In this book Michael Bernhard asks two questions: Why do new democracies pick the institutions that they do? How do certain patterns of institutions affect the chances of democratic survival? And his theoretical framework, in its initial formulation, makes two claims. First, institutional choices are a product of powerful agents pursuing their narrow interest. If you know who is powerful and what they want, you will be able to predict the institutional outcomes. Second, the key to democratic survival, at least in the initial posttransition phase, is the breadth and strength of the original institution-forming coalition. A key conclusion of the study is that democracies are in trouble when the coalitions that brought about their institutional arrangements lose power too soon, as happened in prewar Poland and in Weimar Germany. And this can occur for a variety of reasons: New actors may be brought in, for example, or the salient issues may change.
After briefly presenting this framework, Bernhard elaborates, testing its usefulness on four cases from two countries: Weimar Germany and interwar Poland provide examples of democratic failure; the Bundesrepublik and post-1989 Poland provide examples of democratic success. The core of Institutions and the Fate of Democracy consists of detailed analyses of these four cases, and the book is worth reading just for those examples. But there is the further issue of the value of combining the two questions he asks, of the usefulness of his theoretical framework, and, finally, of whether considering these four cases together adds anything we could not get by studying each case separately. I am not completely convinced on any of these counts. The framework seems to be missing some important ingredients. And actual cross-country comparisons play a surprisingly small role in the author's argument.
Nevertheless, the book shows how large n-studies can be complemented by more detailed historical accounts, allowing us to see better the actual politics that brings democracy into being, or destroys it. One can only hope there will be more and better such studies in the future. With Bernhard's approach, we can learn things about democratization even sometimes despite his framework, with its narrow empirical focus on the executive and legislative institutions and on electoral rules. A number of elements struck me as important in these four stories, despite not being featured in the theoretical framework.
First, peace treaties and the development of “boundary regimes” play an important role in these stories of democratization. Constitutions are also in part peace treaties, as I recently had many occasions to repeat in discussions of the Iraqi constitution. And peace treaties are also in part constitutions, we might add, thinking of Versailles and Riga as constitutive of the short-lived democratic systems of Germany and Poland after World War I. Constitutions are not to be identified with legal documents that bear that title.
Both external and internal boundaries in federal systems form a large part of the nature of the regime. In Germany after both wars, much time and effort was expended on deciding the nature of the federal regime. Bernhard tells the story, yet it is not reflected in his framework. And he neglects completely the debates for and against federalism that were so important in Polish politics during and immediately after World War I, because they were settled more by peace treaties and boundary decisions than by constitutional choice narrowly understood (that legal document again). Yet the way these issues were settled made a federal consociational regime extremely unlikely in Poland in the 1920s, and without that, democracy was going to have a hard time.
The peace treaties after World War I and the Soviet-Polish war left Poles with 69% of the population in Poland, with various minorities dividing up the remainder. This demographic split is perhaps the most difficult for democracy to work (think of Cyprus or Iraq). In such countries, minorities are big enough to require special protections, but the majority is sufficiently large that granting the minorities veto power is unlikely to be accepted as democratically legitimate.
A related lesson about democratization that emerges from Bernhard's account, without being reflected in his conclusions or his framework, is the importance of consociational and consensual institutions (Lijphart is almost completely ignored in this book). The presence and absence of consociational or consensual features in a system, such as federalism or constitutional courts, emerge as significant elements of the story. But they are not fully told, because they are not recognized in Bernhard's framework. As I read the German story, contrasting the Weimar Republic with the Bundesrepublik, it seems that one key cause of the change in democratic prospects is the shift in location of the “constitutional guarantor.” In Weimar, the monarch's role as guarantor is replaced by the president, with well-known disastrous consequences in the 1930s. After World War II, the constitutional court replaces the president, significantly improving (or so it seems) the capacity of German democracy to survive emergencies.
Bernhard's own conclusions focus on how well his framework has performed. It is not enough, he writes, to know the self-interested preferences of the most important actors to know the political outcome. Some political actors act on the basis of values, he continues. And we may add that the evolution of ideas about democracy is an important feature of the contrast between the two pairs of cases (before and after World War II). What constitutes narrow interest is not a simple question, he adds; preferences may change during the process of institutional choice, and the relevant actors can also change. So self-interest matters, but a flexible notion of interest is required, and context matters as well. We should be careful in building parsimonious models of politics. All true enough.
Bernhard's conclusions seem to me more striking on the question of democratic survival. The four cases strongly confirm, he writes, the importance of the political fortunes of the initial institution-framing coalition in the posttransition phase. This is a striking conclusion, because it suggests that the nature of the institution-forming coalition may be more important than the institutions they adopt. He does not discuss in any depth why this might be so. Let me suggest a possibility. Legitimacy matters (this would be a nice addition to Bernhard's framework), and to be seen as legitimate, the institutional arrangements of a democracy must be perceived as impartial or neutral. Over time, this perception develops as different groups win at different times. But at first the perception of neutrality is more difficult to establish. Perhaps the first winners are so because the system is biased in their favor. Hence, the initial importance of maintaining in power a broad institution-making coalition. For them, the rules of the game are neutral because they all agreed to them.
