Yellen Supports Bank Reporting Proposal

Sincerely,

Janet L. Yellen

Enclosure

---

September 14, 2021

MEMORANDUM

FROM:
Mark J. Mazur, Assistant Secretary for Tax Policy (Acting)

SUBJECT:
The Revenue Impacts of Compliance Proposals

This memo summarizes the Treasury analyses estimating the effects of tax compliance proposals, incorporating the Office of Tax Analysis (OTA) estimates as well as other considerations that would have predictable effects on the revenue potential of possible Congressional compliance proposals. Together, these factors imply that tax compliance proposals along the lines of those being contemplated by Congress have the potential to raise around $700 billion over the ten-year budget window and much larger amounts in subsequent years.

In developing the FY 2022 Budget proposals, OTA (working with Internal Revenue Service staff) analyzed the administration's tax compliance proposals using methods similar to those used previously by OTA and IRS (and, as you know, these methodologies are similar to those used by the Congressional Joint Committee on Taxation (JCT) and the Congressional Budget Office (CBO)). However, the Administration proposals were larger in scale than recent proposals and historical experience and therefore the estimating approaches required some adaptation.

Our current understanding is that Congress is considering the adoption of financial reporting proposals that are likely to be narrower in scale and scope than those proposed by the Administration. Clearly, this will lower the estimated revenue raised from the proposed reporting regime relative to earlier Administration estimates.

Conventional revenue estimating approaches incorporate the behavioral responses to legislative changes. For instance, if tax rates change on some economic activities, taxpayers respond by reducing those activities. We see similar results in the information reporting area, where income sources with detailed third-party reporting result in taxpayers reporting 90 percent or more of this income, while income sources without robust third-party reporting generate reporting rates of 50 percent or even lower. In this case, improved information reporting leads taxpayers to increase their voluntary reporting of previously under-reported income sources.

However, when analyzing proposals that provide increased enforcement resources to the IRS, the conventional approach is to merely estimate the increased revenue collected by the IRS through additional audits and additional collection activities, but not to ascribe any changed behavior on the part of taxpayers who potentially would be subject to added enforcement scrutiny. This convention leads to an underestimate of future revenues generated by the proposal.

The revenue potential for the two aspects of the compliance agenda can be analyzed as follows:

IRS Funding Proposals

• Estimates for the effect of IRS resource changes on increased enforcement revenue (revenues associated directly with increased audit or collection activity) come directly from the IRS. IRS estimates the return on investment (ROI) for most of its enforcement activity based on historical enforcement data. The IRS collects this information on an ongoing basis.

  • The expected return is estimated as the increase in the direct enforcement tax revenue, or revenue generated by an examination or a collection activity (or both). Importantly, these estimates do not incorporate a potential deterrence effect from having more IRS personnel undertaking enforcement action in future years.

  • IRS does not provide any ROI estimate associated with non-enforcement functions, such as IT improvements or enhanced taxpayer services, because these activities do not by themselves generate enforcement revenue. However, these investments in services and IT will positively impact tax administration.

• The estimated revenue effects for proposals to increase IRS funding proposals are based on the IRS ROI estimates. In general, new hires are expected to take several years to reach the full ROI potential due to the time needed for hiring and training. IRS also assumes a declining marginal return for its enforcement activity as the level of enforcement function increases.

• The estimated revenues are clearly backloaded. Under the current funding proposal, investments in IRS personnel and fixed assets would be phased-in over a 9-year period, and then remain at that level (in real terms) for the rest of the budget window (and on into the future).

• The estimates for the direct effect of $80 billion in additional resources derived from the IRS's revenue estimation model for budget initiatives is that it raises about $320 billion in additional tax collection over the ten-year budget window. This estimation reflects the additional revenue from enforcement activities; but not the increase in compliance that will result from investment in taxpayer services and IT modernization.

  • This is because the IRS does not calculate a ROI in terms of enforcement revenue for these investments, and as a consequence, neither Treasury nor CBO account for the effects of such spending on revenues. However, there is a large scholarly literature that supports beneficial effects on compliance from investments in service and technology.¹

• The conventional revenue estimate described above does not include any deterrent effect of the increased IRS resources. Just like have a police officer on the median of an interstate will result in fewer cars excessively speeding, having more IRS agents able to audit potential tax evaders should increase voluntary compliance with the tax system, or deter tax evasion. Recent academic work speaks to the importance of deterrent effects. DeBacker et al (2018) find that increased income reported in the 5-8 years following a random individual audit is about 1.5 times the revenue generated by the audit itself.² Studies in the U.K. and Denmark find similar effects.³ Boning et al. (2020) find that in-person collection visits raise as much revenue from firms sharing a tax preparer with the visited firm as from the visited firm itself.⁴

• Using the results from the more conservative of these estimates suggests that the deterrent effect from additional IRS resources is about as large as the direct effect of additional revenue collected from enforcement activity. That means that the $320 billion of enforcement revenue generated over a decade from the additional investment in IRS resources would lead to about $640 billion in additional revenue over the budget window if the deterrent effect is incorporated into the revenue estimate (leading to $560 billion over 10 years net of the $80 billion investment).

• The deterrent effect associated with different types of enforcement activities varies, and these effects can be challenging to identify. For example, evidence for the follow-on effects of corporate audits is mixed; as a result, we suggest only including a 50 percent boost for deterrent effects. Conservatively, about $400 billion of additional revenue can be collected (incorporating both the direct and indirect effects of enforcement investment) from the President's proposals, net of the $80 billion investment.

Information Reporting and Interaction Effects of the Proposals

• The Treasury's revenue estimation for a financial information reporting regime begins with an estimate of the tax gap for business income and employment taxes. This includes Schedule C, E, and F income for individual income tax, corporate income tax, and employment taxes (SECA, NIIT, etc.).

• The amounts of business income reporting are estimated using random audit data from the IRS' National Research Program (NRP) for the individual income tax. The additional pass-through entity gap relies on the preliminary results of the NRP S-corporation study. For small corporations, the tax gap comes from IRS tax gap estimates.

• OTA converts the estimated tax gap for business income and employment taxes into a revenue estimate by assuming increased voluntary compliance as taxpayers react to the increased information reporting. This is assumed to be a gradual increase in voluntary compliance that phases in over time.

• OTA estimated the revenue effects of the Administration's financial reporting proposal using these basic steps. That proposal included information on inflows and outflows for a wide range of financial accounts, additional account information including additional information on business accounts, and an annual account threshold of $600.

• We do not know the exact parameters of Congressional proposals, but we expect that they are likely to incorporate a narrower reporting regime than was the basis of the Administration's information reporting proposal scored by Treasury. Although we have not scored any particular proposal, we suspect the revenue potential of a narrower reporting regime would be lower, perhaps in the range of $200-250 billion over the ten-year budget window. The estimated path of revenues for this proposal is backloaded, and expected revenue generated in the decade after the budget window would be larger than the revenue estimate for the initial ten-year period. This is because information reporting proposals take some time to implement and for the IRS to determine how best to deploy this new information. Moreover, the reactions of taxpayers to this increased visibility into business income also will lead to gradually increase voluntary reporting over time.

• Neither of the estimates of the revenue generation potential for the compliance proposals (increased resources and increased information reporting) reflects the interaction effects of these two major changes to the tax compliance environment. Additional information reporting is valuable to address the tax gap not just because of an increase in voluntary compliance (which is reflected in the OTA scoring approach described above), but also because additional information can help the IRS better target enforcement activities. With the resources to meaningfully pursue tax evasion that is identified through improved information reporting, there will be a meaningful direct effect on additional tax collections. In addition, the IRS investments in IT and enhanced taxpayer services will also increase the efficacy of IRS enforcement resources. While there is not an official estimate of the size of these interaction effects, due to the size of the compliance initiatives outlined here, it is reasonable to expect that the interaction effects would serve to increase revenues by at least $50 billion over the 10-year budget window, with an increase of at least that size in the subsequent decade.

Overhauling the IRS by giving it the information and resources that it needs to pursue tax evasion will create a more equitable tax system. The pieces of this proposal work well together: additional information for the IRS is crucial to its ability to efficiently and effectively select enforcement actions; and additional resources are essential to help the IRS use the information at its disposal. To truly overhaul tax administration, both components of the plan are essential.

The factors detailed above imply that tax compliance proposals along the lines of those being contemplated by Congress have the potential to raise around $700 billion over the ten-year budget window. Much more will be raised in the second decade, and in the decades that follow, because revenue raised from these reforms is backloaded, in part because investments in the IRS often take several years to reach their full potential. Outside estimates are more optimistic.⁵ But it is also important to appreciate that this is an area with substantial uncertainty. Therefore, I have been cautious in evaluating the revenue that is likely to accrue from Congressional compliance proposals.