Political philosophers have long argued that democratic states foster the welfare of their citizens better than autocratic ones. Recently, those normative arguments have been supported by both theoretical and empirical research, which claims that democracies produce better social policy and provide more health and education services to their citizens. This literature argues that regularly held elections create incentives for the provision of public goods: in democratic countries, citizens elect politicians who deliver policies and services that improve citizens' welfare, and vote out of office those who fail to do so. However, this parsimonious argument does not explain why some democratic governments successfully improve the wealth and health of their citizens, while others suffer from public policy failures - "the underprovision of public goods and the overprovision of regulations and laws that benefit special interests at the expense of the whole society" (Keefer, 2005, p. 314). Nowhere are the differences between good and poor policy, and their consequences for human welfare and misery, more visible than in the ways developing countries have responded to AIDS. The research question guiding this dissertation project is: what accounts for the differences in the way democratic governments in the developing world have responded to the epidemic? I argue these differences are due to the variation in imperfections in their electoral markets. Specifically, in countries where 1) the citizenry lacks access to information, 2) the society is ethnically polarized, 3) politicians face commitment problems, and 4) a tradition of political organization is lacking, the response to the epidemic is slow, policies and institutions aimed at fighting AIDS take longer to develop, spending on AIDS and the availability of prevention and treatment services is low, and the results of efforts to fight the disease, such as behavioral change in those at risk of infection, are limited.