US Corporate Taxes I: Mystery Solved.

Melissa and I have been puzzling over the significant difference between the NIPA measure of corporate profits taxes and the revenues actually collected by the IRS from corporations. NIPA stands for “National Income and Product Accounts,” which are compiled by the Bureau of Economic Analysis to measure GDP.

NIPA, therefore, has to be comprehensive to capture all components of the economy that add up to GDP.

The IRS, on the other hand, is only interested in measuring how much tax revenues are being collected.

Let’s see how all this impacts the measurement of corporate tax revenues:

(1) The NIPA’s all-encompassing approach can be seen in the way it measures corporate profits. For example, it includes the profits earned by the Federal Reserve Banks (Fig. 1).They are incorporated entities. There are 12 of them in the Federal Reserve system. Collectively, their profits have soared as the Fed’s balance sheet was loaded up with bonds acquired through QE programs from December 2008 through October 2014 (Fig. 2). Over this period, the Fed’s balance sheet ballooned by $2.4 trillion to $4.4 trillion. That’s after it had ballooned from about $800 million to about $2.0 trillion from September to December 2008 as a result of numerous emergency liquidity facilities that were replaced by QE programs.

(2) The Fed has been earning interest on all those bonds. As a result, the NIPA show that the Fed’s profits jumped from $27 billion (saar) during Q1-2009 to a record $105 billion during Q2-2014. They’ve come down a bit since then to $82 billion during Q3.

(3) The Fed is required to return the profits from its operations, net of expenses, to the Treasury. This item is buried in Table 4 of the Monthly Treasury Statement of Receipts and Outlays as “Miscellaneous Receipts: Deposits of Earnings, Federal Reserve System.” Not surprisingly, the NIPA measure of the Fed’s profits tends to coincide with the 12-month sum of the Table 4 item.

(4) So if we add the 12-month sum of the Fed’s earnings as reported by the IRS to the 12-month sum of the corporate tax revenues reported by the IRS, the result is a series that is closer to the NIPA measure of total corporate tax revenues (Fig. 3).

(5) The remaining difference between the NIPA and IRS series, as adjusted by us, is mostly attributable to state and local taxes on corporations, which are included in the NIPA measure but not in the IRS tally (Fig. 4).

US Corporate Taxes II: Not So Taxing. The obvious conclusion is that measuring the average effective federal corporate tax rate using the NIPA data will overstate it by the amount of “taxes” collected from the Fed and by the amount of taxes paid to taxing authorities other than the IRS.

Nevertheless, the comprehensive NIPA-based effective corporate rate has been below 25.0% since Q1-2008 (Fig. 5). It was 20.7% during the four quarters through Q3 of this year. Even lower is the comparable tax rate based on IRS data on federal corporate tax receipts, excluding the Fed’s contribution. This tax rate has been below 20.0% since Q2-2008, and was just 13.0% over the four quarters through Q3!

That’s well below the current statutory rate of 35.0%. It’s also well below the 20.0% rate that Republicans are aiming to enact as part of their tax reform plan. This raises two relevant questions:

(1) Are we comparing apples and oranges? Melissa and I now feel quite comfortable using the IRS corporate tax revenues measure as the numerator in calculating the effective federal corporate tax rate.

But what about the denominator? Removing the Fed’s profits from the NIPA measure of corporate profits in the denominator doesn’t make much difference. That’s because those profits are a much more significant percentage of corporate tax revenues than corporate profits.

We are investigating the possibility that the treatment of sole proprietorships and LLCs in the NIPA data might be overstating the denominator relative to the numerator of our effective tax calculation. We doubt it since the NIPA include sole proprietors’ income in personal income rather than in profits, as we discussed last Tuesday (Fig. 6). We presume (and are checking whether) the same holds true for the taxes paid by proprietors. We are quite certain they are included in the IRS and NIPA measures of individual income taxes.

(2) Are small corporations paying full fare while large corporations are free-loading? The IRS-based effective federal corporate tax rate suggests that corporations are using all sorts of tax deductions and dodges to very effectively lower their effective tax rate well below the 35.0% statutory rate.

The urban legend is that large multinational corporations have the resources to play this game, and are paying almost nothing in taxes, while smaller corporations are paying something close to 35.0%. I asked Joe to find the amount of taxes paid by the S&P 500 corporations. He found annual data showing that for most years since the late 1990s, taxes paid by these companies actually exceeded the federal corporate tax receipts at the IRS (Fig. 7). Their effective tax rate was 26.4% during 2016 (Fig. 8). That may reflect that the S&P 500 data aren’t strictly comparable because they include taxes paid to other entities, including state and local governments, as well as to overseas taxing authorities.

The bottom line is that getting to the bottom line when it comes to matters of taxation is a very taxing exercise. We’ll keep at it, but our conclusions so far are that corporations, on balance, may actually be paying less than the 20.0% statutory rate that the Republicans are aiming to enact, and large corporations may not be free-loading at the expense of small ones.

Dr. Ed Yardeni is the President of Yardeni Research, Inc., a provider of independent global investment strategy research.