Thursday, February 26, 2009

Raising tax rates doesn't guarantee more revenues

This chart is worth at least 1,000 words on the subject of tax policy, and the Obama administration and Congress would do well to study it carefully before plunging ahead with Obama's plan to raise an extra trillion or dollars in revenues over the next 10 years by raising all sorts of taxes (e.g., higher income and capital gains taxes, new climate taxes, limitations on deductions, etc.).

Note how capital gains revenues soared in the year before the capital gains tax was scheduled to rise in 1987. Taxpayers respond significantly to incentives: if they know taxes will rise in the future they will accelerate income and accelerate the realization of gains. The CBO projected in 1986 that the rise in the capital gains tax would lead to a revenue bonanza over the next several years that completely failed to materialize. In fact, capital gains revenues were abysmally low for the next 8 years.

Note also how a reduction in the capital gains tax in 1997 resulted in a huge increase in capital gains realizations in subsequent years. I would argue that a lower tax on capital helped boost the stock market and the economy, through increased investment, and that in turn led to higher capgains realizations. Of course, then we had the stock market collapse of 2001-2002, brought on largely by tight monetary policy, which resulted in a huge decline in capgains revenue.

The Obama folks are projecting that a 2010 rise in the capital gains tax rate for the rich, to 20%, will result in an extra $118 billion in revenues over 10 years. They are also projecting that the expiration of the Bush tax cuts, and new limitations on personal exemptions and itemized deductions for the rich, will generate over $500 billion in new revenues over 10 years. If history and human nature are any guide, the actual revenue gains could be much less.

As Art Laffer always says, when you tax something less you usually get more of it; and when you tax something more you usually get less. Higher taxes on anyone, coming at a time (next year) when the economy will likely still be struggling to recover, are quite simply a very bad idea.

And this helps explain why the stock market is celebrating Obama's "new era of responsibility" with lower and lower prices.


Jon S. said...

Great chart, Scott. I wish I could say that maybe some people in the halls of Congress will see this and have second thoughts, but the chance of that happening are zero.

Maybe Kudlow will talk about this chart on CNBC -- and speaking of which, with the exception of Larry and only a few others, the cluelessness of most CNBC commentators and guests is mind-boggling. Lots of expertise on technical aspects, very little strategic wisdom. But I digress....

Mark A. Sadowski said...


The Tax Foundation's data is a little off. I prefer to use data from the following two nonpartisan sources:

One of the problems with the chart is that it makes a mention of tax revenues but they don't appear anywhere on the chart. I think that the relevant figure would be capital gains tax revenue as a percent of GDP. Here are the figures from 1978 through 2005 (I don't have 2006 handy):

Year Tax Rate Tax Revenue
1978 18.0 0.397
1979 16.0 0.459
1980 16.8 0.447
1981 15.9 0.411
1982 14.3 0.396
1983 15.2 0.529
1984 15.3 0.545
1985 15.4 0.627
1986 16.1 1.186
1987 22.7 0.711
1988 23.9 0.762
1989 22.9 0.643
1990 22.5 0.480
1991 22.3 0.415
1992 22.9 0.457
1993 23.7 0.542
1994 23.7 0.512
1995 24.6 0.598
1996 25.5 0.849
1997 21.7 0.955
1998 19.6 1.018
1999 20.2 1.206
2000 19.8 1.297
2001 18.8 0.648
2002 18.3 0.469
2003 15.9 0.468
2004 14.7 0.627
2005 14.8 0.822

I ran a linear regression on the data and came up with the following equation:

Revenue = 0.0091*Rate + 0.4838

In short, there is a positive relationship between revenue and effective tax rate over this period. Furthermore, the R-squared value was 0.0161 indicating the relationship is very poor in any case.

Furthermore, if you take the 14 years with the highest effective tax rates (1987-2000) and compare it to the other 14 you'll find that revenue from capital gains averaged 0.746% of GDP versus 0.574% of GDP the years when it was lower.

This should not be too surprising from a supply side theoretical perspective. Effective (and for that matter top marginal) capital gains tax rates have been too low to have much of an incentive effect on investments. A more likely explanation for the pattern in the capital gains revenue data is economic activity.

In the final analysis when one uses nonpartisan data there is no evidence that lowering effective capital gains tax rates increases capital gains tax revenue as a percent of GDP. On the contrary the empirical evidence suggests a modest positive relationship between effective tax rates and tax revenue over this period.

Bruce Bartlett (who worked for Ron Paul, Jack Kemp, Ronald Reagan and the CATO Institute) has a good article in Forbes today where he makes the case that tax increases will not necessarily harm the economy:

Mark A. Sadowski said...

I should also mention that the effective capital gains rate averaged 22.6% during 1987-2000 and 16% the rest of the period from 1978 through 2005. So revenues as a percent of GDP averaged 30% higher when the effective tax rate was 41.3% higher. This of course suggests that there are decreasing margins of returns with tax increases (consistent with theory), but it also suggests that we have been on the left hand side of Laffer's Curve during this period.

Scott Grannis said...

Mark, I don't know what you're doing with the data, but my chart does show realized capgains as a % of GDP and they bear no relation to your numbers.

Maybe I'm cross-eyed, but the Forbes article says Reagan proposed a huge tax increase in 1983. In fact it was a huge tax cut, taking the top marginal rate down from 70% to 35%. The economy subsequently boomed.

On that score, I would note that with top tax rates today about half of what they were in the late 1970s and early 1980s, tax revenues (federal) as a % of GDP in 2008 were approximately the same as they were back when tax rates were twice as high. If that is not empirical evidence of the power of the Laffer Curve, I don't know what is.

Mark A. Sadowski said...

First, realized capital gains are not the same as revenue from capital gains taxes. And I think the whole point the graph was supposedly making (which it clearly failed to do) was that capital gains tax increases fail to increase tax revenue (show me where revenue is shown on the graph).
Second, the massive tax increase (the largest in history) ennacted under Reagan was the the Tax Equity and Fiscal Responsibility Act (TEFRA) of 1982 which was signed into law on September 3rd. Most of its provisions took effect on Jan. 1, 1983. It did not raise marginal rates but in fact raised revenues mostly through hikes in excise taxes. (I remember it distinctly because I had a cigarette smoking friend in college who griped endlessly about its effects on the price of a pack of Camels.)
Third, the only way you can really reduce marginal tax rates and raise more revenue at current levels is by reducing exemptions, deductions and credits (closing loopholes) which is what we did with the bipartisan Tax Reform Act of 1986.

Laffer's Curve is not necessarily wrong, it's just that most marginal rates are now so low that we are on the left hand side of his curve. In fact I read one research paper that estimated that the peak amount of static revenues occurs at a marginal tax rate of about 80%. That paper suggested that the Johnson marginal tax cut would have raised revenue but no subsequent one would have raised revenue in the absence of other tax changes (closing loopholes). Raising marginal rates might increase deadweight losses from taxation, but they certainly will not reduce revenues at their current low levels.

Mark A. Sadowski said...

Don't confuse TEFRA with the Economic Recovery Tax Act (ERTA) of 1981 (otherwise known as the Kemp-Roth Tax Cut) which reduced the top marginal tax rate on unearned income from 70% to 50% and also reduced the top capital gains tax rate from 28% to 20%.

P.S. Bill Roth was my Senator.

Gene Prescott said...

As I was in active tax practice during all of those years I have personal anecdotal observations that a reduction in capital gain rates did stimulate more property with gains being sold. I consider a 'realized capital gain' as revenue. The capital gains I encountered in practice were more often real estate related than securities related although I encountered all kinds.

Jon S. said...

Rates were cut, revenue rose -- period. In 1980, at higher rates, federal receipts were $244 billion. In 1988, they were $446 billion. And there was a recession the first few years of the '80s, a big one by current standards.

Even Clinton's Council of Economic Advisers in 1994 agreed with the basic proposition under discussion here, with direct reference to ERTA: "It is undeniable that the sharp reduction in taxes in the early 1980s was a strong impetus to economic growth."

Scott Grannis said...

Mark: The only good thing about the capital gains tax is that it can be legally avoided by not selling the appreciated asset. Showing capgains realizations (i.e., decisions to sell realized gains) thus tells us exactly how much people's behavior was affected by changes in the capgains rate. Realizations change by an order of magnitude more than the rate did in the chart, so revenues must have changed in line with realizations.

Once you get up to well into the top marginal bracket, let me know how you would feel about paying 80% of any additional dollar you were able to earn. Perhaps you wouldn't mind, but I for one would not try too hard to earn more. Indeed, the prospect of higher marginal rates was one factor I considered when deciding whether to retire.

I think the only sensible income tax rate is a flat tax rate of no more than 20-25%. I believe that would maximize peoples' incentive to work, and minimize peoples' incentive to evade taxes, and thus maximize the size of the tax base and revenues to the government.