Does the earned income tax credit encourage or discourage work effort?: an empirical study

The purpose of this dissertation is to examine the effectiveness of the earned income tax credit (EITC) in meeting the objectives that Congress intended. The refundable EITC was first enacted in 1975 to reduce the burden of the Social Security tax for the working poor, to increase the progressivity of the tax system overall, and to increase the incentives of low-income individuals to work. A better understanding of the EITC is important because the credit is becoming the major government support for low income families, and the total expenditure on EITCs are expected to exceed total government spending on welfare.

This study utilizes empirical analysis of actual individual income tax return data to evaluate EITC participants' and comparable nonparticipants' income over time for indications of statistically significant increases. The incomes of participants and nonparticipants are also compared cross-sectionally for evidence of a statistical difference over time. Observations are grouped into income classifications based on the income ranges of the credit: phase-in, plateau, and phase-out. Economic theory predicts a potential work incentive in the phase-in range and a potential work disincentive in the phase-out range.

EITC participants' incomes are found to have increased significantly over the selected time windows during the 1979 to 1990 time period. The incomes for a comparable group of nonparticipants also increased significantly over the same time windows. Income growth for both participants and nonparticipants was greater in the phase-in range than in the phase-out range. There is generally no significant difference in the income growth between participants and nonparticipants until the time windows for 1987 to 1988 and for 1987 to 1990. In those periods, there is indication of a statistically significant association of the EITC with higher income growth for participants in the phase-in range. The differences in the income growth between participants and nonparticipants in the phase-out range is not statistically significant for any of the time periods. An association of the credit with a disincentive in the phase-out range cannot be inferred.