The Tax Gap and the Manchin Memo
The pivotal senator wants action to address the "tax gap." But what exactly does that mean?
Janet Holtzblatt, Steve Rosenthal, and I posted a new working paper yesterday on “The Tax Gap’s Many Shades of Gray.” You can download it here. Comments and suggestions appreciated. From the abstract:
The “tax gap”—the difference between the amount of “true tax” and the amount of tax actually paid—has garnered widespread attention in recent months. Much of the commentary on the subject equates the tax gap with “tax evasion,” a term broadly understood to connote intentional (and potentially criminal) underreporting. This paper cautions against conflating the tax gap with tax evasion. The tax gap includes substantial gray areas where the law is ambiguous and the IRS’s determination of “true tax” is debatable. On top of that, the IRS’s methodology for measuring the tax gap includes upward adjustments that are recommended by front-line examiners but reversed on administrative appeal or judicial review. Moreover, a substantial portion of the estimated tax gap is derived from a statistical technique called “detection controlled estimation” that potentially magnifies the impact of later-reversed recommendations on the ultimate tax gap measure. Weighing in the opposite direction, the IRS’s approach to measuring the tax gap excludes some amounts that clearly constitute tax evasion (most significantly, underreporting of tax on illegal-source income).
Understanding the tax gap’s shades of gray can inform discussions of tax law and policy. We explain how proposals to use the tax gap as a performance target may produce perverse incentives for the IRS. We further explain how additional IRS funding—though necessary to improve the agency’s ability to enforce the tax laws—may have counterintuitive effects on the estimates of the tax gap. We also illustrate—using examples from the taxation of passthrough entities—how legislative reforms can reduce the size and scope of legal gray areas that contribute to the tax gap. Our analysis highlights the importance of increased IRS funding levels and substantive tax law changes as complementary strategies for improving tax compliance.
Hours before we posted our paper, Politico posted a one-page memorandum of understanding between Senator Joe Manchin and Senate Majority Leader Chuck Schumer dated July 28, 2021, in which Manchin laid out his conditions for a budget reconciliation bill. The memo is startling in several respects. Manchin’s demand that the “Federal Reserve ends quantitative easing” as a precondition for the Senate to move forward on budget reconciliation reflects a rather bold rejection of central bank independence. Moreover, Manchin’s $1.5 trillion topline spending number is way below what other Democrats have been contemplating for months. Of particular interest to me was one item at the very end of Manchin’s list of revenue-raising measures that would offset the new spending: “Tax Gap.”
Speaking only for myself and not for any coauthors: I had three reactions to Manchin’s inclusion of the “Tax Gap” (without elaboration) in his list of offsets.
1. It’s a good thing that Manchin supports measures to address the tax gap because that probably means he supports more money for the IRS. As our paper emphasizes, the IRS desperately needs more money. The IRS has endured a decade of deep budget cuts that have hampered the agency’s ability to retain skilled examiners and to hire and train new ones. The agency’s starvation diet also has undermined its ability to provide high-quality services to taxpayers (e.g., answering questions from individuals who are confused by ever-changing and often-complicated tax laws). And the IRS is long past due for a technological upgrade—it still stores tons of taxpayer information on an IBM mainframe that dates back to the Eisenhower administration. A functional IRS can’t live long on a shoestring, and I definitely don’t want to dissuade Manchin from boosting the agency’s budget.
2. More money for the IRS *will* yield more revenue (virtually everyone agrees on that), but more money for the IRS won’t necessarily narrow the measured tax gap. The tax gap—as measured by the IRS—largely depends on the results of random audits conducted as part of the IRS’s periodic National Research Program studies. If those random audits detect more underreported income as a result of improved IRS capabilities, then—holding behavior constant—the measured tax gap will likely rise (though deterrence effects could cut in the other direction). Other factors outside the IRS’s control also can affect the measured tax gap. For example, absent any change in law, the measured tax gap is likely to rise in tax year 2026 when most of the Tax Cuts and Jobs Act individual income tax provisions phase out. That’s because the tax gap is the difference between “true tax” and what’s actually paid, so an increase in “true tax” mechanically widens the gap.
Lawmakers and administration officials therefore run a risk when they frame the case for more IRS funding as “Tax Gap.” If, say, the tax gap in tax year 2026 is larger than today, that doesn’t mean the IRS has failed—indeed, it might be because the IRS is getting better at detecting underreported income. The tax gap is just not a very good performance metric for the IRS. In the short term, the laser-like focus on the tax gap may be one way to generate political support for more IRS funding. But it’s not so obviously a wise strategy for sustaining long-term support for the agency.
3. There is something positively perverse about focusing on the “Tax Gap” while rejecting other key elements of the Biden administration’s and House and Senate Democrats’ tax proposals that actually address the tax gap. For example, as the paper points out, one driver of the tax gap is noncompliance with respect to partnership income. The partnership tax rules in Subchapter K are fatally ambiguous, and disputes between examiners and taxpayers about the application of these ambiguous rules are part of the tax gap. Senate Finance Committee Chair Ron Wyden has proposed a set of sensible reforms that would rationalize Subchapter K. But partnership tax reform doesn’t make it onto Manchin’s list of approved offsets. Of course, it would be crazy for Congress to give billions of dollars to the IRS to duel with partnerships and their partners over ambiguous tax laws while refusing to reform the ambiguous laws that give rise to those duels in the first place. Our paper offers a number of other examples—involving reasonable compensation rules for active S corporation shareholders and limited partners, valuation of illiquid assets in IRAs and for charitable-contribution purposes, and more—where a simple statutory change would be far more effective than thousands of audits.
That’s all for now—but there’s much, much more in the paper for fun (?) weekend reading.