August 5, 1996

Taxes: The Highest in History?

Some columnists claim that tax receipts are now a bigger share of the economy than ever before, presumably the result of tax increases enacted in 1993. As a result, they say, federal taxes and especially federal income taxes should be cut. But the claim is false.

Here is a sample of the argument, from an April column by Bruce Bartlett [1]:

"Recently released data from the Census Department reveal an amazing fact. In 1995, total taxes as a share of gross domestic product were the highest in U.S. history...Virtually all of the growth in the total tax burden [federal, state, and local revenues, from 1992 to 1995] is accounted for by the rise in federal receipts, which reached 20.4 percent of GDP in 1995. This is the second highest level in history...The growth in federal taxes since Mr. Clinton took office is remarkable...there has been a rise of 1.2 percentage points in the federal tax burden just during Mr. Clinton's first three years... Federal income taxes could be cut 15 percent across the board and total federal receipts as a share of GDP would be only about the same as they were in 1992."

Receipts as a Share of the Economy

Measuring tax receipts by portraying them as a percent of GDP is the best way to compare tax levels over time since it adjusts for the growth of the economy. For this reason, when Mr. Bartlett discusses the level or growth of receipts, he always means as a percent of GDP. This analysis adopts Mr. Bartlett's shorthand: all references to the level of receipts means as a percent of GDP.

There are at least four points to be made about Mr. Bartlett's argument, each of which will be examined at more length in this analysis.

Non-standard definitions. The standard definition of federal revenues used by the Congressional Budget Office and the Office of Management and Budget includes only those revenues generally considered to be taxes. Mr. Bartlett does not use this standard definition, but instead relies on estimates of government receipts published by the Commerce Department. What Bartlett does not explain is that the Commerce Department estimates include Medicare premiums and other amounts not generally considered as tax receipts. Because Medicare premiums and other non-tax items included in the Commerce estimates have grown substantially in recent years, the Commerce data that Bartlett cites overstate both the total amount of tax receipts and the rate at which they have grown over the decades. (It also may be noted that under Bartlett's definition, Congressional proposals last year to increase Medicare premiums would constitute tax increases.)

Receipts are not much higher than historical averages. Using the standard CBO definition, federal tax receipts reached 18.9 percent of GDP in 1995, not the 20.4 percent figure cited by Mr. Bartlett. This is not the highest in history or even the second highest. Federal tax receipts have fluctuated over the past 30 years, exceeding 19 percent in both the early 1970s and the early 1980s. Average tax receipts over the past five years equal 18.2 percent of GDP, only trivially higher than the 18.1 percent average rate over the previous three decades.

What about 1992? Mr. Bartlett exaggerates the recent growth in tax receipts by comparing current tax receipts to tax receipts in 1992. Tax receipts were considerably depressed in 1992 because of the recession. Now that the economy is operating at capacity, receipts are no longer depressed. CBO estimates that tax receipts would have risen about 0.3 percentage points from 1992 to 1995 even if tax law had remained unchanged. An improving economy is a blessing, not a curse.

Most taxpayers have not seen their federal taxes increase in recent years. The tax changes enacted in 1993 account for some of the increased revenue in recent years. The 1993 tax increase, however, applied overwhelmingly to high-income taxpayers. The vast majority of American taxpayers saw no increase at all in their income taxes and were touched only by a 4.3 cent increase in the gas tax, which costs a typical household $38 per year.

These points are described in more detail below.

How are Tax Receipts Measured?

The Congressional Budget Office and the Office of Management and Budget define federal government receipts as the total of individual income taxes, corporate income taxes, social insurance taxes and contributions, excise taxes, and selected other receipts, which include estate and gift taxes, and custom duties and fees. In other words, government receipts include all revenues commonly considered to be "taxes."

The Commerce Department collects data on government receipts and expenditures as part of its National Income and Product Accounts. NIPA data on government receipts differ from CBO and OMB measures of government receipts in several important respects. Most important, NIPA measures of government receipts include several accounts that are not considered taxes by CBO or OMB. The most significant of these accounts are Medicare premiums and government contributions to federal employee retirement funds.

Over the past 30 years, both Medicare premiums and government contributions to federal employee retirement funds have grown significantly, widening the gap between NIPA and CBO/OMB measures of government receipts. In the early 1960s when those two activities represented a negligible proportion of NIPA receipts, NIPA and CBO/OMB measures of federal receipts were virtually identical. By 1994, however, receipts as measured by NIPA were $100 billion higher than receipts as measured by CBO and OMB. Of that difference, more than $85 billion was due to the inclusion of Medicare premiums and government contributions to federal employee retirement funds in the NIPA measures of receipts.

The most recent Republican budget plan included proposals to increase Medicare premiums. If this proposal were to be enacted, federal government receipts as measured by NIPA — and cited by Bartlett — would climb to a still higher percentage of GDP.

Relative to the standard CBO/OMB measures of taxes, NIPA measures overstate the increase in tax receipts as a percentage of GDP. Total NIPA receipts grew from an average of 18.1 percent of GDP in the mid-1960s to an average of 19.6 percent in the early 1990s, an increase of 1.5 percentage points. In contrast, CBO and OMB measures of government receipts increased from an average of 17.6 percent to 18.1 percent over the same period, an increase of only 0.5 percentage points, or one-third as much. While there has been some growth in total tax receipts as a percentage of GDP, the growth has been modest.

How Do Current Taxes Compare with Historical Levels?

Both NIPA and CBO/OMB measures of tax receipts have increased modestly over the past 30 years. When the cyclical fluctuations caused by the business cycle are removed, CBO/OMB tax receipts as a percentage of GDP have changed little in recent decades, ranging from an average of 17.6 percent of GDP in the 1960s, to 18.5 percent in the 1980s. (NIPA measures showed a greater increase because of the inclusion of non-tax receipts, rising from 18.2 percent in the 1960s to 19.8 percent in the 1980s.)

Furthermore, CBO and OMB estimate that tax receipts under current law will edge down slightly in the coming years as the current unusually high level of business profits recedes to normal levels. CBO and OMB project that by 2000, tax receipts will decrease from their current level of 18.9 percent of GDP to 18.5 percent of GDP, a level comparable to the average level of tax receipts in the 1980s and only slightly higher than the 18.1 percent average level of tax receipts over the past 30 years.

How Big Was the 1993 Tax Increase?

Despite evidence showing total tax receipts are only slightly higher now than average tax receipts over the past 30 years, the tax increase signed into law by President Clinton in 1993 is sometimes described as "the largest tax increase in history."

After controlling for cyclical effects, CBO data show that total tax revenues increased by 0.8 percent of GDP between 1992 and 1995. This is far from the first time tax receipts have increased by 0.8 percent of GDP in a three-year period. In fact, in the past quarter century alone, tax receipts after controlling for cyclical effects have increased by increments larger than 0.8 percent of GDP over a three year period at least six times. The table below shows some of the increases in tax revenues over the past 25 years.

Not all of the tax receipt increases over the past 25 years were due to legislated changes in the tax code. Many of the revenue increases shown occured simply because tax brackets were not indexed to inflation.

What about 1992?

The year 1992 is often used as a base comparison year by commentators alleging that taxes have grown explosively in recent years. But 1992 is not an appropriate comparison year. Because of the recession of the early 1990s, tax receipts were unusually low during 1991 and 1992. As the economy emerged from the recession, tax receipts increased in substantial part because of economic growth, not simply because tax rates on the wealthiest taxpayers were raised.

After the tax increase of 1993, tax receipts increased from 17.7 percent in 1992 to 18.9 percent of GDP in 1995. According to CBO, if the 1993 tax increase had not taken place and tax laws had remained unchanged, tax receipts would have increased about 0.3 percent of GDP — to 18.0 percent — simply as a result of the improved economy.

Similarly, tax receipts were $97 billion higher in 1995 than CBO estimated they would be when the Clinton Administration came into office in 1992. CBO estimates that fully half of this increase is due to economic growth.

Because short-term fluctuations in the economy related to the business cycle can disguise long term trends, it is misleading to compare tax receipts from a recession year to tax receipts three years into a recovery. To gauge the magnitude of current tax levels relative to the recent past, current tax receipts must be compared to a base year that represents a similar point in the business cycle, or to long-term trends adjusted for cyclical variations. When this is done, tax receipts are found to have climbed only modestly and not to have skyrocketed.

How High are Taxes Paid by Typical Taxpayers?

The 1993 budget legislation did increase federal tax receipts. One can not, however, draw from this the conclusion that taxes increased significantly for the majority of taxpayers. The 1993 changes in the tax code increased federal income tax rates only for high-income taxpayers.

Because taxes paid by wealthy taxpayers increased significantly, average tax burdens climbed. But this tells nothing about the taxes paid by the typical taxpayer. Consider four middle-class families with taxable incomes of $30,000 and one wealthy family with a taxable income of $500,000. The four hypothetical middle-income families in the middle of the income spectrum each paid $6,000 — or 20 percent of income — in federal taxes both before and after the 1993 tax code changes. The wealthy family paid $140,000 — or 28 percent of income — before the tax code changes, and $160,000 — or 32 percent of income — after the changes. The average tax increase paid by all five families was $4,000. But all of this increase was borne by the one wealthy family.

As this example illustrates, using the increase in average tax payments produces a misleading picture of what has happened to the typical family tax burden. In this example, the typical family — the family that falls in the middle of the income distribution, with half of families earning more income and half earning less — pays no more in federal income taxes before the 1993 tax code changes than after.

The vast majority of taxpayers saw no change in their income taxes as a result of the 1993 law. CBO estimates that most households paid only $38 more per year, as a result of the 4.3 cent per gallon increase in the gas tax.

Other Factors Leading to Increased Tax Receipts

Tax rates are not the only factor that determines tax receipts; individual income taxes also fluctuate as a result of changes in taxpayer behavior. Many market analysts believe that part of the increase in federal tax receipts in 1995 resulted from investors selling off profitable stocks and paying capital gains taxes on the profits [2]. Some investors may have held off selling stocks after the election of the Republican Congress in 1994 in the hopes that Congress would quickly pass a reduction in the capital gains tax. If these investors concluded such a tax cut was unlikely, they may have begun selling their stocks by the end of 1995. Stock prices had soared by 1995, yielding large profits to investors — and a surge in capital gains tax receipts when these stocks were sold. The sale of such stocks would increase receipts, but have no impact on the taxes paid by the typical taxpayer. Nor should the decisions of investors to sell stocks in 1995 and realize large profits lead to the conclusion that tax burdens have reached unprecedented heights, only that the stock market has.

Footnotes

2. See Jackie Calmes, "Scary Deficit Forecasts For Clinton Years Fade as Tax Revenue Grows," The Wall Street Journal, August 1, 1996.