May 3, 2021

GILTI and California: Show Me the Money Edition

Back of the envelope calculations indicate significant revenue can be raised if California conforms to GILTI.


[Cross-posted from Medium]

By Darien Shanske

The California Legislature is considering a bill (AB 71) that would subject 50% of a category of income derived from federal tax law, known as Global Intangible Low-Taxed Income (GILTI), to California’s corporate income tax. In short, GILTI represents an attempt by the federal government to estimate, by formula, how much income — really earned in the US — has been shifted to low-tax jurisdictions to avoid US tax.

I have written extensively about what GILTI is, why states should tax it, why they can tax it and even why this is likely to raise a lot of revenue.


The current serious consideration of including GILTI in the California corporate income tax base merits a deeper dive into the data in order to arrive at a necessarily rough range of estimates. To be clear, I am not an economist and am not offering a model. Rather, I am extrapolating from publicly available information provided by leading economists through the use of what I believe to be reasonable assumptions.

On the high side, I will start with an analysis done by the Penn Wharton Business Model (PWBM), based primarily on IRS data. According to the PWBM, the GILTI formula should produce $388 billion of GILTI at the national level in 2021. This number, not surprisingly, reflects a good estimate as to the total amount of income shifted by US-based MNCs. Based on this number, the California proposal would raise about $1.7 billion in revenue per year. To arrive at this estimate, I am assuming that 10% of (50%) GILTI would be apportioned to Ca, which is a reasonable guess based on recent reports issued by the FTB (average factor of about 11% in 2017, 8% in 2018).

New Jersey has seen its CIT flourish after conforming to GILTI (and making several other substantive changes in 2018, including a rate increase and taxing 5% of the repatriation) and so the notion that GILTI conformity can yield substantial revenues is not outlandish.

The steadily declining yield of California’s corporate income tax similarly suggests there is a substantial upside to reforming the corporate income tax base. (Observe that the small reduction in rates on its own does not look like a good explanation for such a steep decline.)


[See chart on page 3 of this document:]

(Note that this high-end estimate derived from PWBM might actually be conservative if one believes that there is going to be an economic boom over the next few years.)


What of the low end of the range? We do have one piece of actual data on the amount of GILTI that has actually been reported to the IRS, which is that 81 of the largest corporations, with ¼ of total corporate income, reported about $100bn in GILTI in 2018. Note that if we just multiply 100bn by 4, then we are back to our high-end estimate of about $400bn in GILTI per year. Yet one might reasonably believe that these large companies represent a disproportionate amount of GILTI. One might also believe that these sophisticated taxpayers have found numerous ways to reduce their tax liability — perhaps by gaming the formula or by transmuting GILTI income into something else (e.g., so-called subpart F income). In California, it might turn out that some of the taxpayers with GILTI are already not taking a water’s edge election, which take them out of the GILTI inclusion proposal because, as to these taxpayers, income shifting is being countered in another way.

Note that California conformity proposal is written in such a way that it does not conform to federal regulations governing the GILTI calculation that are very likely to undermine it further. And, California already includes a different kind of suspect income (subpart F) into which GILTI income can be transformed in its tax base and so that particular stratagem will not reduce California revenue.


And so I think a lower bound of 200bn in GILTI at the national level for 2021 is reasonable — that means that between the vagaries of revenue estimation and taxpayer strategies, aggressive and benign, the estimate from PWBM is about 100% too high. In that case, conforming to GILTI would raise about $850 million/year.

This would mean a range of between $850mn and $1.7bn annually, subject to substantial caveats.

AB 71 would also subject the deemed repatriation of IRC 965 to taxation. This is another topic I have written about. The amount of revenue that could be raised by this change is also hard to estimate. It is not clear how much revenue was deemed repatriated, how much might have already been subject to California tax and how much corporations have chosen to pay on a deferred schedule (which is what AB 71 would tax). First, as to the repatriation that was formally subject to California tax already because the revenue was actually repatriated (about $1 trillion — again, based on FTB data), I think that not much tax was already paid on this revenue because so many of these same taxpayers had (and have) a surfeit of tax credits to use against the California corporate income tax. AB 71 caps the use of these credits against the repatriation (or GILTI) at $5mn. Second, I think a conservative estimate of how much of the repatriation taxpayers chose to defer is about $1.6 trillion. This takes into account the original $2.4 trillion estimate for the total amount of revenue to be repatriated, as well as the reported fact that 2/3 of taxpayers with repatriated tax liability chose to defer their tax liability on the repatriation (and only these deferrals would be taxed). Note we do not know how much of the repatriation those taxpayers represented.

Based on these reasonable, but rough, assumptions, taxing the repatriation would result in roughly $4.3 billion in taxes paid to California over the next 5 years. Acknowledging the significant uncertainties here, starting with the original estimate of how much revenue there was to be repatriated, a lower bound of figure of $2 billion over 5 years seems reasonable or again discounting by a factor of 2.

A final note on the larger bill. One of AB 71’s core policy insights is that the state needs to provide a stable stream of revenue to combat homelessness. Tying the state’s contribution to the vagaries of how much taxpayers pay on their GILTI income would undermine that stability. What is most important about the revenue estimates is that making a change to California’s corporate tax that should have been made anyway is reasonably likely to roughly pay for the state’s revenue commitment to combat homelessness over time so that these new programs are not taking away from other essential programs.


To make this point more concrete, consider that we want taxpayers to have less GILTI because we want them to shift less income. California’s conforming to GILTI could contribute to taxpayers changing their tax structures in this way. Indeed, the Biden Administration is proposing to strengthen the GILTI regime in various ways. Suppose then that there is less GILTI income because less income is being shifted. If that were to occur, then that would mean there is more tax paid on ordinary corporate income and so California would still receive more revenue from GILTI inclusion, but not with the GILTI label.


Accordingly, it makes sense to appropriate about $2 billion/year to homelessness mitigation programs, knowing that the GILTI conformity and repatriation provisions are at least reasonably likely to roughly cover these expenditures over time if one takes a broad view of how these provisions are meant to operate.