The Earned Income Tax Credit, Expenditures and Labor Supply

The Earned Income Tax Credit (EITC) is a refundable tax credit targeted toward the working poor. In the wake of welfare reform, the EITC program expanded dramatically to become the largest federal cash transfer to low-income families in the United States. Unlike traditional welfare, the EITC requires labor force participation. In this dissertation, I explore the impact of the EITC on expenditures and labor supply. The goal is to provide a more complete picture of how this requirement impacts eligible families.;In the first chapter, I use variation in the EITC schedule generated by federal legislative expansions between 1984 and 1998 to identify the causal impact of the EITC on expenditures. I estimate the expenditure response using a sample of women with and without children who have not attended post-secondary school within the Consumer Expenditure Survey (CE). I summarize multiple expansions of the EITC schedule across time and family structure using a ''simulated'' credit that captures the complexity of the EITC schedule without depending on endogenous changes in other outcomes (such as earnings or income). The CE allows for the estimation of a differential response to the EITC within the calendar year because it is administered at the quarterly level. Finally, I use the Current Population Survey (CPS) to estimate the impact of the EITC on after tax and transfer income and its components. These estimates inform the mechanism behind expenditure responses and can be used to rescale the expenditure estimates in a two-sample two-stage least squares framework, so that they are more easily interpretable by policy makers. Assuming the expenditure response to the EITC works entirely through after tax and transfer income, this final estimate may be interpreted as the expenditure response due to a one dollar change in income.;I find that total expenditures increase with the EITC. The impact on non-durable expenditures is particularly strong. This result is not consistent with the previous literature exploring the relationship between the EITC and expenditures.;In the second chapter, written with Hilary Hoynes, we use variation in the credit induced by legislative changes across tax year, family size and state of residence. We construct a ''simulated'' version of the EITC at the federal and state level to summarize these changes. This measure captures changes in the EITC across many parameters, allowing us to fully exploit variation caused by legislative changes in policy. The estimation is divided into two major parts. First, we estimate the impact of the federal credit on labor force participation. We narrow the identifying variation in ways that are relevant to policy and the literature. We use non-parametric differences-in-differences to visually assess the impact of policy changes on labor force participation while checking for pre-existing trends. Second, we estimate the impact of state-level credits on labor force participation. State variation in EITC generosity allows for number of children by year fixed effects, powerful controls which cannot be used in federal EITC policy estimations. State-level EITCs are smaller than the federal EITC, however, making precise estimation more challenging. Finally, we rescale estimates so that they can be easily compared across specifications and used to draw lessons for policy.;We find that the federal EITC has increased the labor force participation of single women with children, even when we include detailed controls for other tax and transfer programs. Using an entirely different source of identifying variation, we find that state EITCs increase the labor force participation of single women with children. Although the estimates at the state level are noisy, after we rescale the point estimates in terms of eligible benefits, the impact from the federal EITC is roughly comparable to the impact at the state level.;In the final chapter, I extend the analysis of the EITC's relationship with family expenditures by contrasting the results from Patel (2012a) with estimates identified using variation in after tax income caused by the Economic Stimulus Payments of 2008 (ESP). The advantage of using this payment is that the timing of disbursement has been randomized. By comparing expenditures of households that have received an ESP in a given period to those who receive an ESP in another period, I estimate a causal impact of a change in after tax and transfer income without using variation correlated with changes in labor supply, earnings or income.;The results are imprecise. However, point estimates suggest that total family expenditures increase upon ESP receipt by a magnitude that is similar to the EITC expenditure response found in Patel (2012a). The expenditure response is evenly split between durables and non-durables, with non-durable expenditures concentrated in food and recreational expenditures. Unlike the EITC driven response, work-related expenditures do not change with the ESP. (Abstract shortened by UMI.)...