Earnings Responses to Increases in Payroll Taxes

This paper uses SIPP data matched to longitudinal uncapped earnings records from the Social Security Administration for 1981 to 1999 in order to analyze earnings responses to increases in tax rates and inform discussions about the likely effects of raising the Social Security taxable maximum. The earnings distribution of workers around the current taxable maximum is inconsistent with an annual model in which people are highly responsive to the payroll tax rate, even in the subset of self-employed individuals. Panel data on married men with high earnings display a tremendous increase in earnings over the 1980s and 1990s relative to other groups, with no clear breaks around the key tax reforms. This suggests that other income groups cannot serve as a control group for the high earners. Our analysis does not support the finding of a large behavioral response to taxation by wives of high earners. We actually find a decrease in the labor supply of wives of high earners around both the 1986 and the 1993 tax reforms, which we attribute to an income effect due to the surge in primary earnings at the top. Policy simulations suggest that with an earnings elasticity of 0.5, lost income tax revenue and increased deadweight loss would swamp any benefits from the increase in payroll tax revenue. In contrast, with an elasticity of 0.2, the ratio of the gain in OASDI revenue to lost income tax revenue and deadweight loss would be much greater. This research was supported by the U.S. Social Security Administration through grant #10-P-98363-1-01 to the National Bureau of Economic Research as part of the SSA Retirement Research Consortium. The research in this paper was conducted while Liebman was a Special Sworn Status researcher of the U.S. Census Bureau at the Boston Census Research Data Center (BRDC). This paper has been screened to insure that no confidential data are revealed. The findings and conclusions expressed are solely those of the authors and do not necessarily represent the views of SSA, the Census Bureau, any other agency of the Federal Government, or the NBER. Support for the Census research data centers from the NSF (award no. ITR-0427889) is gratefully acknowledged. We thank Erin Metcalf, Daniel Ramsey, and Sarena Goodman for excellent research assistance. We thank Leora Friedberg for helpful comments and discussions.