Data on Tax Rates, by Quintiles

Or, where have you gone, Todd Henderson?
About two years ago, law school professor Todd Henderson complained about the difficulties of making ends meet on slightly more than $250,000, as a way of arguing against the proposal to allow the Bush tax cuts expire for incomes above $250,000, since he was not "rich" in his own estimation. Given the CBO’s release of the actual (data based) estimates of tax rates [1], I thought it useful to revisit the debate over the plight of the top 1%. Professor Henderson took down the blog post soon after posting, and most repostings have been expunged from the Web, but fortunately one can find a repost of this gem here. The key quote:
... Our combined income exceeds the $250,000 threshold for the super rich (but not by that much), and the president plans on raising my taxes. After all, we can afford it, and the world we are now living in has that familiar Marxian tone of those who need take and those who can afford it pay. The problem is, we can’t afford it. Here is why.

Interested readers can read the entire story of the Professor’s plight here.

The graphs based on the CBO data are quite striking, and are at variance with the beliefs held in the fevered imagination of some commentators that taxes have risen enormously under the current Administration.

Figure 1: Average Federal tax rates, by pre-tax income quintiles. Figures in parentheses are the incomes at the bottom of the respective quintiles for 4 person households. Source: CBO, The Distribution of Household Income and Federal Taxes, 2008 and 2009, supplementary Excel file.

Note that the highest quintile bottom income is $148.1K, substantially below the Todd Henderson figure (either self-reported or estimated by others). The following graph shows that the Todd Henderson household is probably at the bottom of the 96-99th percentile.

Figure 2: Average Federal tax rates, by pre-tax income quintiles. Figures in parentheses are the incomes at the bottom of the respective quintiles for 4 person households. Source: CBO, The Distribution of Household Income and Federal Taxes, 2008 and 2009, supplementary Excel file.

One last observation: I know that there will be the usual talk about the onerous burden placed on the “job creators” (a euphemism applied indiscriminately to those households with high incomes) deterring growth, should Federal tax rates rise to those prevailing during the Clinton Administration (during which time per capita income growth exceeded that during the EGTRRA/JGTRRA era). However, there is no obvious correlation between the tax rate applying to the top quintile and per capita GDP growth.

Figure 3: Scatterplot of per capita GDP growth (Ch.2005$) against tax rate on top fifth quintile, 1979-2009. Nearest neighbor fit, tricube local weighting, bandwidth = 0.7. Source: BEA, CBO, and author’s calculations.

For those interested, linear regression yields a slope coefficient of 0.35, significant at the 20% msl. Over the time period of the EGTRRA/JGTRRA experiment (2001-09), the slope coefficient is 1.32, significant at the 6% msl. This is not proof that higher tax rates cause faster growth (since one can’t impute causality without additional structure), but it surely puts the burden of proof on those who assert strong negative effects arising from higher tax rates.

For more discussion of tax rates and elasticities, see [2], [3], [4], [5].