Tax Policy 101

After I suggested inconsistencies in how people apply the ethics of progressivity, a friend asked what it really means for a tax policy to be progressive. Depending on how you define and measure progressivity, you can come to vastly different conclusions using identical numbers. Below is an illustration that, while contrived, is on the same order of magnitude as the Bush tax cuts.

Suppose there is an individual who has $10,000 of income and is representative of the bottom income decile—the “poor”. Suppose this person pays 10% of his income ($1,000) in taxes, leaving him with $9,000 of after-tax income. Now, suppose there is an individual who earns $1,000,000 of income and is representative of the top decile—the “rich”. Suppose this person pays 50% of his income ($500,000) in taxes, leaving him with $500,000 of after-tax income. (Note that I use “poor” and “rich” here only for convenience and contrast. In practice, “rich” is certainly not a misleading title for the top tax bracket, but the bottom tax bracket is better characterized as “middle class”.)

In the next election, suppose there is a supply-sider candidate who wants to cut taxes to “stimulate” the economy. The proposed tax cuts would reduce the tax rate for the lowest earners by five percentage points (to 5%) and by four percentage points (to 46%) for the highest earners. After-tax income becomes $9,500 and $540,000.

Is this tax policy progressive?

Well, it depends how you analyze it.

1. If we look at tax rates we see that the rate for the poor was reduced by five percentage points, from 10% to 5%. The rate for the rich was reduced by four percentage points, from 50% to 46%. So, this tax cut widened the tax rate gap between the rich and the poor. The result is a more progressive tax code. This is the only definition of progressivity that does not require manipulating or interpreting the data in a favorable way.
2. Another way to view the difference in (1) is by noticing that before tax cut the rich paid a tax rate that was five times as large (0.50 divided by 0.10), now the ratio is almost 10 to 1 (0.46 divided by 0.05). Clearly, the tax code is more progressive.
3. The tax cut lowers government revenue by $40,500. The rich receive 99% of the tax cut—a monstrous $40,000. Meanwhile, the poor get a measly $500 back, representing only 1% of the total amount of the tax cut. This is what the media and many public officials who opposed the tax cut in the first place would deem “tax cuts for the rich.” But, this is just demagoguery. The fact is, the rich were already supplying 99.8% of government revenue. Any way you slice it, unless you start redistributing wealth, the rich are going to receive a larger piece of the pie from an across-the-board tax cut.
4. Looking at tax rates again, notice that the poor’s tax rate was reduced by 50% (5% divided by 10%) while the rich’s tax rate was reduced by only 8% (4% divided by 50%). In other words, the poor got 50% of what he was paying in taxes back ($500 of $1,000) and rich got only 8% back ($40,000 of $500,000). Here, the tax cut looks extremely progressive. This is how Conservatives might demagogue the issue (although the Right seems especially inept at articulating tax policy) and is only slightly less misleading than the previous example. In this case, however, the fact that a small difference in the percentage point change in tax rates looks big as a percent of tax rates is a result of an already progressive tax code.
5. Now look at after-tax income. The poor’s after-tax income increased by about 5.6% ($500 divided by $9,000) while the rich’s after-tax income increased by 8% ($40,000 divided by $500,000). So, despite the fact that the tax code became more progressive, it appears that the tax cut was actually regressive, slightly favoring the rich. This is the view favored by Brookings Institute’s Tax Policy Center. But this interpretation is unnecessarily intricate and misleading because it is entirely an artifact of existing inequality and an already highly progressive tax code. In fact, the only way to remove this effect would be to make the tax code much, much more progressive—essentially by only cutting taxes for the poor. In Brookings’ defense, they are emphasizing the “distributional impact” of a tax cut, but that is a measure of inequality, not progressivity.
6. Another alternative distributional measure is the percentage of tax liability. Before, the rich were paying 99.8% of taxes ($500,000 divided by $501,000). Now they are paying a higher percentage, 99.9% ($460,000 divided by $460,500). This is slightly misleading because it’s mostly a function of inequality and the rich’s high tax burden.

In summary, the only reasonable and consistent measure of progressivity is the tax rate gap—the difference between the percentage of income paid by high and low earners [illustrated in (1)]. Any other definition is intended—intentionally or otherwise—to advance a particular point of view or it’s an idiosyncratic result from a specialized example and should be taken with a HUGE grain of salt.

The bottomline is that everyone should favor simplifying the tax code—if you don’t, then you are likely a bad person, a member of Congress, or both. Whether or not a person favors tax cuts at all or prefers a more or less progressive tax code probably depends on their ideological bent or where they fall on the income distribution....and perhaps a little bit ignorance, but that is by no means limited to one side of the issue.