Massachusetts Can Raise Up to 415mn/year by Conforming to GILTI
This reform would be fair, efficient, administrable and legal. Recent arguments made to the contrary by the Massachusetts Taxpayers Foundation do not hold up to scrutiny.
I have already written a great deal on why states should conform to a provision of federal tax law known as “GILTI.” See here for the policy argument, here for the legal argument and here for a short summary of the first two essays. Furthermore, recent analysis done by the respected Penn Wharton Budget Model demonstrates that this change would raise a significant amount of revenue.
In Massachusetts, a serious and thoughtful proposal is on the table to have the state conform to GILTI. The Massachusetts Taxpayers Foundation (MTF) has responded that “taxing [GILTI] would be “complicated, costly and could be unconstitutional.” None of the arguments made by the Foundation are new or persuasive, but they require a response lest their apparent mass hide that they are, in fact, without real substance.
Policy
Clash with federal approach. The MTF argues that a state conforming to GILTI is a problem because “[t]his approach directly conflicts with the federal approach…” This claim makes little sense because conforming to GILTI would be to follow the federal approach. Indeed, there is no reason to believe that states conforming to GILTI would not help the federal approach. After all, even the MTF concedes that GILTI is in part meant “to discourage base erosion tactics by mitigating the benefits” and thus Massachusetts is helping to achieve this goal by conforming. In any event, the presupposition of MTF’s argument is false because there is no good reason that Massachusetts should follow the larger policy prescriptions of the Tax Cuts and Jobs Act (TCJA). At its core, the TCJA radically increased the federal budget deficit in the midst of an economic expansion in order to give enormous tax cuts to the biggest and most profitable corporations. This lost revenue should and could have been used on numerous urgent domestic priorities, including infrastructure. It is wholly appropriate for Massachusetts to try to recoup some of the federal giveaway to finance the kinds of projects a responsible federal government would have been funding.
Double Taxation. The MTF has a whole section entitled “Double Taxation.” Massachusetts’ corporate income tax, like that of all states, does not work in the exact same way as the federal corporate tax. In particular, as MTF notes, Massachusetts would not give foreign tax credits on GILTI income included in the Massachusetts tax base, while the federal government does. This sounds bad and seems to suggest that therefore GILTI will be taxed once abroad and then again in Massachusetts. And yet the claim is nonsense and the MTF apparently knows it (as it should). If one reads the MTF section on double taxation carefully it only claims that Massachusetts “could” end up taxing income twice. Why “could”? The answer is that states do not give tax credits because they have another system, apportionment by formula, that prevents double taxation. The use of this system has been upheld by the Supreme Court for over a century.
Complexity. The MTF claims in its title that GILTI conformity would be “complicated” and “costly,” but these claims are never really developed — and with good reason. A corporation with GILTI is already calculating that number on its federal return. No new work there. The Massachusetts proposal then assigns a fraction of that GILTI number to Massachusetts using the formula that the taxpayer already has to use to apportion income to Massachusetts. Very little new work there. Ironically, there would only be more complexity if Massachusetts were to accept the MTF’s legal contention that the usual formula must be adjusted to take into account foreign business operations. However, this contention is false and so there would be minimal new complexity.
Law
Apportionment. The MTF claims that the current proposal to conform to GILTI is unconstitutional because of how it would apportion GILTI income. I cannot emphasize enough that this claim is false. [Note that just because the MTF’s legal arguments are not consistent with what the law has long been (and should be), I cannot guarantee that some plaintiff won’t succeed on winning through obfuscation; these are hard issues that the courts sometimes get wrong. That said, state tax policy should not be held hostage to aggressive legal entrepreneurship.]
The law surrounding the apportionment of the income of a multijurisdictional business is complicated in its details, but the basic rules are not. Let’s take a simple apportionment formula: a state will tax the income of a multijurisdictional business in proportion to the sales that that business makes in the state. As a matter of law, the Supreme Court has long understood that such formulas are attempting to “slice a shadow” and have reviewed them very deferentially. The simple sales-based formula above is constitutional. Now, if this were the formula for a higher-tax state, then a taxpayer would have a great deal of incentive to contort itself so that it appears to have fewer sales in that state. (As a matter of business logic, the taxpayer will not seek actually to forgo these sales and the profits they generate, only to avoid the tax liability associated with such sales.) It should come as no surprise then that states are not required to take a taxpayer’s word on where its sales (or income) are. The courts have been very clear on this matter. Accordingly, there are numerous aspects of state tax law and doctrine that are used to re-assign for tax purposes the sales or income or other assets of a corporation, and these have all been found constitutional, so long as the method used is reasonable.
GILTI represents income that certain multinational taxpayers claim was earned abroad, but the states need not take their word on it. Indeed, the federal government is not taking their word on it either. GILTI is a practical tool for identifying income generated in the US, regardless of claims made by corporations. Thus, as to the 50% of GILTI that the federal government brings back into the federal tax base, all Massachusetts is doing is agreeing with the feds that 50% of GILTI income was shifted overseas from the US and, accordingly, that shifted income should be apportioned to Massachusetts using its usual apportionment formula for domestic income. Note that 50% is, in fact, a reasonable approximation.
The claim of the MTF is that Massachusetts is barred by the federal constitution from taking this reasonable position and has to treat all of GILTI as really earned abroad. This makes no sense and is simply not true. There are several Supreme Court cases directly on point as to this very issue. As you can imagine, large corporations have long wanted to force states to accept their accounting of where they earn their profits — again, so long as the states are acting reasonably, the corporations lose.
Law/Policy
GILTI is flawed. Buried in the MTF’s argument, but more prominent in the arguments of its predecessors is that GILTI is a rough formula and might include profits that were not really shifted. This is a policy objection, but it is also a legal objection because if the mistake were big enough then what the state would be doing would not be reasonable. This argument is also quite weak. Consider the following chart as to the location of the profits of various US corporations.
What we see here is that US multinational corporations report hundreds of millions of dollars of profit in jurisdictions in which those profits could not have been earned because the reported profits are many multiples of that nation’s GDP. Clearly, these profits were earned elsewhere, including in the world’s largest economy, the United States.
Or consider this chart, showing how unbelievably profitable workers are in traditional tax haven jurisdictions.
In the United States, locally employed workers are slightly more productive than those employed by foreign firms, a pattern true for Australia, France, Germany etc. But in Ireland a foreign worker is astronomically more productive than a local one. How can that be? Again, it is because profits from elsewhere are being shifted to that worker’s firm.
GILTI tries to identify shifted income by looking for inexplicable amounts of profits. As the data indicates, this is clearly a conceptually sound notion. To be sure, the GILTI methodology is not perfect, but it is no more imperfect than numerous other elements of tax law. Furthermore, if GILTI really does misfire, a taxpayer has two options at the state level that it does not have at the federal level. First, if the taxpayer believes that all of its income and assets should be taken into account around the world in order to provide a more accurate picture of its net income, then it can do so under Massachusetts law by making an election for worldwide combination. Second, the taxpayer can petition for equitable relief. That is, the taxpayer can ask the DOR to adjust its apportionment formula, based on specific information provided by the taxpayer showing how it generated such high income with so little capital investment. These safety valves help assure fairness to taxpayers; they also make it even less plausible for anyone to argue successfully before the courts that the Massachusetts proposal is so unreasonable that it is unconstitutional.
Revenue Estimate
The Penn Wharton Budget Model [PWBM] analyzed the revenue potential of GILTI conformity. Its preliminary analysis concluded that Massachusetts likely would apportion to itself some $12.5 billion of GILTI income in 2020. Using this GILTI estimate for Massachusetts, the Massachusetts Budget and Policy Center calculated that the Commonwealth could raise as much as $450mn/year. (PWBM finalized its estimates recently, reducing expected MA GILTI in 2020 to $11.6bn. Using this lower MA GILTI number, MassBudget’s calculation drops to $415mn in 2020.) This is a large number, but note that it only amounts to less than a 20% increase in state CIT collections of about $2.5bn. Put another way, this analysis suggests that the Massachusetts’ corporate tax base is missing out on 20% of its revenue, revenue legitimately owed to the Commonwealth by large multinationals, but which currently goes uncollected because these companies are shifting billions of dollars of profits generated in Massachusetts to their subsidiaries located in overseas tax havens.
Note that the only actual state data point I am aware of — and it is quite noisy — is from New Jersey. As you can see from the chart below, New Jersey’s corporate tax has increased very considerably recently.
As it turns out, New Jersey conformed to GILTI in 2018. However, in 2018 New Jersey also made a variety of other changes to its corporate tax. (New Jersey taxed 5% of the repatriation, moved to combined reporting and imposed a 2.5% surcharge on its CIT for 2018 and 2019 (1.5% for 2020–21).) Thus this impressive — and seemingly durable — increase is not attributable just to GILTI conformity, but surely some substantial portion must be.
How did the PWBM arrive at its number? It relied on data from the IRS concerning the profits of US corporations reported in foreign jurisdictions, along with the assets the corporations have in those jurisdictions. For instance (and this rough example is my work, not PWBM’s), in 2016, US corporations reported earning $31 billion in Ireland, based on $80 billion of assets there. The GILTI formula assumes that a 10% return (so $8bn) is reasonable and that the rest of the income is GILTI. In the case of Ireland in 2016, that would mean $23bn in GILTI. The federal GILTI formula then assumes that only half that amount, so $11.5bn, was shifted from the US. The Massachusetts proposal follows this reasonable assumption. Suppose only GILTI attributed to firms in Ireland were brought back into the US tax base. Massachusetts represents roughly 3% of US GDP and so it might reasonably expect to apportion $345 million to itself (3% * 11.5bn). Massachusetts’ corporate tax rate is 8% (for most industries), which means that the state might reasonably expect to gain about $27 million (8% * $345 mn) only through bringing back GILTI assigned to Ireland. Ireland is but one of many overseas jurisdictions from which Massachusetts can expect to reassign to its corporate tax base shifted income identified using the federal GILTI provision.
Incredibly, the Massachusetts DOR arrived at an annual revenue estimate of $19mn for GILTI conformity. Naturally, the MTF makes much of this. However, it seems clear that somehow the DOR goofed. Indeed, the Joint Committee on Taxation has reported that the 50 largest corporate taxpayers alone reported $65bn in GILTI in 2018. Assuming that Massachusetts included half of that amount in its tax base, then that alone should have yielded the state about $78mn (32.5bn in GILTI * 3% MA share of GDP * 8% tax rate).
According to its analysis, the DOR arrived at its number by working back from how much the Joint Committee on Taxation estimated the federal government would raise from GILTI. As will be explained in a moment, this was not a great place to start, but note that when the Kansas DOR used this methodology it arrived at an estimate of about $25mn raised for Kansas. Kansas’ economy is about 1/3 of that of Massachusetts (and its tax rate is 7%). So, even using this methodology it seems DOR’s estimate is perplexing.
It is not hard to figure out how Kansas came up with its estimate. The primary revenue estimate that I am aware of for GILTI conformity at the federal level that was provided by the JCT was about $9bn for 2020. To generate $9bn in federal taxes per year at a 21% rate, there must have been about $42bn in U.S. GILTI (42 * .21 = 9), which remember is only half of total GILTI. Kansas has about .8% of US GDP and so that means Kansas would tax about $342mn of US taxable GILTI at 7%, thus generating $24mn. Hence the Kansas DOR’s $25mn estimate. Since Massachusetts’ economy is a bit more than three times bigger and Massachusetts’ tax rate is higher (8%), the Massachusetts DOR should have arrived at an estimate closer to $100mn ($42bn in GILTI * 3% * 8%).
Now, the Mass DOR used a 2% apportionment factor rather than a 3% factor. This would reduce the amount by 1/3 or to $66mn or about three times greater than their actual estimate. (I think it is defensible to use a lower apportionment factor here, but it is striking that the DOR thinks it should be so much lower than what economic fundamentals, as calculated by PWBM would indicate. Improving state apportionment factors is an issue for another time.)
Why the Mass DOR’s revenue estimate is 1/3 of what it should be based on its own methodology is a mystery. More importantly, that methodology appears to contain a fundamental and crucially important error. Remember that the federal government gives foreign tax credits, and it is these credits that were and are expected to be the primary way that taxpayers reduce their GILTI liability at the federal level. Again, there is information collected on foreign taxes paid. According to the PWBM analysis, these credits reduced federal tax liability on GILTI by 80%; this means that the Massachusetts DOR estimate is another 5 times too small because it began its analysis with the amount of revenue that JCT projected from taxing GILTI at the federal level. The JCT’s federal tax collection projection already accounts for this 80% reduction in expected revenue due to foreign tax credits. These foreign credits, however, are available at the federal level only. They do not apply at the state level –in Massachusetts or any other state. Correcting this error returns the estimate to the ballpark of $330mn year.
So the DOR estimate of $19mn appears to be based on a flawed methodology — the basic elements of which DOR describes in a brief summary. Whether other factors have been incorporated into DOR’s estimate — such as the effects of various corporate tax credits and/or net operating loss carryforwards — and these legitimately lower the final estimate, we do not know. No mention of such is made in the DOR’s summary.
It is reasonable to conclude then that Massachusetts stands to gain somewhere between roughly $300 million and $400 million annually from taxing GILTI. Suppose Massachusetts were to raise “only” $300mn from income-shifting multinational corporations through adding a few sentences to its tax code. This would remain a great deal.
