Monday, December 6, 2010

Mankiw on UI

A thoughtful post from Mankiw on Unemployment Insurance (UI).

He leaves out a couple of bits that are relevant to the current discussion. The first is that one additional benefit of UI centers on increasing the quality of job matches. Severely credit constrained workers without much in the way of assets will make bad job matches very quickly in the absence of UI and that has some costs for the economy in terms of forgone output and later job switching.

The second is that my understanding of the optimal UI literature is that it does have one fairly robust finding, which is that UI benefits should decline with spell length. This finding pops out fairly quickly from a model in which individuals have heterogeneous tastes for leisure. Among those with long spells, individuals with a strong taste for leisure are over-represented because the dynamic selection process weeds out those who do not like leisure earlier in the spell. As the optimal UI payment is decreasing in taste for leisure, so also is it decreasing in spell length.

The third is that there are dynamic issues here. We are setting up expectations for the next recession. This increases the impact of current extensions on future budgets.

I think I would be inclined to do the following in terms of an extension:

(1) Don't call the checks UI. Call them "Temporary Aid to the Long Term Unemployed". This makes it clear that they are welfare and not social insurance via the parallel to Temporary Aid to Needy Families. It also, hopefully, reduces expectations regarding future, milder, recessions.

(2) Following the optimal UI literature, reduce the replacement rate so that payments decline a bit. One could even think about dropping the replacement rate business altogether (along the theme that this is not a social insurance program and not really UI) and tying the payments to demographics instead. The cost of that is a whole bunch of bureaucracy that it would be hard to do quickly (and for which UI offices are likely ill-suited), so I guess I prefer just reducing the replacement rate a bit for practical reasons. Similar practical concerns arise with otherwise reasonable ideas like making people with family incomes above some level ineligible for the extended benefits.

(3) Attach some time-consuming job search requirements along the lines of those imposed in some states under the Worker Profiling and Reemployment Services System (WPRS). What I have in mind are the sorts of things Kentucky was doing when the data were generated for the Black, Smith, Berger and Noel (2003) American Economic Review paper. These sorts of required services are a two-fer. Based on the results in BSBN, they chase some people off UI who should not really be there. For the people who should be there, they may help them get a job by providing them with job search skills or simply by giving them a needed pep talk or kick in the pants. In this context, it may be useful to include some of the sorts of counseling services sometimes offered to high-wage union workers displaced from contracting industries and designed to help them adjust their wage expectations down to something closer to the value of their marginal product.

Like Greg, I think the petition is telling us about the normative views of those who sign it as well as their positive views, but there is some of both, as the statement of views at the start of the petition makes clear. I should note in full disclosure that the list of petition signers includes several friends of mine.

No comments: