A Reporter’s Guide to Quoting Revenue Effects of “Bush” Tax Cut Scenarios
August 2, 2010
There are a lot of numbers being thrown around in regards to the issue of to what extent Congress should extend the so-called "Bush" tax cuts that were passed between 2001 and 2006 and set to sunset on December 31. One of the key figures is how much extending the tax cuts (or various parts) would cost. A few things should be kept in mind for those reporting such a figure:
(1) Choose your time frame wisely. Fiscal year 2011 estimates should be avoided, if at all possible, because F.Y. 2011 begins in October 2010. Fiscal year 2012 is better, but there are still some issues possibly with the timing of capital gains realizations, which have revenue effects in the years immediately preceding and following the tax change. The best score to use would be the ten-year score, but dividing it simply by 10 to get an annual score will not give you a reliable estimate for 2011 and 2012 (due to nominal income growth that is projected to take off in the latter years of the budget window).
(2) Baselines need to be understood. When we refer to the "cost of extending the tax cuts," that "cost" equals the revenue raised if the tax cuts were not extended less the revenue raised if tax cuts were extended. Hence, the baseline would be "current law," which assumes the tax cuts expire. Therefore, extending any of the tax cuts would be "costly." If one uses the administration's baseline (i.e., OMB Baseline), which assumes that all the tax cuts are extended, any expiration would technically decrease the deficit. Also, whether expiring tax cuts equals a "tax increase" depends on your baseline as well.
(3) Economic assumptions affect the revenue score. When JCT and OMB score important tax provisions like the Bush tax cuts, assumptions about the size of the base can significantly affect the score. For example, as is customary, OMB has a much rosier set of economic assumptions than does CBO. Therefore, because OMB assumes the base (i.e., income) is going to be higher than CBO, when OMB quotes how much expiration would affect revenues, that figure is going to be significantly higher than when CBO or JCT does.
(4) AMT interacts with the "Bush" Tax Cuts. When quoting the cost of extending the Bush tax cuts, one must ask whether that "score" assumes an AMT patch is already in place or not. For example, if one assumes that AMT is not patched and then simply runs the numbers for extension of the Bush tax cuts, the cost of extending those tax cuts is going to look significantly lower since AMT "takes back" much of the savings from the tax cuts for many upper-income (but not super high-income) taxpayers. If one then ran the numbers for patching AMT, the cost of the AMT patch would be high. If one goes the other way and assumes that AMT is patched before running the numbers on the Bush tax cuts extension (which I would recommend), the cost of extension is much greater.
(5) The level of tax rates in the first four tax brackets (10, 15, 25, and 28) have a "cake" effect on high-income taxpayers. Extending the tax cuts "just for people making below $250,000" is technically impossible because even if Congress maintains the lower rates in the bottom four brackets, those cuts add to the after-tax income of high-income taxpayers. Therefore, even if Congress would approve the administration's plan, the cost of that extension (relative to expiration scenario) would still include a large chunk that would benefit high-income taxpayers.
(6) Overall, if one assumes an AMT patch already to exist, OMB cites a figure of around $3 trillion as the cost of extending the tax cuts. This comes from Table S-7 of the budget, which is a bridge between the BEA Baseline (assumes full expiration) and the OMB Baseline (assumes full extension with partial exception of estate tax). If Obama gets his way and the top rates return to their higher levels, along with higher capital gains and dividend rates for high-income returns (relative to 2010), etc., OMB estimates that such a plan would lower the $3 trillion cost figure by about $630 billion, taking it to around $2.3 trillion over 10 years, which is still a significant amount of money.