DeBacker et al. 2013 — Rising Inequality Transitory or Persistent New Evidence from a Panel of U.S. Tax Returns

Summary

Using a confidential Internal Revenue Service (IRS) panel of individual tax returns (1987–2009), DeBacker et al. (2013) decompose the rise in U.S. income inequality into permanent and transitory components using an error-components model. They find that the entire increase in male earnings inequality is attributable to a rise in the permanent variance, with zero increase in the transitory variance. For household income, roughly 75% of the increase is permanent and 25% transitory, with the transitory component traced to spousal earnings and investment income. The federal tax system reduces inequality levels by about 15% but does not attenuate the trend.

Key Claims

• Male earnings: 100% permanent. The variance of permanent male earnings shocks rose substantially over 1987–2009; the variance of transitory shocks did not increase at all.
• Household income: ~75% permanent. For broader household income, roughly 75% of the inequality increase is in the permanent component; the remaining ~25% transitory increase is concentrated in spousal earnings and capital income.
• First use of admin data at household level. Prior error-components studies (Moffitt-Gottschalk 1995, Haider 2001) used survey data; this is the first to apply the framework to U.S. administrative tax records and the first to extend it to the household unit.
• Tax system doesn't offset the trend. The federal income and payroll tax system reduces the level of inequality by ~15% but leaves the rising trend essentially unchanged.
• DI implications. Permanently rising inequality raises the disability insurance (DI) earnings replacement rate mechanically for low-wage workers, providing direct empirical grounding for the Autor-Duggan (2003, 2006) channel: it is not a temporary fluctuation but a durable structural shift in the wage distribution.

Concepts Introduced or Extended

• Income Dynamics
• DI Replacement Rate
• DI Growth Decomposition

Entities Mentioned

• Jason DeBacker
• Bradley Heim
• Vasia Panousi
• Shanthi Ramnath
• Ivan Vidangos

Quotes

"The entire increase in cross-sectional dispersion of male individual earnings is due to a rise in the permanent variance, with no change in the transitory variance."

"For household income, about 75 percent of the increase in inequality is associated with an increase in the permanent variance, with the remaining 25 percent due to an increase in transitory variance."

"The federal tax system reduces the level of income inequality, by about 15 percent, but does not affect the trend."

My Take

This is a methodologically clean paper that settles a key empirical question: the rise in earnings inequality is not noise that will revert — it is a permanent structural shift. The result matters enormously for the DI literature because the Autor-Duggan replacement rate mechanism (low-skill wages falling while the Social Security Administration's national wage index rose, mechanically raising DI replacement rates) requires that the wage gap be permanent to have lasting effects on application incentives. DeBacker et al. (2013) provide precisely that empirical foundation. The household income result is more nuanced — 25% transitory variation suggests some volatility — but the dominant permanent component survives. The main caveat is that IRS data miss informal income and underrepresent the very bottom of the earnings distribution, where the DI-relevant workers are concentrated.