EHP: Tax Policy and Low-Wage Labor Markets: New Work on Employment, Effectiveness and Administration

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

Part 1 of this project will provide new information on the employment effects of tax policies targeted to low-income families. Part 2 will produce new information on the relative effectiveness of welfare policy choices, tax policy, and economic growth in enhancing the incomes of poor families. And part 3 will develop new information on the degree to which administrative problems arise with low income tax provisions. Many previous studies find the EITC has large, positive employment effects. But much of this work uses indirect identification strategies and it has not been able to verify with tax data that the EITC causes the observed employment patterns. We examine the employment effects of the EITC using a sample of California adults who received AFDC or TANF sometime between 1987 and 2000. Our proposed work will be the first study to use information on whether sample members actually filed a tax return and claimed the EITC. Using an identification strategy similar to that used by earlier studies, the PIs document employment patterns across groups, both in the raw data and in regression-adjusted estimates, that are consistent with significant, positive EITC employment effects. Tests using tax data, however, raise questions about whether the EITC employment effects reported by others are in fact caused by the credit. The PIs will also carry out additional work to examine the validity of the preliminary results and help reconcile our work with past EITC employment studies. Employment rates of low-skilled single mothers increased sharply in the 1990s and welfare caseloads declined precipitously. Considerable uncertainty remains about the relative importance of factors accounting for these changes. This project is a systematic analysis of the relative importance of welfare reform, the EITC, and the role of labor demand in understanding patterns of employment and welfare use. Four features make this proposed work promising. Tax data improve the identification of EITC effects. The PIs use more geographically precise measures of labor market characteristics than other papers. And they use much more detailed characterizations of the "treatments" of state (and county) welfare programs than other papers. Finally, samples exceeding 3 million adults allow us to examine interactions in the way welfare and tax policy influence employment and welfare use in different labor market. Part 3 of the project examines the extent to which families respond to welfare program and tax system incentives to inaccurately report their income or household circumstances. The PIs use statistical models to study income reporting to welfare and tax authorities. They will also complete a unique study that examines misreporting of family structure, the largest single source of EITC non-compliance. The work will directly inform policymaking. Our partners in this research include the Internal Revenue Service, California Franchise Tax Board, and the Wisconsin Department of Revenue. They are making data available to (under strict confidentiality agreements) and underwriting much of the cost of data construction and processing, because they are convinced the research promises to improve the tax policy effectiveness and administration. Second, this work has required a substantial investment in data infrastructure. The PI team will continue to make data available to other researchers, except in cases where confidentiality restrictions legally prevent them from doing so.