Who supports social policy in the developing world? Most of what we know about micro-level preferences for social policy comes from well-developed, wealthy countries of the OECD, where governments can credibly commit to policy enforcement and implementation. This dissertation explores preferences for social policy in post-communist countries, where weak constraints on the state challenge the welfare state. In doing so, it provides novel insights both into social policy debates in these countries and the coalitions which support (or oppose) social policy. I argue that support for social policy depends on how institutions shape the expectations of actors about the costs they pay into social policy programs versus future benefits. I draw on existing theories of political economy to propose four mechanisms -- misappropriation, contract enforcement, free-riding, and macro-economic risk -- that alter the distribution of winners and losers from social policy. Misappropriation stems from officials' ability to divert funding away from intended uses. While for most this imposes dead-weight costs on social policy, where institutions are poor. the politically well-connected can benefit from diverted funds to decrease social policy costs. The contract enforcement mechanism emerges due to the inability of weakly constrained states to enforce contracts. Predictions are similar to misappropriation, but actors also cannot trust other private actors with control of social policy. Free-riding emerges when bureaucrats are unwilling to expend effort to ensure tax compliance. Again, this imposes dead-weight costs on most, but garners support from tax evaders, who can free-ride. Finally, the macro-economic risk mechanism suggests that macro-economic volatility is heightened in settings with weak institutions, which increases both individual risk and support for social policy. The empirical portion of the dissertation tests the observable implications of each of these mechanisms. Chapter 2 provides a first-cut, cross-national test of part of the argument using micro-level data from a cross-national survey of 28 post-communist countries. I draw on work on informality in the post-communist world to identify individual characteristics associated with tax evasion to test the free-rider mechanism. Consistent with it, I show that those associated with evasion support social policy more where institutions are weaker. Chapter 3 posits that if the mechanisms I propose matter, actors will appeal to the logic of my theory during concrete reform debates. I test this using evidence from the 2001 pension reforms in Russia. I combine analysis of the legislative debates surrounding reform with in-depth content analysis of the Russian media, which draws on an original dataset of all mentions of reform in 352 Russian newspapers, journals, and trade magazines. I show that all four mechanisms were indeed major concerns. Chapter 4 tests the theory at the firm level, using a survey of 666 Russian firms to look at preferences where institutional quality is weak. I test whether firms that I predict support the welfare state in such settings -- those with political connections and a comparative advantage in hiding from the authorities -- actually do so. In addition to providing some support for the misappropriation and free-riding mechanisms, this chapter is a contribution in its own right: it is among the first to use surveys to study firms' preferences for social policy. Finally, chapter 5 uses a survey experiment conducted on 1600 respondents to attempt to understand the ceteris paribus effect of institutions on the average individual. Using a simple framing experiment, I provide three different treatment groups with information about bribery, tax evasion, and the extent to which private pension funds commit fraud to test the misappropriation, free-riding, and contract enforcement mechanisms, respectively. The chapter offers mixed evidence. The dissertation makes contributions to both the study of the welfare state and the political economy of institutions and investment. First, the dissertation explores preferences for social policy in the developing world and introduces institutional quality concerns to this literature. My work particularly focuses attention on the ways certain groups can abuse social policy to pass costs onto others, adding nuance to existing understandings of who benefits from social policy. Second, it advances our understanding of how institutional quality shapes economic decision making and provides evidence as to how different pathologies of poor institutions shape economic decisions.