One dilemma for policymakers is how to get people to take advantage of social welfare programs. In the case of the Earned Income Tax Credit (“EITC”), the goal is to encourage eligible individuals to claim the credit on their tax return. Take-up rates for the EITC are quite good (about 80% overall), but ideally would be higher. Typically the approach to increasing EITC take-up is information campaigns, like EITC awareness day. The conventional wisdom has been that the more people know about the EITC, the more likely eligible recipients are to claim it. But is it right that advance notice is important? If people use tax software that will automatically calculate the EITC for them, how important is it that they are made aware of the benefit ahead of time? Perhaps not very, as suggested by Jacob Goldin in his forthcoming article, Tax Benefit Complexity and Take-up: Lessons from the Earned Income Tax Credit.
The key insight from Goldin’s article is that in the modern age, virtually anyone who files a tax return is presented with the opportunity to claim the EITC. This is because the vast majority of taxpayers—96 percent in 2015 according to Goldin—use assisted preparation methods (“APMs”) such as self-preparation software or a tax return preparer. Using either of those methods, it is extremely unlikely to fail to claim the credit accidentally. (Though, as Goldin notes, some taxpayers may consciously choose not to claim the credit even though they are eligible.) The paper’s main conclusion is logical yet important: people who are eligible for the EITC but who fail to claim it are generally people who fail to file returns at all. Thus, if policymakers want to increase EITC take-up, they must increase the filing rate.
Goldin frames the article by analyzing the complexity of the EITC. In doing so, he introduces a novel and important taxonomy into the tax literature that is useful beyond the specific context of the EITC. Goldin’s taxonomy distinguishes between “informational complexity” involved in claiming tax benefits and “computational complexity.” Informational complexity involves the level of difficulty in compiling the inputs for claiming a benefit, both in terms of how much information is required and how hard it is to compile the information. For example, the EITC requires information about age, income, marital status, dependents, and several other inputs. On the other hand, computational complexity describes the difficulty in actually calculating the taxpayer’s eligibility for the credit and determining the amount of the credit. For example, determining whether a minor is a qualifying child, and applying the credit’s phase-in or phase-out rules, are both part of the EITC’s computational complexity.
Tax benefits might have high levels of one type of complexity and low levels of another. For example, as explained by Goldin, the charitable deduction is easy to calculate (low computational complexity), but may have high informational complexity because it requires taxpayers to keep good records of what donations they have made during the year. The EITC, Goldin argues, is low in informational complexity because the information required is easily knowable by taxpayers (e.g., age and marital status), easy to obtain if not already known, and is often already required for other purposes on the tax return. The computational complexity of the EITC, on the other hand, is quite high.
The EITC’s classification as low informational complexity/high computational complexity is important because computational complexity is easily overcome by APMs. Software used either on a self-prepared return or by a tax return preparer will calculate the amount of the EITC in seconds once the required inputs are provided. Thus, Goldin concludes, computational complexity should not be a barrier to take-up for anyone filing a return with an APM (i.e., the vast majority of people filing returns). Similarly, informational complexity should not be a barrier because, although APMs do not eliminate informational complexity, informational complexity is low for the EITC.
Important policy prescriptions flow from this analysis. First, Goldin argues, efforts to increase EITC take-up should focus on increasing the filing rate among taxpayers who are eligible to claim the EITC and who do not otherwise file. Relatedly, efforts to raise EITC awareness are probably ineffective at increasing take-up, except to the extent those efforts encourage return filing among non-filers. Further, Goldin argues, current EITC awareness campaigns are unlikely to increase the filing rate among eligible non-filers because awareness of the credit does not help taxpayers understand if they would be eligible and what the amount of their refund would be. This is consistent with research showing that mandatory employer notices of the EITC are not linked to higher take-up rates.
How should we encourage EITC-eligible non-filers to file? Goldin’s article offers a number of creative proposals here aimed at increasing the actual or perceived benefit of filing. For example, one approach would be to alter the withholding tables so that taxpayers eligible for the EITC would be owed a bigger refund at the end of year, thus increasing the economic benefit of filing. One potential drawback to this approach, though, is that EITC-eligible taxpayers are more likely to experience liquidity problems during the year from over-withholding. Another possibility would be to limit government programs that offset tax refunds (e.g., programs that automatically divert tax refunds towards things like child support obligations), since these programs likely deter eligible claimants from filing a return. However, as acknowledged by Goldin, limiting these programs would undermine the policy goals of those programs (e.g., making sure child support obligations are paid).
Yet another approach would be to increase awareness of free preparation assistance (e.g., VITA programs and the IRS’s FreeFile program). Goldin acknowledges that these free programs have some drawbacks and programs like VITA may need to be expanded if the number of low-income filers significantly increased. Other forms of information may also encourage more eligible taxpayers to file and claim the EITC. For example, if taxpayers currently perceive the monetary benefit from filing to be too low, providing taxpayers with more personalized estimates of their potential refund amount may increase filing rates. The unifying theme here is that these informational campaigns would be aimed at increasing awareness of the benefits of filing, as opposed to awareness of the credit itself.
Increasing EITC take-up by increasing awareness of the credit itself made sense in previous decades. As Goldin points out, about one-third of taxpayers self-prepared their returns in 1998, making it more likely that eligible filers would fail to claim the EITC on their returns. But due in large part to the prevalence of tax preparation software, the number of taxpayers filing without some preparation assistance is approaching zero. As the tax return preparation landscape changes, so must our approach to encouraging eligible individuals to claim social welfare benefits through the tax system. Goldin’s article, and his focus on taxpayers’ filing rate, provides important guidance to the IRS and policymakers in this regard.