The Elasticity of Taxable Income: Influences on Economic Efficiency and Tax Revenues, and Implications for Tax Policy
Date of this Version
While research into the elasticity of taxable income (ETr), which measures the responsiveness of reported taxable income to changes in tax rates, dates back to at least Lindsey (1987), recognition of its importance as a central parameter for tax policy design did not begin to take hold until the second half of the 1990s. In fact, a 1998 survey to determine public and labor economists' views on key policy parameters (Fuchs, Krueger, and Poterba 1998) included no questions on the ETI. I suspect that a 2008 survey would include such questions, just as I suspect that a 1998 conference entitled "Tax Policy Lessons from the 1990s" would have no session on the elasticity of taxable income. The two 1998 survey questions most likely to provide some insight into the views public economists then held of the ETI asked about the effect of the Tax Reform Act of 1986 (TRA86) and the Omnibus Budget Reconciliation Act of 1993 (OBRA93) on long-run (steady-state) gross domestic product (GDP). For TRA86 , a fundamental reform that broadened the tax base and substantially lowered marginal tax rates, the median response was that steady-state GDP would rise by 1 percent. However, the interquartile range was large, from 0.20 to 3 percent of GDP For OBRA93, which raised marginal tax rates for primarily upperincome groups, the median response was zero, with an interquartile range from -0.5 to 1 percent of GDP It is noteworthy that half of public economists surveyed thought that raising marginal tax rates for the highestincome groups (in 1993) would not result in decreased steady-state GDP.
Disagreement among public economists as to the effect of taxes on the economy is embodied by the views of two former chairmen of the president's Council of Economic Advisors (CEA). One former chairman, Martin Feldstein (l995b, 1999), estimated that the 1993 tax increases substantially increased deadweight loss (DWL) and that repealing the rate increases could actually increase tax revenue because positive behavioral responses would more than offset the mechanical revenue loss-that is, the loss in tax revenue absent any behavioral responses. Another former CEA chairman, Joseph Stiglitz (2004), viewed the 1993 tax increases in a quite different light: 'The Clinton experience showed that raising taxes on the rich does not have the adverse effects that the critics claimed" (4). Additionally, Stiglitz is very critical of the Bush tax cuts, while Feldstein supports the lower marginal tax rates. One could argue that the two former CEA chairmen take such different positions on recent tax policy because of differing political ideologies or party allegiance. However, a more plausible explanation is that they hold very different views of how responsive individuals are to changes in tax rates. Feldstein's estimates for the effects of repealing OBRA93 , for example, rest on an ETI estimate that is toward the high end of the literature-although not implausible. Stiglitz, on the other hand, while not directly speaking to the ETI, believes that behavioral responses to tax rates are small (at least for high-income individuals). If the ETI is very small, then the revenue and efficiency implications from repealing OBRA93 would be quite different from those estimated by Feldstein.
Published in TAX POLICY LESSONS FROM THE 2000s, ed. Alan D. Viard (Washington, DC: AEI Press, 2009).