Borrowing the dominant theoretical frameworks of studies of voter turnout in the developed world, this paper tests the applicability of these frameworks to a transitional democracy in a Muslim society. Using logistic regression, we estimate an individual’s likelihood of voting in Kyrgyzstan’s 2005 parliamentary and presidential elections.We find that traditional rural networks, rather than networks of voluntary associations, play the central role in the mobilization of voters. In addition, turnout is affected by long-standing cultural cleavages based on religion and ethnicity, and by education, occupation, and trust in government. Our findings suggest that the political behavior of the people of the Kyrgyz Republic is supportive of democracy, in spite of elite-level obstacles to a successful transition.

This study introduces basic models of voting behavior and describes in greater detail the findings and starting points of the model of party identification. It recapitulates findings related to the voting behavior in the Czech Republic throughout the nineties, as well as the reasons leading to the application of the party identification model to the analyses of voting behavior. The influence of positive feelings toward the elected party and negative feelings toward the other parties are tested here. The study explores in particular the influence of inter-party hostility on the voting behavior, and tests it both separately and in combination with the subjective social class and subjective political orientation.