Odds are better than even that the incoming Biden administration will jettison the current requirements that significant tax regulations be subject to economic analysis by Treasury. Despite what undoubtedly have been prodigious efforts on the part of Treasury and the Office of Management and Budget staff, the analyses provide the public with meager new insights into the effects of regulations. Tight time constraints, unique estimating challenges, the enormous complexity of the law, and a fundamentally flawed framework all contribute to the shortcomings of these analyses.
As Greg Leiserson of the Washington Center for Equitable Growth, formerly senior economist with the Council of Economic Advisers in the Obama administration, and Adam Looney of the Brookings Institution, formerly deputy assistant secretary for tax analysis in the Obama Treasury, wrote in 2018: “The experience to date suggests the review process itself may not pass a cost-benefit test.” Now two years later, experience with the new procedure hasn’t changed Leiserson’s view: “It is clear this experiment in cost-benefit analysis of tax regulations has failed. . . . The cost-benefit analyses released alongside proposed and final regulations provide little information relevant to assessing the merits of those regulations.” It’s likely the Biden administration’s starting point on this issue won’t be far from that of Leiserson and Looney (Leiserson and Looney, “A Framework for Economic Analysis of Tax Regulations” (Dec. 20, 2018); and Leiserson, “Cost-Benefit Analysis of U.S. Tax Regulations Has Failed” (Sept. 2020)).
Even in the absence of such a critical assessment of the new procedures, it’s hard to understand why the Biden administration would maintain Trump administration rules for OMB oversight of tax rules. It isn’t something any elected Democrat has expressed interest in, and it only makes extra work for Treasury. As for the OMB, this new task is far removed from its historical role of reining in sometimes overzealous federal agencies that pursue their missions without adequately considering economic costs.
The Trump administration’s move to restrain tax regulations seems to have been motivated more by anti-regulation ideology than practical procedural or political considerations. The push for OMB oversight can trace its academic roots to the work of Kristin E. Hickman of the University of Minnesota Law School (and hired as a special adviser by the OMB in 2018). She wrote in 2014: “As the attentions of the Treasury Department and the Internal Revenue Service shift away from raising revenue . . . the revenue based justification for tax exceptionalism from general administrative-law norms fades” (Hickman, “Administering the Tax System We Have,” 63 Duke L.J. 1717 (2014)). “The exemption for IRS rules from OMB oversight has gone unexamined for too long,” wrote James Valvo of the Cause for Action Institute, an oversight group that advocates for economic freedom. “OMB review of agency rules ensures that the president is able to exercise his proper constitutional role to direct the Executive Branch” (Valvo, “Evading Oversight: The Origins and Implications of IRS Claims That Its Rules Do Not Have an Economic Impact” (2018)).
In April 2017 President Trump issued Executive Order 13789, which directed Treasury and the OMB to review the exemption of tax regulations from OMB review. An April 2018 memorandum of agreement between Treasury and the OMB required that Treasury analysis be consistent with OMB guidelines and subject to OMB review. The incoming president could end this awkward process with his own executive order.
Possible Revision
The Office of Information and Regulatory Affairs is the part of OMB that reviews regulations. The 48-page OMB guidelines for economic analysis of regulations, written in 1983, make up Circular A-4, the premier government reference on the topic of performing cost-benefit analysis of nontax regulations, such as those issued by the Environmental Protection Agency or the Occupational Safety and Health Administration. But much of the discussion in Circular A-4 has no or only tangential relevance in evaluating the economic effects of tax regulations. As Leiserson and Looney emphasized, its most glaring shortfall is its omission of recognizing any benefit from raising revenue.
In the current framework, a tax regulation that eases tax burdens and loses significant revenue will always register as a net positive in any cost-benefit analysis. The process has a built-in bias against raising revenue and for cutting taxes. If the Biden administration continues Treasury economic analyses of tax regulations, it should and almost certainly will adopt a new framework considerably different from that in Circular A-4.
How could the framework be improved? We’ll divide our suggestions — many of which overlap with those of Leiserson and Looney — into three categories: (1) What are the economic effects of tax regulatory decisions that should be estimated? (2) With what alternatives should regulatory decisions be compared? (3) To which regulatory decisions should estimating procedures be applied? It is important to distinguish between regulations that interpret recently enacted laws and regulations written under a general grant of authority. Since the passage of the Tax Cuts and Jobs Act in December 2017, the Treasury Office of Tax Legislative Counsel has been occupied mainly with the enormous task of writing regulations to assist in implementation of that law. The focus here is on tax regulations that interpret recently enacted law.
Estimated Economic Effects
Treasury can and should estimate the (1) revenue impacts, (2) compliance costs, and (3) administrative costs of economically significant regulatory decisions. It could also measure (4) distributional effects. For tax rules that affect individuals, estimated distributional effects could include conventional distributional analysis by income category. For tax rules affecting business, they could include the effect on different business sectors and on businesses of different sizes.
There is also interest in (5) the economic impact of taxes that distort private decision-making and behavior. Economists call this “deadweight loss.” In general, taxes are expected to reduce economic efficiency and economic growth (based on the fundamental assumption that a free market will produce the most efficient results). That generalization has two important exceptions: taxes that serve as user fees for some government-provided service; and taxes that reduce private sector activity that generates negative externalities. As a practical matter, it probably makes sense that any future formal requirements for economic analysis of tax regulations exclude estimation of deadweight loss. These efficiency effects are extremely difficult to quantify and, like so much in economics, subject to uncertainty and controversy.
Compared With What?
To estimate the economic effect caused by a change in policy, we must have a point of reference — some state of the economy without the policy whose effects are being estimated.
Sometimes the point of reference is what is referred to as a baseline; sometimes it is a policy alternative. The three leading candidates for points of reference for estimating the economic effects of tax regulations are: (1) a no-action baseline; (2) a pre-statutory baseline; and (3) an alternative regulatory approach.
Least informative are estimates that use a no-action baseline, which compares the potential economic effects of a regulation with what would happen in the absence of that regulation. For regulations that fill in the details of recently passed legislation, there is little to say beyond the obvious: Taxpayers and their advisers are assisted by the certainty and instruction provided by the regulations compared with the uncertainty that reigned in their absence and would reign in the future without their promulgation. Jerry Ellig, research professor at George Washington University’s Regulatory Studies Center, has correctly chastised Treasury for citing “clarity” and “certainty” as meaningful analysis. He writes that “clarity or certainty are not benefits attributable to a specific regulation, because just about any other regulation the IRS could have issued could also improve clarity or certainty.” (Prior analysis: Tax Notes, May 20, 2019, p. 1181.)
Of course, regardless of the point of reference against which policy is measured, Treasury may be making important choices. Because Treasury must interpret the statute consistently with legislative intent, the use of a no-action baseline leaves little room for Treasury to claim any significant economic effects of a regulation beyond the obvious policy-neutral benefits of the public knowing how the tax administrator interprets the law. In effect, a no-action baseline allows Treasury to move the economic goal posts with any alternative it chooses as the correct reading of the statute.
A second point of reference is a pre-statutory baseline. Under this procedure, analysis would evaluate the economic effects of not only Treasury’s interpretation of that statute but also those effects combined with economic effects of the statute. Ellig argues in favor of using a pre-statutory baseline. He makes the logical argument that evaluation of the economic efficiency effects of a regulation must follow from the economic efficiency effects of the statute.
Put into a simple example: If (A) a tax break provides positive economic effects, then (B) a regulation that expands the scope of that tax break will also provide positive economic effects. You must know the effects of a statute to determine the effects of regulations modifying the scope of that statute. The public provision of Treasury economic analysis of separate tax provisions enacted by Congress would certainly be of great interest, but the obstacles to this approach are overwhelming. In most cases, it would be difficult, uncertain, and time-consuming. Treasury staff would be required to critically review and publicly disclose all their professional views on legislation that, in many cases, was supported by the Treasury Department and the administration. This type of analysis, welcome as it might be in an ideal world, seems operationally unrealistic. In any case, it would probably be better for this type of big-picture analysis to be performed by congressional staff before legislation is enacted.
Under a third approach, economic analysis of a tax regulation would compare Treasury’s chosen regulatory approach with one or more alternative regulatory choices. This has the advantage of identifying and highlighting important and perhaps controversial alternatives (more on that below) and of maintaining focus on the immediate issue of adopting the best regulations (rather than evaluating legislation that is already enacted). It doesn’t require economic analysis of the underlying statute. It doesn’t require knowledge of the assumptions that underlie the original revenue estimate (which congressional staff do not routinely disclose).
For example, suppose Treasury is interpreting a statute that is ambiguous about whether some costs would remain deductible or whether a major industry was subject to a tax increase. Under a no-action baseline, no revenue effect would be assigned to that regulatory decision (regardless of the alternative Treasury chooses). But if Treasury’s chosen regulatory action is compared with an alternative interpretation of the statute, Treasury would be providing meaningful and relevant analysis that should be available to Congress and the public. Under any revision of procedures for Treasury economic analysis of regulations, comparison with major alternatives should be emphasized and, in the interests of providing the best information at the least cost, comparison with a no-action baseline or pre-statutory baseline should be eliminated.
Significant Regulatory Decisions
Much of what has been discussed in this article so far draws heavily from analysis by Leiserson and Looney. In this section we suggest a novel approach for economic analysis of tax regulations. Instead of identifying which particular regulations are sufficiently significant to warrant analysis, we suggest here that the procedure should identify particular decisions that arise in the process of writing regulations that are significant. So, for example, a set of regulations may specify that a category of costs is deductible, that an industry isn’t subject to tax, and that a specific procedure can be used to allocate costs (among different categories of taxable income). We suggest here that each of these, if deemed significant, should be separately subject to economic analysis.
The procedure could work as follows. As Treasury gears up to write regulations that interpret recent legislation, it should publish and ask for comment on a list of potentially significant regulatory decisions anticipated to play a role in rulemaking. The public could then comment on whether they believe these pending decisions are economically significant and identify any items they believe have been omitted from the list. The procedure up to this point shouldn’t significantly add to Treasury’s regulatory workload. Undoubtedly, Treasury in private and informally in its interaction with the public identifies potentially significant decisions it will be making. The requirement described here only systematizes the disclosure of those issues as well as Treasury’s interaction with the public. This should be a welcome step that can only increase transparency.
Next, concurrent with Treasury’s promulgation of final regulations, it would address whether all the potentially significant decisions identified (by itself and the public) are indeed significant. For the decisions Treasury doesn’t deem significant, all it needs to do is say so. This is a quasi-quantitative estimate because without the need to provide a precise number, Treasury would be reporting these decisions not to exceed the quantitative thresholds (described below).
For those it does deem significant, Treasury should endeavor to provide quantitative estimates of revenue effects, compliance costs, administrative costs, and distributional effects. Treasury shouldn’t be able to excuse itself from quantification because it lacks precise and detailed data. It should use the same methods and same standards of precision it would use for a revenue estimate of a legislative proposal (for example, in the president’s budget) or the tax expenditure budget. If Treasury for some reason cannot provide what it believes are accurate estimates, it should describe qualitatively the issues that arise in the estimation of revenue, compliance costs, administrative costs, and distribution.
What should be the numerical threshold amount that determines economic significance, and what effects should be included? Under current procedures, a regulation is considered economically significant if it has an annual non-revenue effect on the economy of $100 million or more, measured against a no-action baseline. Under the proposal suggested here, a regulatory decision would be considered economically significant if the combined revenue effects, compliance costs, and administrative costs, compared with a likely alternative, exceed a high threshold — perhaps something like $250 million in any one year.
Revenue effects wouldn’t be net of gains and losses, but the absolute value of gains and losses combined so that a decision that significantly shifted burden among taxpayers didn’t slip under the radar. The dollar threshold should be set high given the difficulty in performing these analyses and the limited resources of Treasury and the OMB.
Conclusion
If the Biden administration continues requiring Treasury economic analysis of tax regulations and OMB review of that analysis, it should make significant revisions to those procedures. The suggestions described above provide several benefits.
First, they incorporate the highly worthwhile traditional aspects of OMB oversight — that is, weighing costs and benefits of regulations. They make sure Treasury doesn’t lose sight of economic burdens. For example, if a tax increase raises $120 million but incurs $150 million of compliance and administrative costs (especially in the first year), they will make it clear that Treasury should return to the drawing board.
Second, our suggestions allow the public and Congress to quantitively assess the impact of Treasury decisions that they otherwise would be unable to ascertain given that most of the public and most of Congress don’t understand the implications of complex changes in tax rules. Third, they require Treasury to perform only the types of analysis that it has long been performing for legislative proposals and for tax expenditure analysis.
The next assistant secretary of Treasury for tax policy and the next OIRA administrator will be key players in the Biden administration’s decision on this topic. Appointees to those critical positions have yet to be announced.