Evaluating the Growth Effect of the Earned Income Credit
August 9, 2013
Jared Bernstein, who is currently a Senior Fellow at the Center on Budget and Policy Priorities and was Chief Economist and Economic Adviser to Vice President Biden from 2009 to 2011, has criticized a new Tax Foundation study regarding the Earned Income Credit (EIC). Bernstein's blog post is here and the Tax Foundation paper is here.
The Tax Foundation study concluded that while the EIC raises the incomes of low-income workers, its net result is to reduce both national output and total hours worked. This result may seem surprising because the credit creates a strong incentive for workers with very low incomes who are within the EIC's phase-in range to work more since each extra dollar of earnings brings a larger credit. Unfortunately, for the larger number of low-to-middle income workers who are within the EIC's phase-out zone, the loss of benefits with rising earnings generates a powerful deterrent against additional work effort. (The phase-in is sometimes described as generating a negative marginal tax rate, in effect, and the clawback as producing a marginal rate spike.) According to the Tax Foundation's dynamic simulation model, the clawback reduces work hours and output by more than the phase-in increases them. The number of taxpayers in the phase-in and phase-out ranges is an empirical question, and the Tax Foundation model answers that question with data from the IRS's Public Use File, a large, anonymous (individual taxpayers can't be identified) sample of tax records.
Mr. Bernstein correctly notes that both the phase-in and the clawback are important, but he then incorrectly claims the Tax Foundation only examines the clawback:
How then, does the TF come up with the claim that it [the EIC] discourages work? That's because they're emphasizing the "substitution effect" in the phase out range of the credit. But of course the question is the net impact-the impact on work incentives across the full program, not just one piece of it (you don't describe the results of a ball game by just telling us how many runs your team scored).
Both sides are indeed important, but, contrary to Bernstein's assertion, the Tax Foundation model does consider the two pieces: the EIC's incentive effects at its bottom (a big reward for additional work) and the disincentive effects at its top (a stiff penalty for additional work.)
To confirm that the model properly handles the phase-in, we reran it keeping the EIC phase-in as is but eliminating the phase-out—that is, all workers would receive the credit—while making the implicit, and probably naïve, assumption that the government cuts enough low-value federal spending elsewhere to avoid increasing the deficit. The model's outcome in that hypothetical case is higher output and more hours worked. The killer from a growth perspective is the current-law clawback.
The Tax Foundation study did explicitly acknowledge one limitation in its estimate, and it is fair to repeat it here (quoting from the original case study). “[S]everal studies have found that the EITC encourages some people to enter the work force who otherwise would not at a rate greater than this model assumes.” The original study also mentioned that the work incentives frequently claimed for the EIC may be overstated because about one-quarter of benefits are believed to be fraudulent or otherwise improper. (See testimony of the Department of the Treasury's Inspector General for Tax Administration here.)
It might be added that just as it would be incomplete to look solely at the phase in, so it would be one-sided and misleading to disregard the phase-out's penalty on additional income when judging the EIC's growth effects. Some of those who assume the EIC must surely be increasing work effort may be committing that error.
The study also estimated the growth and revenue effects if the revenue from repealing the EIC were used to finance an across-the-board individual income tax rate cut of 5.7 percent: $125 billion of added GDP, 826,000 more full-time-equivalent jobs, and a net federal revenue gain of $29 billion. The marginal rate cut would ease tax penalties on work and investment, accounting for the growth.
In a series of case studies, the Tax Foundation has also used its dynamic simulation model to estimate the growth and revenue effects of ending 10 other major tax expenditures. In each case, the Tax Foundation further estimated the growth consequences if the money from repeal were used to finance an across-the-board cut in individual income tax rates. The EIC and other studies are motivated by the “Blank Slate” proposal of Senators Max Baucus (D-MT) and Orrin Hatch (R-UT), the Chairman and Ranking Member of the Senate Finance Committee. The Senators believe that perhaps the way forward on tax reform is “to start with a 'blank slate'—that is, a tax code without all of the special provisions in the form of exclusions, deductions and credits and other preferences that some refer to as 'tax expenditures.'” They would restore only those provisions for which there is “clear evidence” regarding their desirability. The Senators add that much or all of the money from removing various tax expenditures could be channeled into lower tax rates.
The Tax Foundation takes no position on the soundness of the bipartisan “Blank Slate” approach, but has undertaken these studies in the belief that objective information on positive or negative growth and revenue effects would be helpful to policymakers. The Tax Foundation's estimates may be especially valuable because conventional revenue estimates usually rely on the simplifying, but often unrealistic, assumption that tax changes have no impact whatsoever on economic aggregates like total investment, employment, productivity, and output. In reality, when a tax change alters work and investment incentives, it can have a major impact on economic growth and, in turn, the economic feedback can have a substantial effect on federal revenue. Conventional, static revenue estimates can be seriously misleading when growth effects are large.
Of course, growth is not the only factor to consider when evaluating an existing or proposed tax feature, and is often not the main factor, but growth is important and should not be ignored.
Moreover, if it is decided to retain a tax provision, a growth analysis will often be useful in helping to understand how to better structure the provision. For example, one of the original motivations of the EIC was to compensate those with very low incomes for the payroll tax, which begins with the first dollar of income. The concern was that the payroll tax reduces the after-tax incomes of the poorest workers and decreases their incentive to work. An approach that would meet that objective directly would be to establish a zero bracket amount for the Social Security tax: earnings below that amount would not be subject to the payroll tax. Provided the zero bracket amount is not clawed back, it would have a positive growth effect at the bottom of the income scale and no negative effect for people whose earnings are lifting them into the middle class.