A Washington Post column this morning repeats what has become a standard entry on three-by-five cards routinely distributed around this town: “…50 years of economic history show that raising the capital gains tax backfires: It reduces federal revenue, while lowering the tax raises revenue.”
There are several problems with this assertion, perhaps most notably that in the last 50 years, we have increased tax rates on capital gains only twice – in 1969 and 1986. (Footnote: The 1990 and 1993 deficit-reduction tax increases also caused very small increases in the maximum tax rate on capital gains. More on that later.) The 1969 tax rate increase was highly complex, and applied only to taxpayers with more than $50,000 in capital gains. Back in 1969, $50,000 was real money. (Of course, even today, $50,000 is real money to most people. But just to provide some realistic scale using the standard consumer price index, $50,000 in 1969 would be worth more like $325,000 today; and the 1969 law applied only to people with $325,000 (today’s dollars) in asset appreciation – not the total value of the assets, just the appreciation.)
In contrast, the 1986 law change was broad and generally applicable. It is what is routinely cited by some as the “economic history” that demonstrates that increasing capital gains tax rates reduces revenue. However, that conclusion is based on a highly selective reading of the facts.
From the very beginning of the 1986 legislative process, it was clear that if tax reform in fact occurred, the tax preference for capital gains would be at least reduced. The prior law provided a 60 percent exclusion for long-term capital gains. Senator Bill Bradley’s proposed 1982 “Fair Tax Act” would have eliminated that preference entirely. The so-called “Treasury I” proposal by the Department to President Reagan in 1984 eliminated the exclusion in exchange for inflation indexation of the cost basis of assets, which would have increased taxes significantly for comparatively large gains accrued over comparatively short holding periods. The “Treasury II” proposal by President Reagan in early 1985 cut the exclusion to 50 percent. The 1985 House bill cut the exclusion to 42 percent. And then the 1986 Senate bill (which passed the Finance Committee unanimously on May 6), and the conference agreement (which was passed on August 16), eliminated the exclusion entirely – circling back to the original Bradley proposal.
Throughout this process, the designers of the 1986 Act, including Bradley, Senate Finance Committee Chairman Bob Packwood, and House Ways & Means Chairman Dan Rostenkowski, recognized that the significant reduction in the capital gains preference could raise questions of fair transition. Accordingly, they wrote the law with a prospective effective date (January 1, 1987), so that taxpayers who had contemplated selling assets using the prior law preference would have a window to do so. This, of course, created a “fire-sale” opportunity for realization of gains at what was in effect a temporarily low rate. Realizations in 1986 were expected to increase substantially, and they did. Revenues from capital gains realized in 1986 were fully double the level of 1985 – the largest increase on record by a factor of two. But that was fully intended and expected to be a one-time phenomenon, as should be obvious from the following chart:
So the big, temporary jump in capital gains realizations in 1986 was fully expected, and even intended. Comparisons of later data to this one-time event clearly are not instructive. The open question instead was what would happen to capital gains tax revenues in 1987, given that many of the gains that were realized in 1986 likely were accelerated from 1987 (and even later years) to take advantage of the “fire sale” on the expiring exclusion. Opponents of the 1986 Act predicted that capital gains tax revenues would decline sharply – because of the increase in the effective tax rate on gains, even beyond the clear incentive for people to realize gains in 1986 instead of in 1987. But surprising to them, tax receipts on capital gains were higher in 1987 than they were in 1985, the prior law’s last “normal” year – i.e., without the influence of the transition to taxation of gains as ordinary income. This was true both in dollars and – even more significantly – as a percentage of GDP. In fact, in 1987, tax revenues from capital gains hit the largest percentage in recorded history, discounting the “fire sale” year of 1986 – and then were superseded immediately by a new record tax revenue year in 1988.
In 1989, three years after the 1986 Act, the economy began to enter a recession (it started officially in July of 1990), and capital gains tax revenues declined. This is of course what the economy did and what capital gains tax revenues did in 1981 and 1982 – when capital gains tax rates were cut. The obvious lesson is that levels of capital gains tax rates do not have much influence on capital gains tax revenues – clearly not in the highly vaunted “supply-side” direction – but business cycles do. And furthermore – and highly important in this context – the occasion of a change in tax rates can cause substantial temporary changes in taxpayer behavior that have little or no importance in the long run.
In fact, this conclusion extends to the 1969 rate-increase experience, as well as the instances of capital gains tax cuts, as is clear from the following chart (which shows the entire available history of Treasury statistics on tax revenues on capital gains):
As with the 1986 Act, the 1969 tax rate increase was made effective prospectively in 1970. That provided an incentive to realize capital gains in 1969, and 1970 realizations were lower. However, the economy already had entered a recession in December of 1969, related to the financing of the Vietnam War and the Federal Reserve’s action to control inflation. The 1978 capital gains tax rate cut has been hailed by some as a major supply-side success, but as can be seen in the data, the revenue effect was minimal, and not clearly positive – in fact, immediately after the second and final dose of that tax cut, tax revenues declined for three straight years. Capital gains revenues did increase two years after the 1981 capital gains (and general) tax rate cuts, as the economy recovered from the 1981-82 recession. However, after the 1986 capital gains tax increase, capital gains tax revenues continued to increase along the same trend (again, with a large spike in the 1986 “fire-sale” year). The 1990-91 recession flattened capital gains tax revenues – and everything else in the economy. Capital gains revenues increased sharply after the 1990 and 1993 deficit-reduction tax increases (which together increased effective marginal tax rates on capital gains by a little over one percentage point – a much smaller change than for any of the other instances cited here). Revenues continued to increase, at about the same rate, after the 1997 capital gains tax cut – for three years, then falling precipitously with the 2001 recession. Revenues increased again after the 2003 tax cut, as the economy eventually recovered from the recession, but plunged again with the financial crisis of 2008. Unfortunately, the data end there.
In sum, larger macroeconomic forces are the primary drivers of realizations of capital gains (and surely asset appreciation in general). To a lesser degree, changes in capital gains tax rates, much more than differences in their levels over time, can induce taxpayers to change their behavior – which usually means to accelerate or postpone the realization of gains, much more than to realize or not realize gains in an absolute sense.
And no, increasing capital gains tax rates does not decrease tax revenues – no matter what the three-by-five card says.
So if increasing the tax rates on capital gains does not mean death to tax revenues or to the economy, what lesson should we take for tax policy decisions going forward? The answer is that reducing the capital gains preference should be on the table, if it would aid overall tax reform, deficit reduction, and tax equity. (CED recommended ordinary-income taxation of capital gains in our tax-reform proposal.) More on that broad question in a future post.