President Biden’s Build Back Better (BBB) plan “is fully paid for, or even more than fully paid for. And CBO just completed a comprehensive review of it in which they found essentially the same thing.”

— Treasury Secretary Janet L. Yellen, in response to Sen. John Neely Kennedy (R-La.) during testimony before the Senate Banking Committee, Nov. 30

“I want to … correct a record that got real fuzzy when Senator Kennedy asked Secretary Yellen a couple of times about debt and deficit, and she said that the Build Back Better plan is completely paid for and to prove the point, she said the CBO agrees with her. …. I want to read directly from the Congressional Budget Office’s score: ‘CBO estimates that enhancing this legislation would result in a net increase in the deficit totaling $367 billion over the 2022 through 2031 period.’”

— Sen. Kevin Cramer (R-N.D.), later in the same hearing

“I’m sorry, but you didn’t read it completely. It does say $367 billion over 10 years effect on the deficit. It then notes that it did not include the revenue that would come from enhanced resources for tax enforcement.”

— Yellen, in response to Cramer

Yellen’s first comment that the Congressional Budget Office (CBO) verified that the BBB plan was deficit-neutral caught our attention and we had already queried the Treasury Department by the time Cramer decided to call Yellen on it. But then, as Yellen noted, Cramer left off a caveat that the CBO included in its statement. The $367 billion deficit estimate did not include the impact of $80 billion of increased investments in the Internal Revenue Service.

Still, in a separate informal estimate of the impact, the CBO said IRS investment would result in $207 billion in additional revenue, for a net impact of $127 billion in deficit reduction. That still would leave Biden’s plan in a deficit hole. (To be precise, a $160 billion deficit. We’ll explain the math below.)

Yellen acknowledged that as she continued responding to Cramer. But she offered a caveat to the CBO’s caveat: They “have indicated that their scoring of that does not take account of behavioral changes that would result from a regime of stricter tax enforcement and Treasury put out its own estimate.”

Cramer sniffed: “And fairy dust creates energy, I understand.”

What’s going on here? Let’s explain.

The Facts

As regular readers know, we have been tracking Biden’s commitment that his spending plans would not add “a single penny to the deficit” or would cost “zero dollars.” The president has sought to show he is being fiscally responsible — in contrast to former president Donald Trump’s deficit-financed tax cut — but the numbers often rely on budget gimmicks, such as only funding expansions of programs for a few years.

This issue, however, does not involve a budget gimmick. Instead it concerns uncertainty over how to estimate the impact of a huge increase in the IRS’s budget. Biden’s proposed $80 billion infusion would double the size of the IRS over the next decade. Presumably more audits and technology upgrades would mean fewer tax cheats could get away with their schemes. But how much additional money would be raised?

Officially, the CBO cannot include an estimate of the impact in its overall estimate because two key guidelines imposed by Congress — specifically scorekeeping guidelines 3 and 14 — prevent the scoring of indirect effects (such as people getting more honest because they were afraid of an audit) from spending proposals. The same guidelines apply to budget estimates made by the Office of Management and Budget.

As an interesting Yale Law Review article from 2018 noted, this has had the perverse effect of counting cuts to enforcement spending as deficit reduction, thus encouraging bigger cuts to the IRS budget. Ironically, the guidelines were created in part because the IRS in 1987 greatly overestimated how much additional revenue would be raised by a big staff increase.

In a Nov. 18 letter to Congress on the House bill, CBO Director Phillip L. Swagel said the question of how taxpayers would react to an enhanced IRS is uncertain, with different results found in academic literature. In a September blog post on Biden’s proposal, he wrote that “the estimate reflects CBO’s expectation that the increased enforcement activities would change the voluntary compliance rate — that is, the share of taxes owed that are paid voluntarily and on time — only modestly.”

But that section of the blog post linked to a July 2020 report that stated the “CBO has not estimated the deterrent effect of increased enforcement on other taxpayers — and thus its estimates do not show increases in revenue as greater enforcement influences more taxpayers to comply with tax laws.” So some officials believe it’s unclear the degree to which human behavior was part of the CBO’s estimate of the revenue that could be raised.

In a statement to the Fact Checker, CBO said estimates in the letter and blog post “reflect the expected impact on voluntary compliance, that is the deterrent effect of the changes,” while the July 2020 report “did not reflect any changes in voluntary compliance.” The statement again noted CBO expects the increased enforcement activities would “modestly increase the voluntary compliance rate” but “the magnitude of that effect is highly uncertain.” If the legislation is enacted, CBO says, the anticipated additional revenue from enhanced enforcement would be reflected in baseline budget projections.

We should note that the $367 billion deficit figure cited by Cramer includes the $80 billion cost of enhancing IRS funding. So, if you add in the $207 billion in additional revenue estimated by the CBO, that gets you down to a deficit of $160 billion.

But the Treasury Department says the CBO estimate shortchanges the impact of greater enforcement.

Treasury estimated that, before indirect effects, the $80 billion infusion into the IRS would result in $320 billion additional revenue over 10 years. That’s higher than CBO’s $207 billion because CBO believes the diminishing returns from greater enforcement will happen more quickly than Treasury. (Treasury argues that CBO does not account for how underfunded the IRS had become.)

Then, Treasury adds a 50 percent boost for deterrent effects, resulting in a $400 billion gain net of the $80 billion investment. To Yellen’s point, if the Treasury estimate is accurate, this would more than make up the projected shortfall of $160 billion.

Officials say that using 0.5 is more conservative than many other estimates on the deterrence impact, given that every year the IRS fails to collect an estimated $600 billion in tax revenue. Estimates in the academic literature suggest the multiplier effect would be four to 12 times. The Trump Treasury Department in 2019 said the figure was three times as much: “The indirect deterrence value of IRS enforcement programs, which is conservatively estimated to be at least three times the direct revenue effect.” So the Biden variable is one-sixth of the Trump number.

Yet the respected Penn-Wharton Budget Model did its own calculations, after reviewing the academic literature, and concluded that the $80 billion in IRS funding would yield only a net gain of $190 billion over 10 years. That’s higher than the CBO, but half the Treasury estimate. “Our reading is that richer and more-sophisticated taxpayers — those who the administration is explicitly targeting in the proposal — appear to be less responsive to the threat of audit than lower-income and less-sophisticated filers,” said John Ricco, associate director of policy analysis.

The Penn-Wharton estimate, though slightly higher than CBO’s, would still leave the bill in deficit territory.

We mentioned at the beginning that there were budget gimmicks in the bill. We won’t go through all of them, but one especially noteworthy element is a proposed boost in the cap for the deduction of state and local taxes (SALT) from $10,000 to $80,000.

On paper, the provision reduces the deficit by $15 billion. But that’s because the SALT cap, part of the 2017 tax bill, was due to expire after 2025. So even an increase in the SALT cap is technically a revenue raiser in the later part of the 10-year budget window because the CBO has to assume it would have been repealed. But the bill leaves untouched all sorts of other tax elements — such as rate cuts and a doubling of the standard deduction — that were also due to expire.

In this case, the conventions of CBO scoring work in the administration’s favor. The Committee for a Responsible Federal Budget estimates that SALT provision costs $600 billion over 10 years, even though the official CBO score counts it as reducing the deficit.

Administration officials have said that future deficit-neutral legislation will address such temporary measures, such as the SALT cap and programs that have been only partially extended.

“Analysis by CBO, the Joint Committee on Taxation, and Treasury collectively make it clear this legislation would be more than fully paid for,” said Treasury Department spokeswoman Alexandra LaManna. “It has long been understood that CBO’s methodology understates the revenue potential of tax compliance investments by essentially ignoring the impact of changed behavior by tax evaders, so that they are more likely to voluntarily pay the taxes they owe.”

The Pinocchio Test

No one can accurately predict what will happen to tax collections after the IRS is suddenly flush with cash, personnel and new technology. The Treasury Department has grounds to dispute the CBO’s estimate. But the fact remains that the CBO is regarded as a neutral arbiter, the official scorekeeper, and the rules on budget scorekeeping help the administration claim its plan is budget-neutral as much as it hurts it.

Yellen cannot pick and choose what part of the CBO score she chooses to accept to claim that the CBO found that the Build Back Better bill is deficit-neutral. To her credit, when challenged, she acknowledged the difference of opinion with the CBO on modeling behavior effects. But her statements are still worthy of Two Pinocchios.

Two Pinocchios

Send us facts to check by filling out this form

Sign up for The Fact Checker weekly newsletter