In this dissertation, I study the institution of informal risk sharing as a consumption insurance mechanism. Although this topic has global applications, I chose to focus on Russia. The transition to a market economy there introduced substantial income uncertainty, whereas the underfunded social security system failed to provide adequate insurance. Thus, Russia provides a unique opportunity to test the various theories of risk coping. The first section develops the model of informal risk sharing in continuous time, which is then used to demonstrated that risk sharing is facilitated when shocks occur frequently, are of short duration, and mostly idiosyncratic in nature. Cooperation is also promoted when shocks entail heavy, when households are strongly risk averse, and when they expect to cooperate for a long period of time. It has repeatedly been stated that public and private transfers are close substitutes. In this model, the crowding effect is more than one for one. Earlier research has shown that altruistic partners find it harder to implement a scheme of complete income sharing than non-altruistic partners. Chapter one incorporates altruism into the model of informal risk sharing and shows with the help of numerical simulations that this result does not hold for incomplete risk sharing arrangements. This suggests that the altruistic family can more easily provide consumption insurance. Chapter two relaxes the standard assumption that income can not be stored and is the first to model the alternatives of precautionary saving and informal risk sharing. By using a functional form that allows for a closed form solution, it can be shown that precautionary saving is the preferred mechanism for consumption smoothing in the absence of borrowing constraints. The latter appear to be the most important factor in explaining the institution of informal risk sharing. Some of the hypotheses suggested in the theoretical chapters are tested in chapter three, using panel data of the Russian Longitudinal Monitoring Survey of 1995. A measure of the income shock is constructed for each family, and used to estimate the responsiveness of gifts, loans and withdrawals from wealth to the magnitude of the shock. It is found that risk sharing is more important than self-insurance. Risk sharing happens mainly through loans. These loans are more responsive to unpaid wages than to unemployment or unpaid pensions. There is also evidence that loans, and in particular gifts, are exchanged within the extended family, with younger families being the prime beneficiaries. Finally, the data indicate that smaller shocks are better insured against by the social network than large income shortfalls.