Professor Cheibub's book is a significant contribution to one of the longstanding debates within comparative politics: whether presidentialism is systematically more unstable and prone to regime breakdown than parliamentarism. The book, then, intends to settle the question that was initiated in the late seventies by the work of Juan Linz, who stated his now well-known view about the perils of presidential regimes. Since the third wave of transitions to democracy, the differing contribution—if any—of presidentialism and parliamentarism to political stability and economic development has been thoroughly analyzed in the context of Latin American politics. Excellent contributions, such as those by Alfred Stepan, Scott Mainwaring, John Carey, and Matthew Shugart, have argued that the basic regime-type distinction proved insufficient for a detailed account of how different patterns of executive / legislative relations really work. However, Cheibub's analysis takes a different approach. The author builds on formal analytical tools, rather than purely institutional description, and with plenty of quantitative data to test whether presidential regimes are associated with the disastrous consequences for democracy that the classic Linzean version predicted. The combination of formal and statistical analysis makes Cheibub's argument—that presidential democracies are not intrinsically unstable—forceful. The principal explanation for this finding is that coalitions do form in presidential regimes and that instability is the result of historical legacy of authoritarianism. Cheibub also considers and rejects research using traditional political economy variables, such as country size or the region's economic relationship with the rest of the world. In finding that regime-type is irrelevant to stability, Cheibub's work confirms the findings of the collective enterprise headed by Adam Przeworksi (in which Cheibub participated) that resulted in the book Democracy and in Development (2000). In his new book, Cheibub takes the argument one step further, addressing the Linzean view step-by-step and analyzing its claims carefully. The plot of the book is straightforward. Cheibub first explains that Linz's argument does not accurately describe the performance of presidentialism because he ignores the role of coalitions in presidential systems. This leads Cheibub then to scrutinize the relationships between presidentialism, democratic performance and minority government, constitutional powers of the president, and party discipline. He also analyzes the historical legacy of authoritarianism in the region, concluding that these legacies are more important to democratic stability than presidential systems.
The first chapter reviews and criticizes the classic Linzean argument about the “perils” of presidentialism. The case advanced by Linz is that while presidential regimes generate mutually independent branches, parliamentarism does exactly the contrary, producing mutually dependent branches. The separation of powers in presidential regimes and the lack of votes of confidence generate both a zero-sum game (where there are no incentives for coalition formation) and a party system that lacks discipline. These ingredients combine to create a system prone to divided government and legislative deadlock. The fixed term for the president means that the regime lacks a safety valve that can help solve these deadlocks. Therefore, incentives are given for extra-constitutional solutions (namely, coups and military uprisings) for deadlock and stalemate. This is the Linzean view, which Cheibub will analyze (and criticize) in the next chapters. A methodological chapter follows, in which the concepts of democracy, presidentialism, and parliamentarism are carefully defined. The author here presents a definition of regime types that follows the methodology that Przeworksi and associates advanced in Democracy and Development. The argument of the book starts in chapter 3, where Cheibub proposes a rigorously formal model to show that incentives for coalition formation do exist in presidential regimes. The results of the formal model are, first, that coalitions can take place under presidential regimes. The model also predicts that parliamentary governments should be more successful in forging legislative coalitions than presidential ones but that presidential coalitions are less likely with strong legislative powers. In spite of the second finding, the model predicts that majority status of presidential systems should be uncorrelated with legislative paralysis and the demise of the regime. The status of the government (whether it is minority or majority government, single party or coalition) does not affect the probability of survival of the regime. In other words, there is no observed correlation between the majority status of a government and the possibility of a democratic breakdown. Hence, this chapter provides an alternative logic against the Linzean argument. The following chapters are devoted to investigating these claims empirically with statistical support. Cheibub shows that coalitions do, in fact, form in presidential regimes and that minority governments do not necessarily pose a threat to presidential democracies. These two claims contradict the core of the Linzean traditional view. In his last chapter, he also states that if presidential regimes die more often than parliamentarian ones (which they do), it is due to the contexts in which these regimes were installed and the problems democracy faced in these countries. In particular, a “military-presidential nexus” exists that affects the relationship between presidential regimes and the survival of democracy. Presidential regimes tend to follow military dictatorships rather than civilian ones. And the former are more likely to become dictatorships than the latter ones. According to Cheibub, if Latin America had chosen parliamentarian instead of presidential regimes in the mid-nineteenth century, the discipline would not be discussing the critical instability of the latter, but probably of the former.
The argument presented by the author has the merits of not only being able to dismiss the myths about presidential regimes but also giving an adequate explanation to an undeniable fact: presidential regimes are (were?) more unstable than parliamentarian regimes. More than proving Linz wrong (or at least partially wrong), Cheibub proves him incomplete. In other words, Linz was right in what he had observed, but not in his explanation. The author accomplishes these tasks using an admirable blend of qualitative and quantitative tools. Finally, the explanation draws attention to the importance of the path-dependent approaches: What other type of regime could the Latin American republics have chosen in the 1850s, given that they were under the intellectual influence of the United States Constitution, and at a time when parliamentarism, as we know it nowadays, was yet to be born? One aspect that the book does not examine profoundly is the overall role of institutions. First, if the historically institutional explanation accounts for the democratic problems presidential regimes experienced, should we conclude that institutions do not matter? What is the role of institutions in the democratic performance of the region? Related to this, the book does not answer the question of how presidential regimes can make coalitions more likely. In other words, Cheibub makes clear that presidential regimes have the ability of generating coalitions, but not how presidential regimes differ in this matter, or how presidential regimes differ from parliamentarian regimes. It is a valuable contribution to show that presidential regimes are able to construct coalitions. However, we still do not know what institutional component (if any) contributes to coalition formation other than the president's relative strength. Cheibub gives party discipline an important role in the formation of coalitions. But the relationship between the former and presidential strength remains unexplored. He also states that discipline can be generated under presidential regimes, but we do not know how. Is it not also a possibility that the political party variable (strength, organization) plays some part? His quantitative attempt to disentangle the effects of legislative fractionalization and presidential power does not offer conclusive evidence about the broad question about the role institutions can play in generating discipline and (eventually) a more collaborative relationship between governmental branches. The next step in this research should be an in-depth analysis of the conditions under which coalitions are likely in presidential regimes. We know that presidential regimes are not created equal. Are some institutional features of presidential democracies more conducive to coalition formation? In this respect, what the book does not answer is, are coalitions equally distributed among presidential regimes and, if not, what makes this distribution unequal? A qualitative study of a handful of coalitions in Latin America could be helpful in unraveling the conditions under which coalitions take place. Furthermore, this kind of study could provide more definitive evidence as to whether coalitions help a regime avoid a lethal crisis. Presidentialism, Parliamentarism, and Democracy is a significant contribution to the field of comparative politics. It is also a good example of how scientific knowledge is built, as it resumes a relevant institutional discussion with sound arguments and appropriate methodologies. Further, the discipline will benefit from the discussion it will generate in terms of both substance and method.
