The Facts on Capital Gains

Moderator Charles Gibson made a surprising claim while asking a question during the Democratic presidential debate of April 16. Mr. Gibson wondered why Mr. Obama wanted to raise the capital gains tax rate since "in each instance [1997 and 2003], when the rate dropped, revenues from the tax increased. The government took in more money. And in the 1980s, when the tax was increased to 28 percent, the revenues went down."

Unfortunately, neither Mr. Obama nor Ms. Clinton took on the false assertion that the question presents as fact, namely that tax revenues and tax rates on capital gains vary inversely (when rates go up, revenues fall, and vice versa).

Revenues from long-term capital gains are more volatile than most other revenues because they are more subject to strategic behavior by taxpayers. A person cannot easily postpone earnings for a year or two. It is much easier to advance or postpone the sale of some assets – especially financial assets – by a few months as expectations change about future tax rates. Indeed, as we look at the history of realized capital gains, they show great sensitivity to the date of tax rate changes (see Figure 1). In 1986, especially, as the battle to raise the capital gains rate was nearing its conclusion, the wealthy realized capital gains of $338 billion dollars. That was almost double the capital gains of 1985 and a level of capital gains realizations that would not be seen again for eleven years.  Capital gains realizations collapsed in 2001, as promises of a future tax cut (a promise kept in 2003) coincided with the dot com bust and a recession.  In both 1986 and 2003, wealth owners timed their capital gains realizations to minimize their taxes over a couple of years. 

But over the longer run, these timing issues that lend credence to Mr. Gibson's point, fade out.  The 1997 capital gains tax cut was not well anticipated and so the increase in capital gains realizations and revenues in 1997 just fell in line with the trend before and after.  Sure, capital gains tax revenues rose in 1997 after the tax cut, but that just continued what had been happening in 1995 and 1996.  In fact, throughout, capital gains tax revenues are heavily dependent on the business cycle: the stock market crash of 1987, the recessions of the early 1990s and early 2000s, and the boom of the late 1990s. 

Finally, there is a strong upward trend in capital gains realizations as the economy grows. So capital gains realizations depend on a) the long-term growth of the economy, b) the business cycle and c) strategic behavior of taxpayers. In spite of substantial study of all of this, the Congressional Budget Office apologizes that,“The inherent volatility of gains realizations means that CBO’s projections are still subject to substantial uncertainty.”

As Figures 2a and 2b show, long-term capital gains tax revenues move in close alignment with long-term capital gains realizations. Both spiked before the tax increase of 1986 and both rode the business cycles of subsequent years. Both also show a strong upward trend.

One important difference emerges, however, between capital gains realizations and tax revenues. In recent years, capital gains realizations have been considerably above the long-term trend. But capital gains tax revenues have fallen below trend. Why is that? The obvious explanation is that today’s low tax rates produced sub-par capital gains tax revenue even though capital gains realizations in the mid 2000s were near historic highs.

In short, there is nothing at all remarkable about the capital gains tax. Like every other tax, its revenues depend on taxable income (realized capital gains) multiplied by the tax rate. Capital gains realizations are more volatile than most income series, and they are more easily advanced or postponed in time. But apart from a few brief peaks and valleys that anticipation of tax changes brings to the tax series, they are dominated by economic conditions that affect capital gains.

When asset prices are growing rapidly – as in the mid-1990s before the capital gains tax cut of 1997 – capital gains realizations rise steeply in spite of a high tax rate. Similarly, if asset prices are falling, as they were in the early 2000s, capital gains realizations will fall in spite of low capital gains tax rates.

Over the long haul, if we cut capital gains taxes relentlessly, as we did in 2003 following the cuts of 1997, we will depress tax revenue. By the same token, even when tax rates are historically high, as they were in 1995 and 1996, realized capital gains and capital gains revenues will surge as the wealthy stop playing games of asset sale timing and get on with their business.

Returning to Mr. Gibson’s question, it is quite true that revenues rose after the capital gains tax cuts of 1997 and 2003. They also surged in 1995 and 1996 without tax cuts, and they collapsed in 2001 – as they will again in 2008 – because of changes in the underlying economy.  It is nonsense to suggest that over the long haul cutting capital gains taxes increases revenues.

The capital gains tax is paid overwhelmingly by the wealthiest people in the United States. They are lucky to be able to present, as facts, in a presidential debate, the propaganda that is marshaled to reduce their tax load.

Data sources: Capital gains realizations and revenues from the CBO
http://www.cbo.gov/ftpdocs/70xx/doc7047/02-23-CapitalGains.pdf
http://www.cbo.gov/doc.cfm?index=3856&type=0
Capital gains tax rates from the National Bureau of Economic Research , http://www.nber.org/~taxsim/marginal-tax-rates/federal.html

Note: the “average marginal tax rate” is the marginal tax rate adjusted for who receives income. It incorporates the distribution of the income.