Violent crime is becoming one of the most pressing problems in many developing economies and citizens usually associate spiking crime levels with political failure. How does the risk of crime victimization affect preferences for redistribution compared to other risks that the government is harder to blame for? While socio-economic and context factors affect individual preferences for social policy, not all explanatory factors are purely domestic nor equally influential for social protection demand. Risk that emanates from crime is more easily attributed to state failure than economic risk associated with market liberalization. Different types of risks carry different information about the role of the state in the development of the particular risk. I therefore argue that risks which are less easily attributed to government performance will be met with increased demand for compensation, whereas risks that signal low state capacity increase withdrawal from public social protection programs and support for redistribution. This study compares the impact of crime risk and economic risk on social policy preferences in Brazil, an emerging economy with high variation in globalization-related and domestic risks following from crime. I test the argument with a survey experiment employing original data from a randomized face-to-face survey from Brazil in 2019. The vignette randomizes risks of violence, trade liberalization and migration to compare domestic versus international risks on redistributive preferences. Empirical results corroborate the argument. Brazilians are less supportive of more generous social protection programs when confronted with an increase in violence compared to globalization risks.