Distribution: How should progressivity be measured?
A tax system is considered progressive if, on average, households with higher incomes pay taxes that are a larger share of that income. Thus, in a progressive tax system, the average effective tax rate-tax paid as a percentage of income-rises as income rises. This implies that, under such a system, the ratio of after-tax income to pre-tax income falls as income rises. Hence a natural measure of the impact on progressivity of a change in tax policy is the percentage change in after-tax income. A tax cut that gives all households the same percentage increase in after-tax income is neither progressive nor regressive but distributionally neutral; it leaves the relative distribution of after-tax income unchanged. A tax cut that increases after-tax income proportionately more for lower-income households makes the tax system more progressive. One that increases after-tax income more for higher-income households makes the system less progressive.
- Although the percentage change in after-tax income is likely the most informative measure of the distributional impact of a tax change, the Tax Policy Center also reports several other measures in its distribution tables, each of which can be useful but can also be misleading if interpreted incorrectly.
- Comparison of the shares of a tax cut received by each income group can be misleading because the individual income tax is highly progressive. High-income households may receive what appears to be a large share of an income tax cut, but the tax system could still end up more progressive if their share of the tax cut is much smaller than their share of overall tax liability.
- The average tax cut in dollar terms is another often-used measure of who benefits from a tax cut. For example, in 2010, as a result of the 2001-06 tax cuts, households in the middle of the income distribution will receive an average tax cut of $814 (see table). Those in the top one-tenth of 1 percent will receive an average cut that is almost 400 times larger ($314,150). These numbers alone, however, do not tell us who benefited proportionately more, since those at the top of the income scale have significantly more income-but also pay significantly more tax-than those in the middle.
- An alternative distributional measure is the change in tax as a percentage of the household’s total tax liability. This measure can be extremely misleading. Because low-income households pay less tax than high-income households under a progressive tax system, a small tax cut in dollar terms for low-income households can be a large percentage cut, and thus can appear to be a huge reduction in tax liability.
- For example, consider a family earning $20,000 that pays $1 in taxes and another family earning $2 million that pays $500,000. Now suppose that legislation provides a $1 tax cut for the low-income family and a $100,000 tax cut for the high-income one. Using the percentage change in tax liability makes it appear that the cut is tilted toward the low-income family: its taxes fall by 100 percent while those of the high-income family fall "only" 20 percent. In fact, the cut increases the after-tax income of the poor family by just 0.005 percent while increasing that of the wealthy family by 5 percent, or 1,000 times more as a percentage of income.
- Thus it would be extremely misleading to characterize the 2001-06 tax cuts as benefiting the middle class more than the wealthy simply because the percentage change in tax is, on average, higher for those in the middle quintile. For example, in 2010 federal taxes fall by an average of 11.6 percent for those in the middle quintile and 9.1 percent for those in the top quintile. But in terms of what really matters-economic resources as measured by after-tax income-those in the top quintile benefit proportionately more. They see an increase in after-tax income of 3.6 percent, while those in the middle of the spectrum receive only a 2.3 percent increase.
- Another measure of the impact of a tax policy is the change in the share of the overall tax burden paid by different income classes. This concept can also be misleading as a measure of the progressivity of a tax cut if the tax burden is changing at the same time.
- Consider the example discussed above. After the tax legislation that provides the low-income family with a $1 tax cut and the high-income family with a $100,000 cut, the high-income family pays 100 percent of the total tax burden, since the other family has had its tax liability completely eliminated. But just because the high-income family now pays a larger share of the tax burden, it does not mean that it did not benefit the most from the tax cut. As already shown, in percentage terms its after-tax income went up 1,000 times more than did that of the lower-income family.
- The problem with using the change in the share of the tax burden can be seen when examining the 2001-06 tax cuts. As a result of the cuts, the share of the federal tax burden paid by the top income quintile rises by 0.3 percentage point in 2010, whereas the share paid by the middle quintile falls by 0.2 percentage point. But, again, that does not imply that the 2001-06 tax cuts made the tax system more progressive, nor does it mean that those in the middle quintile benefited more than those at the top.