The Misleading Effects of Averages in Tax Distribution Analysis
20 Pages Posted: 11 Feb 2005
This study examines how tax distribution tables often can provide misleading results about the impact of pending tax legislation. These tables rely excessively on comparisons of average tax liabilities of various income groups, and are typically used to oppose broad income tax relief and foster class warfare notions in tax policy. However, tax distribution tables typically are defective in several ways that once recognized undermine their statistical validity and raise serious questions about their value to policymakers and the public.
This study analyzes tax data to examine whether the use of averages in the typical distribution table is statistically sound. These tables usually rank tax filers from lowest to highest incomes, and subdivide them into income groups, such as fifths (quintiles). Then the average tax liabilities of each fifth can be calculated, as can average projected benefits from pending tax relief legislation. Finally, the average tax benefits provided to the various groups can be compared and contrasted. The underlying assumption is that these averages are representative of each income group.
However, this Joint Economic Committee analysis of Internal Revenue Service data for individual income tax returns finds that the use of averages does not accurately reflect the tax payments or potential tax changes of most taxpayers in each income group. In fact, most taxpayers in each income group have tax liabilities considerably different from the group averages. This fact is not surprising given the different characteristics of tax filers at similar income levels.
For example, in the middle fifth, only 23.9 percent of taxpayers had income tax liabilities that were within 25 percent of the group average. In other words, over three quarters had tax liabilities that were greater or less than 25 percent of the average tax liability of $1,780 in 1999. Therefore, 43.8 percent of taxpayers had income tax liabilities more than $2,230, while 32.3 percent paid less than $1,340. Although these taxpayers fall into the same income group typically used in distribution tables, their taxes vary greatly. Thus the average tax liability for this group substantially misrepresents the tax burden of most of its taxpayers. Therefore, unqualified comparisons of average tax liabilities and average tax changes would be misleading. Yet this is the usual procedure employed in the construction of distribution tables.
Keywords: Tax, tax distribution, tax anlaysis, tax policy
JEL Classification: H2
Suggested Citation: Suggested Citation