The conventional wisdom on corporate taxes holds that while the U.S. statutory rate of 35 percent is among the highest of our peer countries, widespread loopholes in the corporate tax system mean that the rate actually paid by U.S. corporations is far lower, and actually firmly in line with these international peers. And this is one of the times where the conventional wisdom is actually correct. Because of a lack of data, it’s hard to put an exact number on it, but it’s clear that the actual rate faced by U.S. corporations is far lower than the headline 35 percent rate.

But recently, some have been trying to refute this conventional wisdom by brandishing a recent Congressional Budget Office (CBO) report. Summary Table 1 in the report has led some to believe that even the effective rates that U.S. corporations actually face are high among peer countries. But as we describe in our recent paper, once you dig into the details of the measures that CBO is reporting, this isn’t actually the case. And in fact, their findings simply bolster conventional wisdom. See EPI’s analysis of the CBO report:

Does a recent CBO report contradict our findings about U.S. effective rates? No—and here’s why.

CBO recently released an updated report comparing corporate income taxes across G-20 countries (CBO 2017). Uncareful readers might be led into thinking that the CBO report overturns the empirical evidence cited above that indicates that corporate taxes actually paid by American companies are not notably high relative to international peers. But a careful read shows that the CBO report does not contradict this other evidence.

The headline findings of the CBO study claim that the United States has the highest statutory corporate tax rate, the third-highest average corporate tax rate, and the fourth-highest effective marginal corporate tax rate. However, when we dig into the study, we are able to show that the latter two claims are simply not true. We’ll take each of the CBO estimates in turn and show why they do not contradict the evidence presented in Figure C.

The statutory tax rate. As we have already noted, it is absolutely true that the United States has one of the highest statutory rates, but widespread loopholes make this rate irrelevant to the broader question of what corporations are actually paying in corporate income taxes.

The average corporate income tax rates. CBO notes in its own table that the average corporate tax rate estimated for the United States is not directly comparable with the rates for other countries. And far from refuting earlier evidence that loopholes substantially erode the corporate taxes paid by American firms, CBO’s measures of average effective tax rates actually reinforce this evidence.

To explain why, we will first start with CBO’s estimates of rates for other countries, which are far from intuitive. CBO defines these rates as the average worldwide corporate tax rates faced by U.S.-owned foreign companies (companies in which more than half the stock is owned by a single U.S. taxpayer). As a rough approximation,9 we can think of these rates as the average worldwide corporate tax rates faced by the offshore subsidiaries of multinational corporations. But this estimate of these firms’ tax rates clearly hinges on profit-shifting and on the firms’ use of the deferral loophole.

In lieu of some measure of taxable income, the denominator in the CBO estimate is worldwide before-tax profits and earnings. Now, if a multinational corporation is deferring the taxes it owes on its offshore profits, then those profits and earnings will show up in the denominator of CBO’s estimate but won’t show up in its worldwide taxes paid—the numerator. This has the effect of driving down the average rate faced by a U.S.-owned company’s foreign subsidiary, and hence makes other countries’ tax rates look low relative to those of the U.S.

Further, using the worldwide profits and earnings of the offshore subsidiary as the denominator means that the effects of further profit-shifting are also being captured. If a U.S. multinational’s offshore subsidiary shifts its profits out of that country into a tax haven with an even lower tax rate, this will likewise drive down the CBO measure of the average effective rate in other countries.

This result isn’t a surprise; it just bolsters our earlier evidence. Closing the deferral loophole and clamping down on profit-shifting would increase the average corporate income tax rates faced by U.S. multinational corporations’ offshore subsidiaries. In the CBO data, this would boost the tax rates of other countries.

Moving on to the estimated U.S. tax rate, CBO notes that the best comparison for this would be the U.S. average tax rate for the U.S. affiliates of U.S. multinationals. But because that data are not available, CBO instead uses a measure of the average U.S. corporate rate faced by foreign-owned U.S.-based corporations. Significantly, rather than using worldwide income and taxes, CBO creates this measure of average corporate rates using the U.S. taxes paid and U.S. corporate income. But this choice of sample by definition misses the effect of profit-shifting that holds down effective U.S. rates.

To see why, imagine that a U.S. multinational decides to invert (merge with a foreign company to incorporate abroad as a non-U.S. corporation) to avoid paying its U.S. corporate income taxes. It may now instead appear as a foreign-owned U.S. corporation. But the entire point of inverting is to strip income out of the U.S. and move it into the tax haven for better access to the now tax-free profits.

Since the CBO measure only takes into account U.S. taxes paid and U.S. income, the money that is stripped out of the U.S. by the inverted multinational corporation does not show up in the CBO estimate. Instead, what is left behind in CBO’s sample is those companies that haven’t (yet) engaged in accounting gimmicks and profit-shifting to avoid their U.S. income taxes.

It is thus unsurprising (and consistent with the evidence we presented earlier) that this rate looks high. As CBO notes, its measure of the U.S. average tax rate would likely be lower if it were estimated with worldwide income and taxes. It is likely further biased upward by CBO’s assumption that average state tax rates are the same as state statutory tax rates, whereas CTJ finds that average state rates are closer to half the state statutory rates due to tax breaks (McIntyre, Gardner, and Phillips 2014b). Presumably for all of these reasons, CBO includes a footnote in its table of average effective rates telling readers that the U.S. average tax rate is not comparable with the rates estimated for other countries.

The effective marginal tax rate. Finally, CBO’s measure of the effective marginal tax rate also doesn’t deal with profit-shifting and tax avoidance. CBO’s effective marginal tax rates are not based on data. They are based instead on a hypothetical corporation in a theoretical model. CBO notes that the model emphasizes two features of countries’ tax systems: the statutory corporate income tax rate and the treatment of depreciation, including the tax treatment of debt vs. equity. CBO’s estimates are certainly interesting estimates as far as those features are concerned, though CBO does not include the effects of the research and experimentation credit and bonus depreciation in its headline estimates, which CBO note in its appendix would lower the U.S. effective marginal tax rate. CBO also does not address the effects of multinational corporations’ tax avoidance (which makes the statutory rate largely misleading if not totally irrelevant), and are therefore not germane to the discussion of whether the U.S. corporate tax burden is particularly high or is a drag on growth.