflict. Helmke argues that presidential attacks against the judiciary are fueled by a dynamic similar to that of attacks against the legislature: aversion to being ousted. The core hypothesis, however, differs. She argues that increases in presidential powers make judicial crises more likely, but that this effect is not conditional on the president being in the minority (as it was in the other type of crises). Helmke’s explanation makes intuitive sense, but raises some questions about the hypothesized motivation behind judicial crises (i.e., the notion of presidents as fearful of being ousted). The empirical results show that prior instances of presidential crises make judicial crises more likely, while confidence in the courts reduces the likelihood of such crises. More central to the argument is the finding that presidential powers make judicial crises more likely. Also revealing are the results of the full model of presidential attacks on the courts, which shows that minority presidents are less likely than majority presidents to initiate judicial crises. Since prior chapters argued that majority presidents are not likely to fear being ousted, this finding seems to raise questions about whether presidential attacks against the judiciary are, in fact, driven by a dynamic similar to that of attacks against the legislature.

In conclusion, Helmke’s timely book presents a well-crafted theory of interbranch crises. The underlying logic builds on, and successfully extends, prior insights about the workings of presidentialism. The empirical analyses are revealing and confirm the author’s core arguments. Institutions on the Edge makes a significant contribution to the literature on presidentialism and Latin American politics. It should be essential reading for anyone interested in executive-legislative bargaining and institutional crises.