GAO Compares Apples to Oranges to Find Low Corporate Effective Tax Rate

A new study by the Government Accountability Office (GAO) claims the corporate effective tax rate (ETR) was 12.6 percent in 2010, which is about half the standard estimate found in other studies cited by the GAO and summarized here, here, and here. Based on IRS data, the corporate effective tax rate is about 26 percent on average, though it dropped in the most recent year of data, 2009, to a little over 22 percent, due to the recession and temporary tax incentives meant to stimulate investment.

While this is still considerably lower than the statutory federal corporate tax rate of 35 percent, nearly all of the difference comes from the fact that U.S. corporations by law are given a tax credit against their U.S. tax liability for foreign corporate taxes paid. This is to prevent double taxation of foreign income. When the foreign tax credit is accounted for, and these foreign corporate taxes are counted, the overall effective corporate tax rate on the worldwide income of U.S. corporations is about 32 to 33 percent.

So how did GAO come up with such a low effective tax rate? Mainly by comparing apples and oranges. Particularly, GAO takes the smallest measure of taxes paid and divides it by the largest measure of net income according to financial statements, even though this net income is not the tax base that the corporate tax was meant to apply to. The corporate tax rate applies to taxable income, as defined in the tax code. According to GAO, taxable income in 2010 was $863 billion for profitable corporations, while financial statement income was $1.443 trillion. The difference between taxable income and financial statement income varies from year to year due to timing differences in the treatment of cost recovery and losses, temporary tax provisions, etc. Other researchers who use financial statement data account for these differences in various ways and they generally find corporate ETRs of about 26 percent. GAO’s ratio of 12.6 percent simply does not take these differences into account.

Nonetheless, many in the press depict this is as a statement of how little corporations pay relative to what they should pay under the law. Nothing could be further from the truth.

The specific problems with the GAO report include:

1. We can’t check the numbers since the GAO report does not include all the data used in their main calculations. Based on the few numbers provided in the text, all calculations appear to be slightly off, perhaps due to rounding error but in all cases my calculations produce higher ETRs. Also, the GAO report fails to provide any data on total worldwide income taxes to compare against total worldwide income. However, based on the tax rate given, 16.9 percent, GAO must be assuming the total of foreign, state and local income taxes to be $62 billion. In contrast, our analysis of IRS data indicates foreign income taxes alone are at least $100 billion in every year since 2004.
2. One big problem, as mentioned, is failing to account for foreign taxes properly, and foreign income. GAO's 12.6 percent figure comes from dividing U.S. federal corporate tax by worldwide income. It does not count foreign taxes at all or the foreign tax credit against U.S. tax liablity. In a 2008 report, the GAO did properly account for these things, by separating domestic and foreign income as well as domestic and foreign taxes. They found that federal corporate tax as a share of corporate domestic income was 25.2 percent.
3. Another big problem, as mentioned, is using a measure of income that has little to no relationship to the tax base to which the corporate tax applies. GAO uses income as reported on financial statements, which is different from taxable income reported on tax returns for many legitimate reasons that GAO describes. The big differences are timing differences in the treatment of cost recovery, losses, and other items. Researchers who use financial statement data generally get around these differences by averaging multiple years of data, but the GAO says they have insufficient data to do such an average. That means the 12.6 percent figure is completely unrepresentative, and should not have been published.
4. Compounding the problem of using a snap-shot single year of data, GAO does not count companies with losses, but does count them when they return to profitability but have low taxes because of losses carried forward from the years in which the companies were not counted. This skews the data significantly in 2010 due to massive losses incurred during the 2008-2009 financial crisis. The GAO claims that only profitable corporations should be counted, since only they pay taxes, but this is not accurate. Corporations and the IRS often take years to resolve tax settlements, and current payments can and often do occur independent of current income. When including all corporations, including those with losses, the GAO finds that worldwide effective tax rates are roughly equal to the statutory rate of 35 percent in 2007 and 2009 (2008 is “not available”, see Figure 5).

Follow William McBride on Twitter @EconoWill