You'd better believe we pay careful attention to capital gains here. Friday, the Congressional Budget Office released an analysis of the rise and fall of federal individual income tax revenues from 1994 through 2004. It showed that capital gains accounted for half of the non-legislative changes to individual income tax revenues over the period. Ironically, capital gains revenues increased 0.7% of GDP from 1994 through 2000 under President Clinton, and they fell 0.6% of GDP from 2000 to 2004 under President Bush.
The headline could have read: "Democratic President Presides Over Capital Gains Bonanza While Republican President Gets a Capital Gains Bust." Like many political headlines, that would be inaccurate and unfair. President Clinton should get credit for a robust economy, but he didn't cause the tech bubble which peaked in 2000. He kept the top capital gains tax rate at 28% until 1997, when he agreed to lower it to 20%. President Bush lowered it to 15% in 2003, where it stands today. Mr. Bush was not responsible for the tech bubble bust or the recession in 2001, which drove capital gains revenues down. Furthermore, CBO stated that capital gains revenues have rebounded 0.4% of GDP through 2007 as overall individual income tax revenues rose 1.5% of GDP to 8.5% of GDP. CBO couldn't extend its analysis through 2007 because required data is not yet available from the IRS.
The conclusion I have drawn as a longtime tax economist is that capital gains revenues fluctuate wildly without regard to short run tax policy or to who is president. Two earlier Congressional Budget Office capital gains studies in 2006 and in 2002 make this point. CBO concluded in the 2006 study that about half of the 2004 surge in capital gains realizations remained unexplained. CBO concluded in the 2002 study that "the relationship of realizations and receipts to gains tax rates is neither predictable or obvious."
Pete: Unlocking Effect and other timing factor
Pete,
I would think that, other things equal, there would be a lower level of capital gains realization immediately prior to the implementation of a capital gains tax cut, followed by a higher level immediately after (due to the postponed sales of assets, plus the unlocking effect).
Is this correct?
If so, has anyone been able to measure the magnitude and specific timing of these effects? And has anyone quantified the portion of the revenues gained from the unlocking effect that are mere shifts of timing -- i.e., how much of that incremental revenue reduces future revenue beyond that point (because people just sold sooner rather than later)?
thanks.
Capital Gains Realizations
Brooks,
As usual, you're on the money.
Capital gains realizations would plunge in anticipation of a rate cut and would accelerate to beat a rate increase. For that reason, capital gains rate changes are invariably made effective as of the first date of public discussion in a hearing or markup. Financial markets prefer to avoid the volatility if taxpayers were informed in advance, and policymakers don't want to lose the extra revenue that would go out the Treasury door from advance announcement.
As for measuring capital gains realizations and tax revenues, there is no data on daily, weekly, or even monthly capital gains realizations or revenues. The IRS reports gains data annually about two years after the fact because it's difficult and expensive to extract that information from tax returns. Once we get to complete electronic filing that might change. One of the problems of trying to do any econometric work with capital gains data is that with only annual data, you don't have many observations.
Pete
Thanks Pete (n/t)
Thanks Pete (n/t)
Post new comment