The EITC is now the single most important public insurance program in place in the US. It provides wage-subsidies to households that are sharply dependent on their demographic status, especially the number of children present in the household. In addition to productivity risk, it is the case that early in life, neither future marital status, nor the number of dependents, is known with certainty. Therefore, an important dimension of the EITC to act as insurance against such risk, particularly the state in which one experiences marital separation at any time after generating dependent children. Given the fundamental nature of EITC as a public insurance scheme, there are, however, potential effects on incentives. To avoid "abuse" of the EITC, the program must impose limits on eligibility. Providing insurance requires transfers to those doing badly, but the second necessitates having a "phase-out" range for eligibility. The phase out however means that the marginal reward to work may have to fall, sometimes sharply. We ask the following questions. Who benefits from the EITC, and to what extent are beneficiaries recipients of pure transfers relative to pure insurance? What the temporal distribution of benefits? Lastly, how distortionary is the EITC likely to be? For the latter, an important innovation of our paper is to utilize a key insight from work on dynamic models of labor supply under uninsurable risk, especially that of Floden and Linde (2001), and Pijoan-Mas (2006). These papers strongly suggest that the labor supply for the target population of the EITC will be relatively low. Our study, to our knowledge, is the first to measure the implications of the EITC in a setting capable of accommodating the essential features governing its impact. These are (i) a well-defined dynamic setting in which productivity varies with both age, and with uninsurable shocks (ii) liquidity constraints, and (iii) demographic risk. Preliminary results suggest that the EITC provides substantial insurance to young US households, and does not significantly alter, and hence does significantly distort, labor-leisure choices.
