Gregory Mankiw tries to bring Paul Krugman up to speed on the incidence of taxation, arguing that looking at Mitt Romney's low capital gains rate and ignoring any imputed corporate taxes paid by his underlyig investments is not accurate. Regrettably, Mankiw is violating an old, established rule of blogging - don't come between a man and his tirade. Krugman is going to beat to death the horse he rode in on, and delivers howlers like this in the process:
If capital gains and other investment income didn’t receive special treatment, we’d be getting substantially more revenue.
Do tell. In most cases the decision to incur a capital gains tax is entirely voluntary and is based on the decision to hold or sell an appreciated asset. The CBO tackled this in a 2002 paper, noting that higher capital gains rates seemed to reduce the realization of capital gains, particularly in the short ru (so who's Laffing now?):
The sensitivity of realizations to gains tax rates raises the possibility that a cut in the rate could so increase realizations that revenue from capital gains taxes might rise as a consequence. Rising gains receipts in response to a rate cut are most likely to occur in the short run. Postponing or advancing realizations by a year is relatively easy compared with doing so over much longer periods. In addition, a stock of accumulated gains may be realized shortly after the rate is cut, but once that accumulation is "unlocked," the stock of accrued gains is smaller and realizations cannot continue at as fast a rate as they did initially. Thus, even though the responsiveness of realizations to a tax cut may not be enough to produce additional receipts over a long period, it may do so over a few years.
...In projecting realizations beyond the current year, CBO gradually moves them to their historical level relative to output, adjusted for the tax rate on gains. That latter adjustment recognizes that with lower tax rates--even in the long run--realizations should be higher relative to GDP than they would be with higher tax rates.
Of course, higher realizations at a lower rate may or may not increase long-term revenue.
They also admit that the evidence on both sides is murky:
Because of the other influences on realizations, the relationship between them and tax rates can be hard to detect and easy to confuse with other phenomena. For example, a number of observers have attributed the rapid rise in realizations in the late 1990s to the 1997 cut in capital gains tax rates. But the 45 percent increase in realizations in 1996--before the cut--exceeded the 40 percent and 25 percent increases in 1997 and 1998 that followed it. Careful studies have failed to agree on how responsive gains realizations are to changes in tax rates, with estimates of that responsiveness varying widely.
...Estimates of the revenue effects of capital gains tax changes by the Congress's Joint Committee on Taxation (JCT) and the Treasury's Office of Tax Analysis (OTA) also take into account how realizations respond to tax rates.(6) In 1990, when the Congress considered a 30 percent cut in the rate on gains, OTA estimated that such a cut would increase revenues by $12 billion over five years; the JCT projected a loss of $11 billion. If they had not factored in a realizations response, the two agencies would have estimated revenue costs of $80 billion and $100 billion, respectively--effectively illustrating how large a behavioral response is incorporated in capital gains revenue estimates.
What the CBO did not find was unambiguous evidence that "If capital gains... didn’t receive special treatment, we’d be getting substantially more revenue". It looks like Krugman's personal pipeline to the truth is wide open. And delivering Kool-aid.
SO TEN YEARS AGO... Surely we can do better than a ten year old CBO study? Hey, be my guest, and stop calling me Shirley. This table shows realized capital gains through 2008; the Tax Foundation tells me that the top long term gains rate was cut to 20% in 1997 and then to 15% in 2003. Realized gains under Bush eclipsed the Clinton boom years by 2006; my quick calculation (applying the relevant top rate to all realized gains each year) is that capital gains tax revenue rose from 2001 through 2007 even with the lower rate, although obviously that is conflated with an improving economy. 2008, of course, was memorably not an example of an improving economy.
Since the lower gains rate is (casually if not causally) associated with higher revenues I don't think that data will update the CBO effort and provide conclusive evidence that raising the capital gains rate brings in substantially more revenue.
PLEASE MIND MY DELICATELY POISED BLOOD PRESSURE: Somewhere a Krugman acoylyte is teeing up a response along the lines of "Krugman didn't say that raising the capital gains rate would raise revenue; he said that raising the capital gains rate and taxes on other investment income would raise revenue".
Uh huh. And if the Yakees could just sign me and Cliff Lee by next April they would be locks foe the World Series.