Evan Soltas
Jul 1, 2012

He Who Laffers Last...

I do not accept the idea, in general, that a cut in most marginal or effective tax rates in the United States would result in an increase in revenues; that is, I think that we are in most cases on the left side of the Laffer curve, in which an increase in marginal or effective tax rates would tend to result in an increase in tax revenues. So do the vast majority of economists, or at least those who participate in the IGM Economic Experts Forum.

Arnold Kling recently came up with an informal classification of arguments by their intended audience and intended purpose, and this post aims to be his Type B -- that is, an "attempt to open minds of people on the same side as the author." So let me entertain the alternate possibility that Laffer dynamics are highly significant, as is suggested in capital gains tax rates and revenues data since 1954.From 1954 to 1968, the top marginal capital gains tax rate was 25 percent. Effective, a.k.a. average, rates were lower due to deducted capital losses, lower marginal rate brackets, other exemptions (such as on municipal bonds), etc. They remained within a range 13.7 and 15.5 percent, with variation driven by apparent cyclical swings -- the 13.7 percent number came in 1957, at the start of a recession, as did 15.5 in 1961, in the immediate aftermath of another. A useful, albeit incomplete, reference is the nominal Dow Jones Industrial Average during this period.

During this time, capital gains tax revenue rose fourfold in nominal terms, from $1.010 billion to $4.112 billion. That is roughly in line with nominal gains in the Dow Jones index, which trebled during the same time period. Inflation as measured by the GDP deflator was low, varying between 1 and 4 percent  year-over-year -- and because capital gains are not adjusted for inflation, an increase in inflation means an increase in tax rates on real capital gains, and often a substantial one. The combination of large capital gains and low inflation means that effective real capital gains tax rates were only slightly higher on a percentage-point basis than the nominal rates. The quadrupling of revenues during a low-and-stable real rate is consistent with the Laffer model.

From 1968 to 1972, during the Nixon administration and the final year of Johnson's, the top marginal tax rate on capital gains was hiked every year, to 26.9 percent, then 27.5, 32.21, 34.25, and finally 36.5, where it remained until 1975. During the period of rising nominal rates, revenues first rose from $4.112 billion in 1967, immediately before the hike, to $5.943 billion in 1968. But then revenues fell on a real basis, which is consistent with Laffer's point about incentives and higher long- versus lower short-run elasticities to marginal tax rates. By 1975, real revenues were half of their 1968 level; revenues as a percentage of GDP fell from 3.91 percent in 1968 to 1.89 percent in 1975.

The economic environment during this time, in fact, is suggestive of a dramatic increase in real tax rates on capital gains. Both nominal and real performance of the Dow Jones struggled in a broader environment of falling returns on investment -- the real Dow fell 50 percent during this time -- and inflation as measured by the GDP deflator soared. This combination implies a high real capital gains tax rate. Evidence from this period uniformly suggests that the increase in capital gains tax rates led to no increase in revenues, after adjustment for inflation and economic conditions.

In 1976, capital gains tax rates increased again. Important to our discussion, the changes went beyond the top marginal rate, which rose from 36.5 percent to 39.875 percent, where it remained until November 1978. The minimum marginal tax rate on capital gains, paid mainly by middle- and upper-middle-class households, rose from 10 percent -- where it had stood since 1968 -- to 15 percent. The holding period for long-term capital gains also increased from six months to one year, which shifted more capital gains into the short-term, higher-rate taxable category. This explains why the effective nominal capital gains tax rate increased from 14.7 percent in 1975 to 18.0 percent in 1978. This period amounted to another substantial increase of the real tax rate on capital gains, given extraordinarily high inflation.

And yet, the evidence suggests that revenues adjusted for inflation and economic conditions continued to decline. Real capital gains tax revenues, that is, did not recover as they "should have" given the improvement in the Dow Jones following its lows in 1975. One way to see this is to look at capital gains tax revenues as a fraction of GDP; they remained near 2.2 percent, lower than where they had been prior to the rate hike.

1978 began a historic reversal in the ever-increasing top marginal tax rate on nominal capital gains. The Carter administration cut it twice, from 39.875 percent to 33.85 percent, then to 28 percent in 1979; President Ronald Reagan cut it to 20 percent in 1981, where it stood until 1986. (We will soon discuss the rate hike in 1987.) During this time period, real tax rates on capital gains must have plummeted, as Fed chairman Paul Volcker reined in inflation. The result was a large increase capital gains tax revenues after adjustment for inflation and economic conditions. The real Dow Jones rose 30 percent from 1978 to 1985, two years before Reagan hiked the tax rate in 1987. Comparatively, real revenues from the capital gains tax rose 47.5 percent. (The 1986 data appears to anticipate the rate hike.) Another way to make this point is to look at capital gains tax revenues, again, as a percentage of GDP; they rose from approximately 2 percent before the cut to approximately 4 percent post-cut, increasing every year of the low rates.

The capital gains tax rate rose three times after 1986: in 1987 from 20 percent to 28 percent, in 1991 to 28.93 percent, and in 1993 to 29.19 percent. Once again, the evidence suggests that these increases in the tax rate on nominal capital gains resulted in lower revenues after adjustment for inflation and economic conditions. During this time, inflation as measured by the GDP deflator slowed from 4 percent to 2 percent; given the real appreciation of assets, on net it seems that this did not materially decrease the real capital gains tax rate, which therefore rose during this time. (A note about 1986 and 1987: capital gains tax revenues, nominally, in real terms, and as a percentage of GDP, appear very high. This makes sense given historically high advances in the Dow Jones, which doubled during these two years, in the context of a tax hike.) It was not until 1993 when real revenues from the capital gains tax would exceed its 1985 level, over which time the real Dow Jones more than doubled in value. Revenues as a percentage of GDP fell from 4 percent in 1985 to roughly 2.5 percent in the 1990s.

The capital gains tax was cut in 1997 under the Clinton administration from 29.19 percent to 21.19 percent, in 2003 to 16.05 percent, in 2006 to 15.7 percent, and in 2008 to 15.35 percent. Inflation was not a meaningful consideration, in terms of varying the real capital gains tax rate. The evidence suggests that these cuts either increased or did not decrease real capital gains tax revenues when adjusted for economic conditions. The path of the real Dow Jones and real capital gains tax revenue are in rough alignment during this period, and capital gains tax revenues as a percentage of GDP rose from roughly 2.5 percent in the 1990s to roughly 6 percent in the 2000s.

Overall, a close examination of capital gains tax rate and revenue data from 1954 to 2008 does not reveal a single instance in which an increase in the top or lower marginal tax rates leads to a sustained increase in revenues after adjustment for inflation and economic conditions.

(I have also discussed capital gains taxation at length in this other blog post.)