This article shows that highly correlated measures can produce different results. We identify a democratization model from the literature and test it in over 120 countries from 1951-1992. Then, we check whether the results are robust regarding measures of democracy, time periods, and levels of development. The findings show that measures do matter: while some of the findings are robust, most of them are not. This explains, in part, why the debates on democracy have continued rather than been resolved. More importantly, it underscores the need for more careful use of measures and further testing to increase confidence in the findings. Scholars in comparative politics increasingly are drawn to large-N statistical analyses, often using datasets collected by others. As in any field, we show how they must be careful in choosing the most appropriate measures for their study, without assuming that any correlated measure will do.

The logic of historical explanation obliges one to understand temporality as a moderator of various effects on political outcomes. Temporal problems remain in the empirical analysis of political phenomena, however, especially as it pertains to categorical data and long-term time dependence. Many theories in political science assert that sequencing matters or that political outcomes are path dependent, but they remain untested (or improperly tested) assertions for which sequence analysis may be valuable. This article briefly reviews the disciplinary origins of sequence analysis and applies the method in order to understand bargaining between actors during national crises. Finally, it explores the robustness of a commonly used sequence analysis metric. The ability to demonstrate and separate sequential effects from accumulative effects—made possible through sequence analysis—constitutes a major step in political science toward analyses that are truly time sensitive.

This article considers the experiences of 24 countries that began a transition toward democracy between 1973 and 1993. To understand why some of these transitions succeeded while others failed, it compares the explanatory power of three factors: the path a country followed through the transition process, the type of authoritarian regime that was in place, and the protracted nature of the transition. It shows that in countries where actors hold diverging preferences, one-party regimes refuse to exit, or transitions are protracted, the result can as easily be a continuation of authoritarianism as of democracy. However, if they successfully install democracy, then these cases are the most likely to continue towards consolidation. Thus, difficult transitions, while risky, offer the highest payoff for democratization.

This article reexamines military intervention in the overthrow of civilian governments in the Third World. First, it reconsiders three major theories developed to explain the military overthrow of the Goulart government in Brazil in 1964, each of which focuses on a different independent variable: institutional change, economic stress, and political polarization. Second, it applies and demonstrates the usefulness of these three theories to a recent case outside Latin America: the Philippines. Supporting evidence is found for each of the three theories in both the Brazilian and the Philippine cases. Third, it offers new evidence on the Philippine case, based on 33 interviews with military officers and other informants knowledgeable about the military's role in the overthrow of the Marcos regime in the Philippines in 1986. The interviews were conducted in the Philippines from May to July 1988. The informants consisted of 17 active military officers, 8 retired officers, and 8 outside observers. The article concludes by showing that each theory complements the others, since they focus on different aspects of the process, and it argues that a synthesis of these contending theories is possible.

No abstract