Yesterday, I wrote about income cliffs in the ACA and their potential effects on wages. They are real, and simple microeconomic theory predicts that they could be a disincentive to work.
However, a number of people have written to me explaining that while the theory is sound, empirical data show that they don’t seem to pose nearly the threat they might. In fact Jared Bernstein (among others) will be testifying today before the Subcommittee on Human Resources of the Committee on Ways and Means, in a hearing entitled, “How Welfare and Tax Benefits Can Discourage Work“. Key findings:
- While benefits of means-tested programs are, by definition, reduced as incomes rise beyond a certain point, their work disincentives differ, and a number of significant programs, including the EITC and SNAP (formerly Food Stamps), are found to have either positive or neutral impacts on labor supply.
- A recent, exhaustive review of the poverty reduction effectiveness of our safety net and social insurance programs found that “…the combination of the means-tested and social insurance transfers in the system have a major impact on poverty, reducing deep poverty, poverty, and near-poverty rates by about 14 percentage points in the U.S. population as a whole in 2004.”
- Importantly, the study concluded that “…this impact is only negligibly affected by work incentives which, in the aggregate, have almost no effect on the pre-transfer rates of poverty in the population as a whole.”
- Recent research also finds positive generational effects of safety net programs on later education and earnings outcomes of children from families that received such benefits. In the full accounting that I’m advocating, these benefits too must be assessed against any costs of work disincentives.
It’s been stressed to me that this type of cliff is a characteristic of any means-tested program, that research shows that in terms of labor supply the effect is less than might be predicted theoretically, and that the only way to avoid the issue entirely is to have a slow phase out (which of course costs more money).
I completely admit that my concerns are theoretical, not empirical. I have no evidence that what I’m worried about will occur. I think Jared makes a good case that the research doesn’t support my fear, at least in domains other than the ACA. Go read.