Medicaid

This post has a few more of my random reflections on OBBBA.¹

Table of Contents

Medicaid
1. Nature of the Cuts
2. Rhetorical Responses
  1. CBO’s estimates are too high
    1. My take
  2. We can fix them
    1. My take
Littering the tax landscape

I blog about taxes and have no expertise in Medicaid, but I can’t let a logical inconsistency in political rhetoric in such a key provision of the bill pass without commenting.

According to the CBO score, the bill will reduce federal spending on Medicaid by almost $1 trillion over ten years (see here, numbers are in Subtitle B, chapter 1 of the Title VII page of the downloadable spreadsheet). If these estimates are remotely close to accurate, this is a very large reduction in the social safety net (by some accounts the largest ever); it will fall heavily on many Trump voters and red states; and Trump campaigned on not cutting Medicaid.²

This situation has led to hand wringing by a few populist type Republicans, presumably because it hurts their base voters and communities.³ By contrast, while Dems are appalled by the potential real-world effects of the cuts, it has them licking their political chops as they plan to make the Medicaid cuts a campaign issue in the midterm and 2028 presidential elections. Whether that will work as a campaign strategy is debatable (see here and here for why it probably won’t), but I’ll do a little amateur analysis of the defenders of the cuts’ rhetorical claims and what I think they reveal about the reality as I see it.

The two main claims that bother me are that the estimated effects of the cuts are wildly overestimated or that because the cuts occur off in the future, they will likely be softened or repealed before they go into effect. If either or some combination of those assertions is correct, then the deficit/debt effects of OBBBA will be even higher. Disassembling about the quality of the estimates and planning future fixes, which you have no plan to pay for, does not change that reality; you can’t have your cake and eat it too.

Nature of the Cuts

Four observations about the Medicaid cuts (again, caveat that this is based on my limited, very high-level understanding of them and Medicaid itself):

First, the cuts largely result from imposing work requirements on able-bodied adults and by capping states’ ability to impose provider taxes that effectively increase the share of program costs paid by the feds. As a very crude explanation of the latter, states “tax” providers while increasing their reimbursements sufficiently to offset the tax. The higher reimbursement holds the provider harmless. Because the feds pay a percentage of the reimbursed tax, this arrangement (“tax”) imposes no cost on the state.⁴ This is widely recognized as gaming the system but has become engrained as part of the system’s overall mosaic – essentially de facto increasing the federal reimbursement rate. It is also subject to complicated federal regulations that restrict states’ ability to do too much of it (e.g., imposing very high tax rates). OBBBA changes those rules by reducing the allowed tax rate in steps over several years. Somehow, this reduced de facto federal reimbursement rate will need to be made up by states paying more or benefit/eligibility cuts or both.

Second, that structure enables proponents of the cuts to say the changes do not reduce individuals’ benefits. Able-bodied adults, if they aren’t already, can go to work to maintain their benefits. States can stop gaming the system by imposing bogus taxes and pay for the share of the benefits that the statutes nominally say they are responsible for. So, it’s just a choice by beneficiaries and/or state governments, not Congress and the feds. The structure of these changes also makes CBO’s estimating task very difficult, because the change in federal budgetary costs depend upon individual and political behavior that is inherently difficult to predict.

Third, the effective dates of OBBBA’s changes are delayed until after 2026 for the work requirement and after 10/2028 for phasing down the provider tax cap. That is necessary as a practical matter for the work requirement, less so for the tax cap, which could be applied in 2027 as well. That means Congress and/or the administration (by granting waivers or delaying implementation, like the TikTok strategy which ignores a technically binding law SCOTUS has upheld) have time to change their minds about imposing or can mitigate the rules’ effects.

Finally, CBO’s estimated savings from the Medicaid changes are crucial to reducing the bill’s effect on debt – both as a political matter (House passage was mildly in doubt because of objections it increased the debt too much) and as an objective fiscal matter (even nonfiscal hawks are concerned that OBBBA may go too far down a seeming ever-increasing debt path, which is already at a historical high-water mark).

Rhetorical Responses

The primary argument I saw made by the proponents of the Medicaid changes (e.g., Speaker Johnson) argues that they protect the program by ensuring that it only covers what everyone thought it was supposed to cover: i.e., those working or unable to (children, disabled, or elderly) and that states pay their expected (nominal statutory) shares of the program costs. That’s not true other than in their re-formulation of the program’s objectives; it’s what they think, not what the law said. A rough translation: we’re cutting it now, so you don’t have to worry about us cutting it in the future, I guess? It’s a way of saying that is what the program’s parameters should be in their view. Fair enough, but it’s not saying benefits won’t be cut. They will be or CBO’s estimates are wildly wrong. What one thinks about the changes depends upon your policy and political preferences and isn’t of much interest to me (it’s not illogical, just an obtuse way of saying it).

I’m more interested in the arguments made by Republican and conservative types who are concerned (politically or policy-wise) about the changes and argue that they really won’t be that bad. I see basically two flavors of these arguments.

CBO’s estimates are too high

A WaPo column by Ramesh Ponnuru, Why the GOP’s Medicaid cuts are less than meets the eye, is a good example. He doubts the estimates that more than 10 million people will lose coverage⁵ because of OBBBA:

But the coverage loss from the bill is likely to be a lot smaller than the predictions.

We should keep in mind, first, that the CBO has a poor track record in modeling the effects of health care legislation. * * *

Medicaid estimates, specifically, have been way off. In 2012, after Obamacare had passed and been modified by a Supreme Court ruling, the CBO thought that in 10 years, Medicaid, along with the Children’s Health Insurance Program, would enroll 43 million people. The actual number was more than 87 million, and it had already reached 70 million before the pandemic sent enrollment shooting up.

* * *

The effects of the work requirement are likely to be blunted by state implementation. A CBO assessment last month found that a version of the work requirements would cause a 4.8 million decline in enrollment. Much of the evidence CBO uses, though, comes from a short-lived experiment in Arkansas, a red state that expanded Medicaid reluctantly. Larry Levitt of the health-policy group KFF, drawing on that example, notes that states “could require Medicaid enrollees to report work or exemptions as often as every month.” They could — but will California or New York really be so stringent? Especially when it would lower their federal funding?

My take

Making these estimates is extraordinarily difficult and I am confident that CBO does a competent and neutral job. The obvious difficulties are the complexities of human behavior and specifically political behavior (i.e., what will legislatures and governors do) for which there is little data to provide guidance. Two states, Arkansas and Georgia, have limited and brief experiences with work requirements.⁶ A large cut in state provider taxes has never been done to my knowledge.⁷ That means that there is a large confidence interval around these point estimates.

That says nothing about a bias in one or another direction (i.e., over or underestimating the size of the reductions). An underestimate, which is certainly plausible, would make matters worse from the perspective of reducing benefits. Why should we assume, if CBO is bad at making health care program estimates, that they always will be wrong in one direction? If they had consistently erred in one direction, one would expect the analysts to assess why and modify their methods. Ponnuru must not think so. He must think they have a persistent bias in one direction.⁸

Should we assume that CBO analysts have some sort of inherent bias to overestimating? I can conjure up two reasons why that might be so (likely subconsciously by the relevant analysts):

Estimating larger cuts will make their Republican employers happy. Although they’re nonpartisans, Congress is controlled by Republicans and so has more control over CBO. Larger estimated cuts will make the deficit effects of the tax cut smaller and will reduce the need to make more reductions, which would be even less politically palatable. Both things will help Republican leadership put a passable bill together, a good thing for them.
If (as the Right always suspects) the analysts involved have policy priors that do not favor the cuts, estimating that they have larger effects will serve those priors. It will allow the program to escape deeper cuts but at the expense of a larger deficit. I’m sure they try hard to avoid explicitly putting their thumb on the scale in that manner, but one’s perception of the world/reality may matter in deciding how much weight to give assumptions for which there is so little data to inform those decisions.

Both factors point toward overestimating the magnitude of the cuts. They are wildly speculative. I’d assign a very low probability to them having meaningful influence on the estimates, but it’s possible.

The key point to keep in mind, which Ponnuru makes to his credit, is that if CBO seriously overestimated the amount of cuts, then the deficit effects of the bill will be much larger. The ship missed the rock but ran into the hard place. Second, the estimates are very large ($988 B), so they need to be off by a lot to matter much in terms of the hardship imposed on recipients and providers. Let’s say they’re too high by 25%, then the cut is ONLY three-quarters of a trillion dollars. Small comfort. In my mind quibbling about estimation accuracy is silly, a false hope, or worse – at least in the way that Ponnuru does it.

We can fix them

As noted above, the Medicaid cuts do not go into effect immediately. That means Congress or the administration can lessen their effects or eliminate some of them altogether. Senator Josh Hawley, one of the GOP populists who opposed the cuts, and Senator Ron Johnson who also opposed them have made this point as a basis voting for OBBBA. Here’s Politico, Ron Johnson believes he will get ‘second bite of the apple’ on Medicaid cuts:

Sen. Ron Johnson (R-Wis.) believes he has a commitment from the White House and Senate GOP leadership to get another chance to repeal an expansion of Medicaid offerings — a controversial proposal that failed to make it the final version of President Donald Trump’s sweeping domestic policy package.

“I think I pretty well have a commitment. They’re going to do that,” Johnson told reporters of the prospects that Republicans will reconsider a provision that would end the federal government’s 90 percent cost share of funding for new enrollees in states that expanded Medicaid under the Democrats’ 2010 health care law.

Here’s NBC on Hawley:

And when asked about the steep Medicaid cuts in the bill, Hawley continued to criticize them. Hawley said his “goal” is to ensure the provider tax changes, which will limit state reimbursement for Medicaid, don’t go into effect in Missouri in 2030 — even as he helped to pass a piece of legislation that will do just that.

My take

There is plenty of time, especially for the provider tax reductions, to modify or mitigate the cuts. As the administrative challenges of implementing them (work requirements) and their social, health, and economic effects become more apparent, changes seem probable. But:

Their gargantuan size means making meaningful changes will be extremely expensive and difficult to achieve without materially increasing the debt.
Voting for something that you firmly oppose and pledge and/or hope to change (e.g., Hawley, Johnson, and Murkowski) is an odd political strategy, since it gives up your strongest leverage in the legislative process. It must mean that you value some combination of tax cuts and deficit concerns and appeasing Trump and MAGA world more highly than your concerns about the Medicaid cuts.
The pattern of structuring payfors (here, Medicaid cuts) in the future and financing the politically desired policy (here, tax cuts) immediately is a time-honored bipartisan tradition. It was done by the Dems with the ACA, delaying implementing some of the tax increases to fund it. That almost inevitably results in unraveling the deficit-mitigating payfors. Republicans and a few Dems joined to repeal key ACA payfors. The medical device tax did go into effect, but was repealed, the Cadillac health coverage tax was repealed before it could be imposed. Some version of that seems likely to happen here unless events intervene (e.g., debt hysteria or crisis). However, there is also a good chance any mitigating factors will be partial or skewed to the politically favored. For example, the rural hospital fund is skewed to red constituencies, as contrast with a more neutral set of criteria that would help any hospital heavily dependent on Medicaid funding or with a high percentage of Medicaid-covered patients, even if it were located in an urban area.
The likely bottom line is that the deficit effects of OBBBA are worse than the estimates say, if past political behavior is any guide.

If you doubt that Hawley and others of similar ideological ilk have little have little to no concern about deficits or fiscal responsibility more generally, check out his proposal to rebate tariff revenues.

Littering the tax landscape

OBBBA’s erosion of vertical and horizontal equity is the biggest policy flaw of its tax changes. The distribution of the income tax will be more regressive (although still progressive) and there will be a wider dispersion of tax burdens for households with equal incomes. Both bad and I’m not sure which is worse.

But the more time I spend going through the details, I’m more depressed by how bad the changes are at a granular level, which likely reflects the several decisional dynamics – the hurry with which it was put together, the lack of public scrutiny, and the impetus to throw every Republican member’s (and his or her staffers’) pet ideas into the bill to get their support, grease the skids, etc. The result is an absolute mishmash, a littering of the tax code with provisions with little to no or contradictory policy rationales.

This has serious consequences for the ability of the IRS to administer and enforce the tax law, as well for taxpayers to comply with it. Janet Holtzblatt at TPC points out that OBBBA requires the IRS to issue about a couple dozen new regulations explaining and filling in the gaps at the same time its enforcement budget is being cut. Something has to give with a system that already had higher demands for guidance than resource to provide.

As Holtzblatt points out, the financing dynamics (e.g., paying for private letter rulings and the interests of those suggesting topics for public guidance) already skew the guidance to those with resources or to issues affecting higher-income filers. This does not bode well in terms of clarification of the multiplicity of provisions affecting the masses (beyond the big-ticket items Trump proposed that have already produced a press release from the Service).

To make matters worse, many of these provisions never had hearings or public scrutiny (hello, Trump accounts) and will affect millions. This is a recipe for confusion and unintentional mistakes and noncompliance. Bad.

The Tax Law Center has a new project focused on improving tax administration with a blue-chip set of advisors and contributors. I’m sure they will come up with many sensible ideas with likely little more effect than yelling for the wind and rain to stop.

Two niche provisions

A sprawling bill like OBBBA inevitably contains odd, narrow provisions. Two that I ignored as OBBBA was winding its way through Congress was the effective repeal of the National Firearms Tax and a new limit on deducting gambling losses. Both strike me as peculiar in different ways.

National Firearms Tax

The NFA was enacted in 1934 (allegedly) in response to the St. Valentine’s Day Massacre and the assassination attempt on FDR. It was more regulation (prohibition) than tax in the days when the Lochner Court imposed serious limits on congressional power under the Commerce Clause.⁹ Imposing a prohibitively high tax was a way to avoid potential limits on regulatory powers. (Second Amendment concerns weren’t much of a consideration in those days.) The tax was/is a flat $200 per item ($4,800 in 2025$) and applies to machine guns, short barrel shotguns (the proverbial sawed-off shotgun, I assume) and rifles, silencers, and similar – weapons and accoutrements Congress wanted to keep off the streets and considered to be criminal tools, rather than legitimate firearms for hunting and protection.

OBBBA’s provision started in the House as an exemption for silencers. The Senate thought that was a good thing and expanded it to exempt everything except machine guns (a bridge too far, I guess). That increased the cost from $1.5 billion in the House bill to $1.7 billion in the Senate bill. Obviously, the tax is mainly being collected on the production of silencers or suppressors. The changes are effective 1/1/2026.

The rationale for repealing this longstanding “tax” is unclear to me. An older CRS report describes proposals to modify the tax on silencers. A case can be made that the tax is unconstitutional now, even though it was upheld back in the 1930s. The Second Amendment doesn’t mean now what it meant then. Congress rarely repeals laws out of a preemptive concern that they may be unconstitutional, especially at a cost of nearly $2 billion, so some sort of pro-gun rationale seems more likely.

This is all new to me. What’s surprised me was how many silencers/suppressors are apparently being sold based on JCT’s revenue estimate. Since the tax is a flat $200/item, the $1.5 billion estimate implies something on the order of 750k silencers will be sold per year. (I would assume a $200 drop in price will stimulate sales FWIW but that is not reflected in estimate of foregone revenue from a repeal I assume.) That’s for items that Congress considered of so little social value (i.e., it was primarily a tool of criminals) in the 1930s that they de facto prohibited their sales. It’s amazing how attitudes have changed in 90 years.

Income from wagering

OBBBA’s changes in how the income tax applies to gambling are even odder. These changes have attracted a modest amount of attention, including claims by some members who voted for the bill that they were unaware of the changes and by even more members who now want to repeal them.

Deducting gambling losses traditionally have been allowed to offset winnings as a miscellaneous itemized deduction, which (obviously) required itemizing.¹⁰ TCJA temporarily eliminated the miscellaneous itemized deduction for casualty losses, but not gambling losses. They continued to be allowed to offset winnings.

OBBBA made TCJA’s disallowance of miscellaneous itemized deductions permanent but added a twist to wagering loss deductibility. It only allows 90% of losses to be deducted. This creates phantom income for someone with net winnings or modest net losses (10% or less). It raises a material amount of tax revenue – $1.1 billion per the JCT estimates for the budget window.

This provision has some very odd and uneven effects. Here are a few:

It has no effect on gamblers whose net losses exceed 10% of winnings.
A losing gambler with small losses (less than 10% of the amount bet) will have income tax liability, even though she has no income from wagering.
For a gambler with winnings, the tax rate effect is higher the lower the return on the amount bet. Put another way, the effective tax rate gets smaller as the return on the amount bet rises. To take an extreme example, it’s meaningless for a lottery winner who pays $5 for a ticket that yields $100,000. By contrast, a high-volume sports bettor who relies on finding bets in which the odds makers are slightly off (essentially playing a narrow margin probability game) will likely be put out of business by the tax. The odds have to be off by a larger amount for her bet to have a positive ROI.
The effect also varies depending upon the other taxable income of the gambler. Someone in the top tax bracket pays a bigger tax price obviously than a low-income person. The phantom income is added to and taxed at their marginal rates. As a matter of algebra, the new rule raises the tax cost of making a bet by 0.1% * the taxpayer’s marginal income tax rate.

I can’t come up with a reasonable rationale for such a tax. It makes no sense as a way to impose a sin tax to deter gambling or as a way to impose a tax on the gaming industry to recover some of the social costs of problem gambling. More accurately, many better structures could easily be devised to accomplish policy goals like those. It’s just one of those head scratcher provisions.

I cannot imagine that it will last long, if it ever actually is allowed to go into effect. According to the NBC story linked above, both chairs of the tax writing committees support repeal. (Q: how the heck did this get into the bill, then? Its inclusion is a massive indictment of the process the two chairs used IMO.)

Matt Levine in his interesting and always entertaining daily newsletter, Money Stuff (if you’re interested in corporate finance and don’t scan it everyday, you’re missing out; it’s free), points out that some commodity markets now sell contracts based on the outcomes of sports contests. The CFTC has approved this, surprisingly. Returns on these contracts, functionally indistinguishable from betting on the games, are taxed as capital gains with no restrictions the deductibility of losses from any other capital gains (and $3k/year to reduce ordinary income). This is a big tax advantage, even absent the silly OBBBA provision.

Charitable contributions

OBBBA makes a host of changes in the charitable contribution tax rules. Some of these were likely intended to raise revenues to offset the tax cuts, while others further the GOP policy agenda (credit for contributions to scholarship granting organizations or SGOs) or to mollify interest groups (charities have been lobbying for a nonitemizer deduction for years, more intensely after TCJA’s standard deduction increase). The changes certainly do not simplify or make the incentives easier to understand.

The table below lists the changes for which the JCT published an estimate (negative numbers are tax cuts and positive one’s tax increases). The amounts are for fiscal years 2025-34.

Provision	Revenues ($B)
Contribution credit – SGOs	($25,930)
Nonitemizer deduction	(73,750)
0.5% AGI floor for individuals	63,107
1% floor for corporations	16,603
Alaskan subsistence whaling	(3)
Net effect on revenues	($19,973)

OBBBA’s charitable contribution changes
Source: JCX-35-25

The overall net effect is a tax reduction of about $20 billion. If one ignores the credit for contributions to scholarship granting organizations (not really a charitable giving incentive since it’s a full offset for amounts given), the net effect is to raise revenues slightly (by about $6 billion).

Not reported in the table is the effect of the increase in the SALT deduction maximum from $10,000 to $40,000. That change will increase the number of itemizers (by about 2 to 4 percentage points according to TPC) who, as a result, will qualify to claim itemized deductions for their charitable contributions. Thus, one of the revenue effects of the higher SALT deduction limit will be more itemizers claiming deductions for their charitable contributions.

Non-itemizer deduction

One of the main criticisms of TCJA’s changes was that its higher standard deductions dramatically limited charitable contribution incentives to a smaller and higher income group of individuals. (TPC has a short piece that describes these effects.) COVID relief legislation allowed a one-year (tax year 2021) deduction for cash contributions by nonitemizers. OBBBA reinstates that deduction and makes it permanent at a higher dollar maximum ($1,000 single/$2,000 married joint versus $300/$600 under the expired provision).

Because of the maximum dollar limit, this is a flawed incentive – it won’t apply to marginal gifts by larger contributors. For larger contributors, it is little more than a modest bonus, a little extra income. But it has a big revenue loss, more than $73 billion.

0.5% AGI floor for individuals

While extending a new benefit to those claiming the standard deduction, OBBBA reduces the benefit for existing itemizers. It does that by disallowing deduction of contributions equal to 0.5% of AGI. This floor will also apply to the new deduction for nonitemizers but will not apply to qualified charitable distributions from traditional IRAs – direct payments made to charities for individuals aged 70.5 years old or older from their IRAs. (At least, that is how I read the law’s language.) This small change increases revenues by over $63 billion, almost totally offsetting the loss attributable to the new deduction for standard deduction filers.

This change has a good policy basis; it makes the incentive more cost effective. Disallowing contributions (scaled to income level) that nearly all contributors would make (w/ or w/o a deduction) minimizes the tax benefits conferred on people for doing what they would do anyway. It focuses the incentive at the critical margin. I have two criticisms, though: (1) the floor should have been set higher (e.g., at 1% or 2%), which would double or triple the savings with little negative incentive effect; (2) the floor should also be applied to qualified charitable distributions from IRAs. The same logic applies in that context and uniform rules are better.¹¹

1% floor for corporations

OBBBA also imposes a 1% taxable income floor on corporate gifts. This raises over $16 billion in revenue.

The logic of this change is similar to the floor for individuals. However, it is worth noting that contributions and advertising/PR spending by businesses are close substitutes. Think of the local retailer sponsoring youth sports or making contributions to get their names listed in arts programs and similar. Those contributions may be motivated by the owners’ charitable impulses or to promote their businesses or more likely a combination of both. The floor will encourage characterizing contributions as ordinary and necessary business expenses. We can expect more “contributions” to be deducted as advertising or promotional expenses. I expect the estimates attempted to take this into account.

Contributions to subsistence whaling

This is one of a suite of minor provisions added at the last minute to secure Senator Murkowski’s decisive vote for the bill.¹²

SGO tax credit

This is a longstanding GOP desidera, not a charitable contribution incentive but de facto government funding of K-12 private schools since the credit equals 100% of the amount contributed.¹³ This is, of course, occurred at the same time that direct federal funding of public schools was being withheld and might be cut.

The interesting issue is that states must opt in for contributors to qualify for the credit. How this plays out in blue and purple states (including Minnesota) will be a potentially hot political issue. (A STRIB op-ed has already been published opposing Minnesota opting in.) It essentially pits the lure of “free” federal money for in-state private schools versus a concern that doing so will undercut the viability of public schools.¹⁴ It’s roughly mirrors the dilemma red states faced in deciding whether to opt in to the ACA’s expanded Medicaid eligibility rules. That, however, bore no risk of undercutting the existing health care delivery system, so it was more a matter of ideology and perceptions about the risks of future cuts in federal ACA funding, which now are, in fact, looming.

Missed opportunity

Overall, the charitable contribution changes seem positive to me (ignoring the SBO credit). They modestly expand who qualifies for tax incentives, while making the existing incentive slightly more cost effective (thanks to the income floors), with little net change in revenues. But it was a missed opportunity to adopt a universal incentive for all contributors, while cutting back on the overly generous incentives for gifts of appreciated property. The right set of changes could have made the incentives more effective at a lower cost. But in a bill and process of this nature, making hard choices to improve tax policy is expecting too much. The process was too hurried and too partisan to expect much.

Not GILTI

OBBBA revises the international tax rules, including significantly modifying calculation of GILTI. Under prior law, GILTI excluded the normal return (i.e., 10%) from tangible assets, like foreign manufacturing facilities. Very simplistically, 50% of the return above that threshold and 50% of any return from intangible assets were taxable at the regular corporate rate. The 50% deduction was scheduled to drop to 37.5% beginning with tax year 2026, raising the effective federal tax rate on GILTI.

OBBBA eliminates the 10%-exclusion for the return on tangible assets and sets the general exclusion at 40% (rather than allowing it to drop to 37.5%). That means saying goodbye to one of my favorite tax acronyms, GILTI, which I assumed was intentionally concocted to be revealing. The acronym (GILTI = Global Low Taxed Intangible Income) reflected the theory that its inclusion focused on income that was either from intangibles or constituted super-returns on tangible investments (well, 10% might just be above average rather than super), so it roughly reflected income that was artificially shifted to a low-tax foreign country to avoid US tax. Now, 60% of most CFC income will be subject to inclusion. As a result, the reference to “low-taxed income” is gone. The law now refers to it as just net CFC tested income. I don’t know what the consensus acronym will end up being (NCFCTI?).

This will present some interesting issues as to how Minnesota should conform to OBBBA’s changes.¹⁵ My instinct is that straightforward conformity (i.e., just updating to the federal changes) would raise revenue, because the effect of increasing 37.5% to 40% will be more than offset by eliminating the 10% threshold for the income from tangibles. However, the JCT spreadsheet does not have the needed detail to see whether that is so. I have a suspicion that fairly significant amounts of new revenue could be involved but have no basis for knowing that. It also depends upon the mix of income of MNCs taxable in Minnesota and their apportionment factors, making a Minnesota estimate slightly more complicated than just allocating the detail from the JCT estimates that JCT typically provides to DOR.

If it is the case that simple conformity would raise material revenues, this should cause the administration and legislators to rethink Minnesota’s very aggressive GILTI tax regime, in my opinion. A few reasons why I think that would be prudent:

First, ITEP has estimates of how much revenue would be raised by mandatory worldwide combined reporting (WWCR). The estimates show that Minnesota’s GILTI regime raises $240 million annually more than WWCR would. WWCR is the basic benchmark for how much a neutral tax apportionment tax regime should raise. (The credible arguments against adopting WWCR are based on administrative, compliance, enforcement, and foreign relations considerations, not theoretical policy.) I have no idea whether the ITEP estimates are accurate (they make quibbles about the possibility they overestimate revenues for GILTI inclusion) but if they are even close, they strongly suggest Minnesota is overtaxing CFC income with its GILTI tax. Conformity to OBBBA would likely make that situation even worse. That combination gives someone like me pause and argues for moderating the tax – i.e., using conformity revenues to somehow cut the tax, not to fund other tax reductions or spending.

Second, there are sure to be constitutional challenges to Minnesota’s GILTI tax. I have tended to discount the probability of their success, but it is certainly not zero or even minimal. The ITEP estimates lend some credibility to opponents’ case.¹⁶ (SCOTUS has simply refused to consider the general issue when petitions for cert are filed, as in the Caterpillar case.) Changes in SCOTUS personnel makes assessing the probability more uncertain. Losing a case could be catastrophic fiscally, similar to Cambridge Bank – with the state obligated to pay refunds, plus interest, for many years and for many taxpayers. That is so, because litigation of this type takes forever and many corporations have open audits (tolling the SOLs) or will file for protective refunds.

Third, it’s a certainty that a fair number of corporations will make cases for administrative relief (even more so if simple OBBBA conformity is adopted) because the treatment of GILTI mismeasures the amount of their Minnesota share income. If a MNC would consistently pay less under WWCR, I would think DOR would have difficulty denying relief if only to head off undesirable facts in a constitutional challenge.

One fix would be to add an option for corporations to elect WWCR for an extended period (e.g., 5 or 10 years, so they can’t toggle back and forth based on bad or good years). That would provide a default alternative to 290.20 relief and could be made the rule that would prevail, if the taxation of GILTI is invalidated (deterring constitutional challenges – not sure that passes due process muster?). Of course, if the ITEP estimate is close to accurate, elective WWCR would likely be scored as prohibitively expensive. Reducing the inclusion percentage would be another obvious strategy for mitigating the tax.

NOTES

That is no longer the official popular name of the bill thanks to a successful Democratic point of order. I assume people will ignore that and continue to refer to it as OBBBA. ↩︎
It is worth noting that reducing Medicaid by a trillion dollars (or whatever the real number is) will not reduce health expenditures and consumption by that amount. The consensus based on quality research by health care economists is that the amorphous health care system (hospitals and other providers) still provides services to the uninsured. That occurs, in part, because federal law requires it of hospitals, emergency care providers, etc. It’s just provided at a lesser level of maybe 75% or so of the insured. So, cutting Medicaid dramatically shifts much of the cost elsewhere. This should be viewed as an implicit social tax of uncertain incidence. ↩︎
This characterization by Ross Douthat, a conservative Republican, captures it succinctly: “And it is coalition-shrinking folly for the G.O.P. to persistently cut programs that benefit its own voters while always lightening burdens on wealthier voters who are trending toward the Democrats.” His general point is that the Republicans are policy prisoners of their anti-tax litmus test. Trump ignores the inconsistency with his promise and the reality by falsely denying he’s cutting Medicaid. Simple but not true. ↩︎
A stylized example will illustrate how this works for those who are not familiar with it. Assume a 50% federal match rate. State imposes and collects $2 tax from Hospital for a procedure. State bills the feds for the $2 tax and gets federal reimbursement of $1. State now has $3 = $2 tax + $1 federal reimbursement (it was reimbursed for the procedure’s cost + tax) – and can pay the hospital its previous rate for the procedure + $3 at no cost to the state budget. Hospital is $1 to the good. ↩︎
The media focus tends to be on the number of people losing insurance. It’s also possible that states will cut benefits by reducing reimbursement rates, procedures covered, and similar to offset the dropping cap on provider taxes. ↩︎
The one ACA estimate that CBO muffed by a lot involved the effect of repealing the mandate to purchase insurance in TCJA. In their defense, unlike the work requirement, there was no experience with that to go on: i.e., a policy allowing annual purchase of heavily subsidized individual health insurance without a mandate. They had to rely on a few state examples that involved community-rated individual policies that were not subsidized and with no annual enrollment period. Here, there are the Arkansas and Georgia examples that are very analogous. I think Ponnuru is profoundly unfair to the CBO analysts. Typical for right-wing commentators like him. ↩︎
It’s worth noting that the type of state action matters. For example, if the state offsets $1 of reduced provider taxes with its own source revenue, the feds realize 50 cents of savings and beneficiaries and providers are held harmless. But the state is out the full $1. If the state wants to hold its budget harmless by cutting benefits – either by reducing benefits or by cutting eligibility – it must do so by cutting benefits by at least twice as much because of the federal match (50% is the lowest federal match). The cuts must be bigger for states or beneficiaries (i.e., beneficiaries who were added by the ACA) with higher match rates. For the 90% match rate, that almost surely requires dramatically reducing eligibility. It’s the reverse leverage effect of imposing a provider tax. ↩︎
The two examples Ponnuru uses in his column go in opposite directions, I’d observe. ↩︎
To clear, there’s not much of a tax policy rationale for the tax. Support for it must lie in regulatory considerations and there certainly must be better ways to accomplish such goals than just raising the price by imposing a special tax. But it raises a material amount of revenue when there is a professed concerns about deficits, and spending is being cut to that end with little analysis or thought about benefits relative to costs. ↩︎
Otherwise, allowing losses to reduce unrelated income would subsidize a hobby, personal consumption. If you can truly prove that gambling is your trade or business, losses can be treated as ordinary and necessary business expenses rather than an itemized deduction. This is very difficult to do. For a period, gambling losses could trigger AMT because the miscellaneous itemized deduction was treated as a preference. That was fixed a while back. Its more modest effects caused a lot of gnashing of teeth. ↩︎
Providing special charitable incentives for old people with traditional IRAs makes no policy sense IMO. Disclosure: that is the way I make my cash contributions, since it yields the largest tax savings. ↩︎
It increases an existing limit of $10K to $50k – essentially treating business expenses as a charitable contribution. In certain circumstances, that provides a tax benefit. ↩︎
It’s easy to speculate about why the GOP favors this – antipathy for teachers unions, a core base of the Dems, and/or favoring parochial schools to which more of their supporters send their children. ↩︎
Political fights have erupted in red states, like Texas and Oklahoma, among Republicans over state voucher laws for private schools. ↩︎
Full disclosure: one of the big benefits of retiring was no longer needing to decipher the federal international tax rules and to try and figure out how they affect Minnesota tax and how to recommend conforming Minnesota law to them. The federal rules are technically very complex and the federal and states systems – apportionment versus separate accounting – simply do not lend themselves to meshing nicely. My comments in the text are high level, rank speculation. ↩︎
The conceptual argument for the constitutionality of a state taxing GILTI is that it reflects Congress’s definition (albeit very rough) of CFCs’ income that is really from domestic sources but that MNCs are artificially shifting to their CFCs to minimize federal tax. That federal characterization, then, allows a state to treat it as domestic income (after a dividend haircut) and apply normal domestic apportionment factors. That argument will be harder to sell when a blanket 60% of CFC income is captured, I would think. ↩︎