By Darien Shanske
[Cross-posted from Medium]
The United States used to tax multinational corporations (MNCs) on the basis of their worldwide income, except that most foreign source income would only be taxed when actually repatriated to the United States. This structure naturally created considerable incentive to strip income out of the United States and then not to repatriate it.
Now, thanks to the TCJA, the United States is ostensibly only going to tax MNCs on their US source income. This shift to a so-called territorial system means, of course, that MNCs will continue to have incentive to shift income abroad in order to avoid US tax. The TCJA has two separate provisions meant to counter this: GILTI and BEAT. One question for the states is whether they should conform to these provisions. I think the answer is yes. I also think it is clear that states should improve these provisions when they adopt them.
There is a preliminary question as to whether states can — as a matter of federal constitutional law — conform to these provisions. I think the answer is again yes, but the details of the state law will matter a great deal.
Not everyone agrees. For instance, the STAR Partnership, argues here that:
“Under the prior federal tax system, States generally did not include foreign income in their own tax bases, for both policy reasons and Constitutional limitations. States should continue to respect these policy and Constitutional rationales and avoid taxation of foreign income by excluding these provisions from their business tax bases.”
In short, according to a group of top state and local tax lawyers and policy wonks — who are quite explicitly representing the “business community” — states should not and cannot conform to the federal provisions meant to prevent base erosion. It is true that states do not tax foreign source income (in general) and that it is a tricky policy and constitutional question whether they can and/or should, but is it true that this is what states would be doing if they tax the repatriation or conform to GILTI or BEAT?
Consider some well-known and curious facts about the income of MNCs. First, there are some jurisdictions that are just unbelievably profitable for corporations. As Torslov, Wier and Zucman put it: “Foreign corporations . . . have extremely high profitability ratios in tax havens, e.g., 800% in Ireland. . . . By contrast, and strikingly, in almost all non-haven countries foreign firms are less profitable than local firms. Thus, there is a clear trace in global macro data of shifting from high- to low-tax affiliates, in such a way that profitability is systematically over-stated in tax havens and under-stated elsewhere.”
Many of those some super-profitable jurisdictions also house profits worth many multiples of their GDP, which would seem to be a real head scratcher. For example, the profits of US controlled subsidiaries in the Cayman Islands represented over 1000% of that island’s GDP in 2014. It is of course not possible for firms to earn profit in a jurisdiction many times the size of the jurisdiction’s economy.
We know why this money is taking a Caribbean vacation and we also know which legal structures got it there — see here for example.
In short, there was already a great deal of base erosion, and the two new provisions of federal law we are discussing are meant to counter it. (Indeed, there is a good argument that these provisions are roughly consistent with the anti-base erosion principles championed by the OECD.)
If a state were to conform to these base protection provisions or tax the repatriation, then a state is not — counter to the STAR Partnership — trying to tax “foreign” income, but trying to protect domestic income from being stripped out of their tax base.
This fact has an important legal implication. Again, it is true that the legal questions would be trickier for states looking explicitly to tax foreign income. Matters are different if a state is trying to tax domestic income and is simply trying to craft rules that lead to a better reflection of domestic income. As to designing their own tax systems to tax domestic income, states have a lot of leeway.
So, states can conform to these federal backup taxes and they should, but it is important to note that they should and could also improve these taxes. For more on how states should tax the repatriation, see here and kudos to NJ for doing so.
Here are a couple of examples of possible improvements to the backup provisions that are drawn in particular from Rebecca Kysar and Dan Shaviro. GILTI stands for “global intangible low taxed income” and the intuition behind the provision, consistent with the evidence above, is that certain low-tax jurisdictions are suspiciously profitable. Therefore, GILTI income is income that is earned beyond the normal expected return on an investment. There are numerous complexities involved in GILTI, many of which are not resolved, but for our purposes we can note that GILTI assumes a normal rate of return of 10%, which is rather high. A straightforward fix that a states could adopt would be conform to GILTI but use some lower rate tied to actual normal returns — perhaps a long-term corporate bond rate +2%. That would make the rate 6% at the moment.
As for the BEAT, the “Base Anti-Erosion Tax,” this is a minimum tax much like the Alternative Minimum Tax (AMT) for individuals. The idea is that a primary way of shifting income abroad is for MNCs to take large deductions for payments to related foreign corporations — say a US corporation paying a foreign affiliate in a low-tax jurisdiction for the use of intellectual property. The BEAT requires an MNC to add such payments back into the tax base; that broader base is then subject to a 10% rate. If the resulting tax liability is greater than the regular liability, then the taxpayer pays the minimum. As with GILTI, the BEAT has many design flaws. For instance, the BEAT only kicks in for relatively large corporations ($500 mn plus in gross receipts); it would make a lot of sense for states to adopt the BEAT, but at a lower threshold.
The TCJA emerged from such a flawed process that even its most reasonable ideas, like GILTI and the BEAT, are deeply flawed. (Again, don’t take my word on it, ask Kevin Brady.) And the states will be negatively impacted if the TCJA encourages MNCs to strip out even more income because the GILTI and the BEAT are ineffective. So, at a minimum, states should conform to these provisions and ideally they should improve them, but it should be remembered that improvements need not be through tinkering with the mechanics of these provisions.
The states pioneered an anti-income stripping regime that I would argue is simpler and likely more effective than even an improved GILTI and BEAT: mandatory worldwide combination using the single sales factor. This approach does not require the government to identify a particularly problematic form of income or deduction, which can have unexpected and undesired consequences. Instead, all income is income, and income is apportioned according to where an MNC makes its sales. We know that it is unlikely that an MNC makes many sales to final consumers in tax haven jurisdictions. Currently, no states require mandatory worldwide combination, but many permit taxpayers to choose it. If a state is unwilling to require worldwide combination, then an appealing option would be to conform to GILTI and the BEAT, with improvements, and then permit MNCs to choose if they would not prefer the simplicity of worldwide combination.
