I just don't have the time today to come fully to grips with what the House bill does to U.S. companies' foreign source income (FSI), as it's complicated and requires studying 100 pages of statutory language that's full of newly defined terms and cross-references.
But, if I am reading it correctly, while generally creating an exemption system, it appears to address base erosion by taxing 50% of certain "foreign high returns." This involves (1) computing a rate of return, based on relevant capital invested, following complicated rules for the treatment of interest flows, etc., (2) comparing that rate of return to the 3-month Treasury rate, and (3) taxing 50% of the excess over a Treasury return plus 7%. It appears that commensurately scaled down foreign tax credits will apply to the taxable piece.
This looks like a version of the approach that I have proposed with regard to taxing FSI. In short, "foreign high returns" face a U.S. marginal tax rate (MTR) that is in between zero and the full U.S. rate, and get a marginal reimbursement rate (MRR) for foreign taxes that is less than 100%, given the foreign tax credit scaledown for the excluded portion. So one gets intermediate rates, as I have been advocating, for both the MTR and the MRR. (By intermediate, I mean an MTR between zero and the full domestic rate, and an MRR between the MTR and 100%.)
Suppose, where this rule applies, that a US company earns 15% (reflecting rents) and the US Treasury bond rate is 2%. Then the excess return is 6% [i.e., 15 - (2+7)], and half of this is taxed at 20% (the bill's corporate rate), which is equivalent to saying that the entire 6% extra return is taxed at a 10% rate. Given that 3/5 is exempted (under my numbers) and the rest is taxed at half the domestic rate, the overall tax rate for the FSI here is 1/5 of the domestic rate (4%). And if the foreign tax credit aspect works as I would presume it is meant to work, there would be, in effect, a 1/5 foreign tax credit and 4/5 at least implicit (whether or not literal) foreign tax deduction under these facts.
Exempting something that is based on the normal rate of return, defined in terms of bond rates, raises concerns about workability, but otherwise has good arguments in its favor.
It's dangerous to comment on this without understanding it more fully than I do at this point. But insofar as the "foreign high return" rule is a key operating feature at the end of the day, and is not overly gamable, they certainly could have done a whole lot worse.
Note also that, while they exempt the Treasury return plus 7%, there are U.S. companies with giant rents that might actually face non-trivial liability under this, again depending on how well the rate of return computation can be done in practice.
I'll also admit that I'm pleased to see ideas that I've advocated, and that I think make some sense, appearing to have some actual policy traction on Capital Hill.
NOTE: I've edited this entry to reflect my initially getting the provision a bit wrong on a quick read. Thanks to a Twitter reader (whom I thanked on Twitter) for setting me straight.