Bunce, Valerie. "Comparative Democratization: Big and Bounded Generalizations". Comparative Political Studies, Vol.33, No.6 (2000): 703-734.

Bunce, Valerie. "Comparative Democratization: Big and Bounded Generalizations". Comparative Political Studies, Vol.33, No.6 (2000): 703-734.

• There is very little agreement within political science about what causes or maintains democratization. This is because there are so many different paths to democratization, making generalization extremely difficult (705).
• "The comparative study of democratization, therefore, has produced five generalizations [...] high levels of economic development function as a virtual guarantee of democratic continuity, political leaders are central to the founding and design of democracy and to its survival or collapse under conditions of crisis, parliamentary systems are [...] far better [...] than presidential systems, settlement of the national and state questions are crucial investments in the quality and survival of democracy, and [...] new and fragile democracies have, [...] uncertain results [...] [and] uncertain procedures" (715).
• Democracies are significantly more likely to be successfully consolidated if the country is wealthier. Although poor countries have become established democracies, and democratization can fail in rich countries, wealthier states are overall more likely to democratize successfully (706). There is almost no consensus on why rich countries democratize more successfully (707).
• Scholars agree that political elites and their attitudes play a significant role in determining the success or failure of democratic transitions, but fail to reach agreement on what values or behaviors encourage democratization or why. While divided elites sometimes allow for democratization, cooperation between elites can encourage its success. (707-708).
• The survival of a nascent democracy depends heavily on the actions of major political and military elites, while the quality of democracy is much more dependent on a wider range of institutional, international, and public factors (710).
• Parliamentary democracies are significantly more likely to survive past infancy than presidential democracies. The sociological and historical reasons for this fact are hotly debated, however, and the effect of system choice varies depending on case-specific factors (711).
• In general, states with multiple national groups or contested borders are much less likely to democratize, however a number of prominent successful cases of democratization is these states demonstrate that other factors might be at play. Moreover, nationalism, contestation of borders and the nature of the state, and recent histories all factor into democratization in unique and confounding ways (712-713).
• Democratization in Latin America and the former Warsaw Pact has produced nations with strong legal commitment to democracy, but poor implementation or absence of the rule of law (713). The retreat of state power in former autocracies should be largely considered in this context, esp. in non-capitalist areas of the world (714).
• This has forced studies of democracy to either accept legal definitions as evidence of democracy, and ignored huge discrepancies in practice, or to modify their measurement of democracy to account for serious issues in the implementation of national law (714).
• All democratizing countries face three similar issues: ending authoritarian rule and beginning democratization, building democratic institutions, and securing the cooperation of the former authoritarian elites. Significant regional variations still occur, however, in what strategies are most successful at approaching these challenges (715).
• For example, one of the key factors to the successful democratization of many Central European nations was their complete rejection of socialism and ostracism of former Communist elites, contrasted to elite continuity in the rest of the Warsaw Pact, whereas democratization was successful in southern Europe largely because authoritarian elites were included in the transition and the new democratizing government (716-717).
• The author questions possible bias in area studies work on regional democratization as being too narrowly focused, but work by political scientists has also demonstrated substantial regional variation in democratization (721).
• There are very large regional variations is the particular threats to successful democratization, which tend to be regionally oriented. For example, former Warsaw Pact nations tend to be highly educated and have strong civilian control over the military, but have inefficient economies and grossly underdeveloped states and civil societies. Whereas Latin American nations have economic wealth and experience with democracy, but have huge economic inequality, disloyal militaries, and personalistic party systems. Different threats are more salient to different democratizing states (720).
• Conceptualizating democratization suffers from the same issue that studies of democracies do, namely that the concept of 'democracy' is poorly defined, and too many states exist in the categories of 'hybrid regime' or 'flawed democracy', vague designations which do not say anything important about the actual function of that democracy (723-724).
• This entire article is just a giant source mine of information about the theories of democratization and comparative democracy.