This dissertation examines variation in welfare expenditure levels in three mineral-rich post-Soviet states: Russia, Kazakhstan and Azerbaijan. The project asks why commitments to welfare spending in three countries with similar conditions—including regime type, vast mineral wealth, and the communist legacy of central planning and a cradle-to-grave welfare system—diverged markedly in the two decades since the Soviet collapse. To explain this outcome, I develop two theoretical frameworks that distinguish between arenas of welfare spending decision-making in authoritarian regimes that have the capacity to capture large rents from oil and gas exports. In the first, a single institutional actor makes all welfare spending decisions, and in the second, multiple institutional actors are involved in the decision-making process. Drawing on these theoretical conceptions as well as the literature on political economy and elite behavior in authoritarian regimes, I hypothesize several factors that may influence welfare expenditure levels: (1) leaders' expectations about staying in power; (2) the strength of civil society; and (3) the number and character of institutional actors involved in welfare spending decisions. I will test these hypotheses using both qualitative and quantitative data collected in the field between August 2010 and August 2011.