clock menu more-arrow no yes mobile

Filed under:

The tax debate needs numbers, not just percentages

There is at least one thing that all the Republicans running for president agree on: Americans are paying more taxes than they should, and this is especially true of the richest Americans. This worldview, grounded in both economics and moral philosophy, has led them to embrace tax policy proposals that would deliver far more money to the least needy than they would to those facing the most financial distress.

Bush, Trump, and Rubio tax plans, compared Vox / Alvin Chang

But rather than make the case for this set of priorities explicitly, many conservatives prefer to keep this kind of information hidden and instead discuss percentages.

"Five percent of a million is a lot more than 5 percent of a thousand," Marco Rubio said in the CNBC debate. "So, yeah, someone who makes more money, numerically, it's gonna be higher. But the greatest gains, percentage-wise, for people are gonna be at the lower end of our plan."

The right-of-center Tax Foundation makes the case for percentages by saying that "measuring the distributional effects of tax changes in dollars leads to some wacky results." For example, they say that if you prioritize dollars over percentages "a regressive tax hike can look relatively benign for the poor."

The case for using numbers

The fundamental case for using numbers rather than percentages is that it is the clearest way to communicate what is happening. And there is something fundamentally illogical about denying that large tax cuts are, in fact, larger than small tax cuts. To say that rich people deserve larger tax cuts than the middle class because they currently pay much more federal taxes than the middle class is an interesting proposition that would be worth debating. But to try to say that Rubio, Jeb Bush, Donald Trump, etc. aren't proposing larger tax cuts for high-income households is ridiculous. They plainly are.

If Bill Gates gets into the back seat of a taxi, finds $900 in an envelope in the back seat, pockets $800, and gives the cabbie a $100 tip, then Gates has come away from the ride with more money than the cabbie. The fact that the percentage increase in Gates's wealth is trivial compared with the cab driver's is irrelevant to the analysis.

Regressive taxes are often good for the poor

The Tax Foundation's counterargument takes the form of a reductio ad absurdum. "Everyone knows that taxes on alcohol are regressive," they write, "so, eliminating taxes on alcohol would be great for the poor, right? Not under Vox’s preferred method for displaying the effects of tax cuts."

Tax Foundation

What's happening here is the Tax Foundation is exploiting an ambiguity between the technical definition of a progressive versus regressive tax (where, by definition, percentages are what counts) and a more ordinary-language sense of a progressive versus regressive policy.

Would eliminating excise taxes on alcohol be a boon for the poor? It likely wouldn't. Because separately from the analysis of tax incidence, we have to consider that taxes finance public spending. If lower booze taxes mean higher bus fares, smaller Social Security checks, or fewer people enrolled in Medicaid, then a cut could be disastrous for the poor. Conversely, successful European welfare states often combine some hefty regressive taxes (most often a value-added tax) with generous provisions for the needy — a policy mix that is highly progressive in its net impact. In general, if a tax finances a service that everyone uses (roads or police protection or national defense), then it is good for the poor as long as the poor pay a smaller dollar amount. If a tax finances a service that is disproportionately used by the poor (SNAP or Pell Grants) then it's even better for the poor. It might be even better for the poor to swap out the "regressive" tax for a progressive one, but the point is that it matters what happens with the money.

This is ultimately another reason to prefer numbers to percentages. The impact of a cut in alcohol taxes needs to consider how they are paid for, and that means you need to be able to compare dollars to dollars.

Broad tax cuts don't have to be skewed to the rich

A related complaint about Vox's tax charts that I've heard from several conservative readers is the view that the hefty skew to the rich of most Republican tax plans is simply a mechanical result of the fact that under the current tax code rich people pay very high tax bills.

But this is simply not the case. The first $9,225 of a person's taxable income (or $18,450 of a married couple filing jointly) is taxed at a rate of 10 percent. That amounts to an identical bill of $925 or $1,845 for rich families and middle-class families alike. Reducing the rate for this bracket to zero wouldn't do much to help the poor, but it would offer across-the-board tax relief — an equal amount of relief — to the top 70 percent or so of the income distribution.

The reason the Republican tax plans are so tilted toward the wealthy is that in addition to cutting rates in the low brackets, they all cut the rates paid in the top few brackets — rates that fewer than 2 percent of tax units are rich enough to pay.

Obviously, if you cut a tax that is exclusively paid by rich people, then you are going to deliver a lot of money to high-income households. And all the Republican tax plans include just that in the form of cuts in the high brackets. That's why the plans all give high-income households so much more money than middle-income ones.