The Optimal Design of the Earned Income Tax Credit

Abstract

This paper uses a microsimulation model calibrated to microdata from the 1999 Current Population Survey (CPS) to illustrate the tradeoffs that arise in designing an Earned Income Tax Credit, focusing in particular on the choice of a maximum credit, a phase-in rate, and a phase-out rate. The paper has two basic results. First, in contrast to the findings of Browning (1995), I find that the overall efficiency cost of transferring income through the EITC is fairly low. While Browning found that it cost $4.03 to provide one dollar’s worth of utility gain to recipients in the phase-out region and almost $2.00 to provide one dollar’s worth of utility to recipients in the phase-in and constant region, I find that it typically costs less than $2.00. An important reason for my lower results is that I take into account the impact of the EITC on the labor force participation of single parents. When the savings from reduced welfare spending are taken into account, the EITC appears to be a much more efficient way to accomplish income redistribution than when they are ignored. In addition, a significant portion of the labor supply responsiveness of married couples is likely to be due to secondary earners leaving the labor force. My simulations indicate that much of this responsiveness is due to the income effect of the EITC rather than the substitution effect. Therefore, for these households, the EITC is close to a lumpsum transfer. Second, the optimal structure of the EITC depends heavily on society’s taste for redistribution, something that is difficult if not impossible to observe. For the current structure of the EITC to be optimal, taxpayers must value one dollar’s worth of utility gain for an EITC recipient receiving the maximum credit at roughly 2.5 times a dollar in the hands of the typical non-EITC taxpayer. If society places a relative welfare weight on EITC recipients of 3 or more, then the credit should be expanded substantially. In contrast, if society places a relative weight below 2 on EITC recipients, then the EITC should be reduced or eliminated.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 40 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.