Regime Change in Africa Joel D. Barkan (bio) Democratic Experiments in Africa: Regime Transitions in Comparative Perspective. By Michael Bratton and Nicolas van de Walle. Cambridge University Press, 1997. 307 pp. Democratic Experiments in Africa is an ambitious and important book that should be read by every serious student of democratization and by every serious student of African politics. It is ambitious and important in at least three ways: First, it seeks to formulate and test a theory of democratic transitions that is applicable to sub-Saharan Africa, a diverse region of 47 countries. Second, it is truly comparative in its approach. No other work has sought to compare the attempted transitions of all African states with one another and with transitions elsewhere. Third, it is systematic and quantitative in its methodology. Although others have measured and compared the extent of democratization across Africa, none have sought to explain these variations rigorously via the specification of a series of causal models, and then to test the validity of these models with comparable data for all countries that have attempted democratic transitions. Indeed, the development of the data set on which this study is based—an effort supported by the National Science Foundation and comprising nearly a hundred measures of democratization—is a Herculean effort for which the authors deserve to be commended. Notwithstanding its importance and rigor, Democratic Experiments falls short of its stated goal of finding a single explanatory theory of democratic transitions in Africa. Put simply, its reach exceeds its [End Page 165] grasp. Yet unlike many quantitative studies that seek to measure the relationships among dependent and independent variables, Democratic Experiments addresses a big question and delivers more than a series of statistically significant but substantively trivial findings. Although this book does not yield a single theory of regime transition, it does make considerable progress toward that goal by rejecting much conventional wisdom about the origins and outcomes of the transition process. As such, it is an example of both the limitations and the promise of this genre of social-science research for understanding democratization. Bratton and van de Walle make three overlapping arguments about the sequences and causes of democratization in Africa between 1988 and 1994, a period during which 40 African countries initiated tran-sitions from authoritarian rule: 1) Democratic transitions unfold in a specific sequence of events beginning with political protest, followed by political liberalization, the holding of a democratic election, and the installation of an elected government; 2) democratic transitions in Africa are largely explained by domestic political forces rather than by domestic economic conditions, international political factors, or international economic conditions; and 3) democratic transitions unfold in a process of “structured contingency”—a combination of individual actions, particularly the actions of elites, that occur within the institutional structures that a country inherits from the pretran-sition authoritarian regime. The most important of these structures is the web of informal rules that served as the basis of “neopatrimonial” political authority in both civilian and military regimes prior to the transition. The extent to which these authoritarian regimes provided for a measure of political competition and participation also deter-mines whether the transition will proceed or be blocked, and if allowed to proceed, whether it will be flawed or successful in in-stalling a new regime that plays by democratic rules. The first and second of these arguments are articulated best, and are relatively easy to follow. Unfortunately, the authors’ analysis of their data set forces them to back away from these two arguments (especially the first), and to supplement them with the third. By embracing the argument for “structured contingency,” the authors acknowledge the difficulty of arriving at a single theory of regime change, since their explanation rests in large part on the individual actions of key actors as they respond to one another’s moves. As Bratton and van de Walle freely acknowledge, some stages of demo-cratic transitions are consequently “more random” than others, while “the role of serendipity and national idiosyncrasies loomed larger as transitions progressed” (pp. 186, 188). The authors begin with an introduction and three chapters on the meaning of regime change and democratic transitions in the African...