Evan Soltas notes that means-tested benefits create work disincentives for workers earning low wages, a subject that John Cochrane, Andrew Biggs, Eugene Steuerle, and, most controversially, Casey Mulligan, among others, have addressed in really interesting ways. Drawing on a recent CBO report, Soltas writes:

The marginal tax rate of earners making 100 percent to 150 percent of the federal poverty level — between $23,050 and $34,575 for a family of four — is more than 30 percent, as compared with a federal statutory rate of 15 percent. Almost one-third of earners in this range — tens of millions of Americans — paid a marginal tax rate higher than 39.6 percent, the top statutory rate. The average person paying a 40, 50, 60, or even 70 percent marginal tax rate isn’t well-off: More likely than not he or she is below middle-income.

Yet he also acknowledges the trade-offs that are at work: phasing out benefits is a way to contain the cost of social transfers, and the most obvious alternative way to contain costs while preserving work incentives — sharply reducing social transfers — strikes many observers as problematic, particularly to the extent that it leaves low-earners with much less disposable income. And so Soltas concludes on the following note:

So these effective high marginal tax rates on low and middle incomes are, to some extent, the downside of a worthwhile trade-off to achieve sharply negative average tax rates for the poor. Although their marginal rates limit their economic incentives to work for more income, it’s thanks to the negative average rates that they have any income at all. In terms of priorities for public policy, incentives for work are important, but it’s unreasonable for them to trump a concern for basic quality of life. [Emphasis added]

Soltas makes a reasonable point. But it does seem sensible to try to flatten some benefits, e.g., subsidies for the purchase of medical insurance, to mitigate the deleterious impact of means-tested transfers on work incentives.