The creation of a stable democratic regime is in part contingent on how effective political leaders within a territory are at building a state, and at consolidating and legitimizing the state's power. At first glance, this relationship seems paradoxical. How could the creation of a democratic regime be affected by the formation of the state, which is a unit committed to protecting its independence from society's dictates, when democracy is driven by the proposition that the state is a decision-making unit that must do what society mandates?

Fernando López-Alves posits an intriguing explanation as to why some Latin American states have been more proficient than others at resolving this puzzle. His analysis is anchored to the imperatives of two questions: 1) How and to what degree does the state gain "control over the principal means of coercion within a defined territory"? 2) What is the relationship between the drive to gain control over the principal means of coercion and the drive to form a regime?

To address these interrelated questions, the author relies on standard definitions of state and democracy. He refers to the state as an entity that has a relatively centralized, differentiated, and autonomous structure, and that controls the means of coercion, within a particular territory. By democracy he means neither social nor economic democracy, but procedural democracy. Though his definitions are conventional, his methodology is not. Via the imaginative combination of the "deep analogy" and the "most differentiated" methods, he carries out a painstaking analysis of state creation and regime formation in five countries. His selection of the cases is dictated by two questions: Why did societies that shared a number of economic, cultural and societal features not end with similar states and regimes? And conversely, why did societies with different economic, cultural, and societal features end with similar states and regimes? He uses Uruguay, Colombia, and Argentina to elucidate the theory, and Paraguay and Venezuela to gauge its applicability.

Though López-Alves is not prepared to contend that his conclusions are definitive, his analysis should force students of democracy in general, and Latin American politics in particular, to reconsider some of their earlier assumptions and arguments. On the one hand, he demonstrates, first, that the composition of a state's ruling coalition does not necessarily determine the type of political regime it forms and, second, that high levels of economic development do not always facilitate the development of more inclusive, democratic regimes. On the other hand, he shows quite convincingly that the types of war, along with the type and scope of mobilization of the rural poor during state formation, shaped institutions, civil-military relations, and regime outcomes. To understand better the last contention, I will break it into two parts.

First, López-Alves proposes that the type of conflict that ensued during the postcolonial period in a particular territory had a greater effect on the formation of the polity than either the number of wars fought or their intensity. By this he means that a political regime had a better chance of being more democratic than of being more authoritarian whenever most of the support for revolutionary wars came from the countryside instead of the city. Revolutionary wars that gained their support primarily from the city typically enabled the military to both increase its power vis-a-vis, and autonomy from, the state. Second, the author suggests that it mattered which group mobilized the rural lower classes into war and politics. The likelihood of a political regime's being more democratic rather than more authoritarian was in no small measure determined by whether mobilization was led by the political parties or by the military.

Though intellectually challenging and empirically sound, López-Alves's work is burdened by two distracting shortcomings, one greater than the other. One of the central hazards faced by any political analyst in search of causality is to decide how far to take the inquiry. To López-Alves's credit, his work removes any doubt we may have had about the effects of state creation on regime formation. But his analysis is partially weakened by his disregard of factors that have long been known to affect the creation of the state and the consolidation and legitimization of its power, such as size of the territory, types of natural resources, topography, and ethnic diversity. Consideration of these factors, for instance, helps us understand why Chile and Costa Rice found it less difficult than Peru and Guatemala to create their states and to establish relatively stable democracies. By no means do I suggest that López-Alves's independent variables are irrelevant. I am merely proposing that their significance would become more evident were they to be placed in the context of the four factors I just identified.

However, my central displeasure with López-Alves's work is of a very different nature. The quality of a comparative political theory is determined by more than whether it is provocative, intelligent, and grounded in important historical debates. Some seventeen years ago, my mother-in-law, after generously volunteering to edit my dissertation, forced me to adopt a simple, but valuable, rule. Confronted with my reluctance to rewrite incomprehensible passages saturated with political jargon, she said: "I do not know as much about politics as you do; but I am an intelligent person. If I do not understand what you have written, it is because it is badly written." As I struggled to unravel Fernando López-Alves's argument, I was reminded how much he would have benefited from an editor with my mother-in-law's acumen. Burdened by a dissertation mentality that seems unwilling to accept that the reader will take his theory seriously even if the main text is not crammed with every available source, and by a propensity not to edit complex ideas into decipherable sentences, López-Alves commits a disservice to both himself and the reader. It could have been a wonderful book.