Whose Incentives?

By Paul Krugman, The New York Times

11 July 12

note on taxes, benefits, and incentives, inspired by the “we are VIP/engine of the economy” crowd.

There is no question that incentives matter, that other things equal, someone facing a high marginal tax rate will work less than he or she would otherwise. How much they matter is another issue; in fact, careful empirical study suggests that they matter far less than right-wing mythology would have it.

But there’s a further point: if you care about incentives, you should care about everyone’s incentives. The top 0.01 percent may like to imagine that it is the engine of the economy, but there is no good reason to believe that there is anything special about their role other than the fact that they make lots of money. Or to be a bit more specific, there is no reason to believe that there is a larger gap between the social marginal product of super-elite earners and their pay than there is for anyone else. (If anything, given the prominence of dubious financial activities in the top .01’s income, the presumption goes the other way). So if you’re worried about the effect of marginal tax rates on work incentives, you should worry about that effect at all income levels.

And here’s the thing: it’s actually a well-documented fact that effective marginal rates are highest, not on the superrich, but on workers toward the lower end of the scale. Why? Partly because of the payroll tax, but largely because of means-tested benefits that fade out as your income rises. Here’s a recent discussion by Eugene Steuerle (pdf), with this figure included:

So if you’re really, really concerned about incentives, you should focus not on top rates but on this lower-income trap. How can this be done?

Well, one answer is to be meaner to the poor; bringing us back to Dickensian conditions would reduce effective marginal tax rates, plus lead to a lot of starving children and such.

Another answer is to do less means-testing — for example, to provide universal health care, free school lunches for all, etc.. Of course, this would require more money, and hence higher marginal tax rates on higher incomes — but remember, those incomes aren’t where the acute incentive problems are to be found.

But in the real world, of course, conservatives want more means-testing, not less, so that they can afford to cut tax rates at the top. Social justice aside, this actually makes the “real tax system”, as Steuerle dubs it, even more distorted.

And all of this is justified with the mythology that the rich make a special contribution to the economy over and above the huge amounts they actually get paid. Funny how that works.