Sharing wage risk

Intro: risk sharing Our paper introduces three innovations with respect to the standard literature. First, labor supply is explicitly recognized and modeled as an endogenous variable, that responds to exogenous shocks. Actually, one of the main topic of the paper is precisely to investigate how adverse income shocks trigger changes in labor supply at the household, and possibly at the risk sharing group level. Second, we consider variations in non labor income, but also in wage. While price uncertainty arguably plays an important role in real life (if only because wage variations are a crucial component of income shocks), not much is known about optimal risk sharing in this context. In the paper, we provide an exhaustive, theoretical characterization of effi cient risk sharing contracts in a general context of uncertainty on wages and incomes; as well as an empirical implementation. Finally, our setting is fairly general; in particular, it allows for different levels of risk aversion, both within and between households. We offer a detailed discussion of identification issues in this framework; we show, in particular, that preferences and the decision process are non parametrically identifiable,1 and we show how long panels can be used in practice to achieve identification. Related literature Outline ∗Townsend acknowledges support from the NICHD-NSF. 1For the distinction between identification and identifiability, see Chiappori and Ekeland (2008).