Given that formal insurance markets in developing countries are very limited, poor households typically rely on the help of family or friends in times of economic hardship. These informal exchanges of gifts, loans or labour, which are motivated by social preferences or strategic incentives, serve de facto as risk pooling devices and are an important, though not complete source for households to cope with negative income shocks. A large body of literature investigated forms, motives and constraints of such informal risk sharing arrangements. However, little attention has been paid to the relationship between mutual support and the extent to which individuals can control their risk exposure. This issue refers to the fact that (positive or negative) income shocks can either be the consequences of risky choices (e.g. investments) or completely random events (e.g. accidents which affect work capacity), a distinction which might be quite relevant for solidary behaviour for mainly two reasons. Firstly, evidence from the Western world suggests that a considerable proportion of individuals favour redistribution when inequalities are caused by exogenous circumstances rather than by factors of personal responsibility. Secondly, while the first point focuses on the behaviour of the beneficiaryof solidarity, there is also reason to believe that it matters for the willingness to give whether the donor deliberately accepts risk for earning income or whether he can earn the same income by pure luck. In the former case he may view part of the earned income as compensation for bearing risk and choice costs resulting from foregoing safer opportunities which may induce him to share less. In this project we investigate in a developing country whether individuals condition their giving behaviour on the extent to which they and their partners can influence own risk exposure using an incentivized experimental approach. We test whether poor individuals' solidarity is sensitive to the degree of control subjects have over their risk exposure in a context without public social safety nets based on laboratory experiments we conduct with slum dwellers in Nairobi. The output from this project is available here.