Sara Murray of the Wall Street Journal aired the opposite side of the unemployment benefits debate yesterday. Could a cut in unemployment benefits make deficits worse?

Extending unemployment benefits isn’t free, of course, and has the potential to keep unemployed workers out of jobs for longer. But it could also be preventing a “lost generation,” economists say. That generation is the crop of 50-somethings who might have worked for another decade. Their outlook isn’t bright…. Unemployment checks have the added benefit of helping these people feel like they’re still a part of the labor force. When the checks run out, and with few glimmers of hope in their job searches, they’re more likely to drop out of the labor force and turn to a program like disability. And unlike a relatively short-term fling with jobless benefits, their attachment to disability is more likely to be permanent.

The immediate, apparent problem with this line of thinking is that it relies on a fuzzy intuition about people’s feelings and self-image about workforce participation, rather than clear, rational economic motives. A more serious problem is that Murray makes a few mistakes with the 34-page social science article upon which she relies heavily. The result is that the article, from the Winter 1995 Social Security Bulletin, is slightly misrepresented.

Murray leaves the impression (surely unintentional) that this study survey offers support for the “fuzzy intuition” noted above. That’s actually a jump made by others, not by the authors of the paper she cites. The paper’s points are far more limited. The authors note that when the unemployment rate goes up, some studies (although not most) suggest that disability applications and/or awards go up as well. The paper also notes that disability rolls sometimes swell when welfare programs change — especially when certain states (Michigan in particular) make a concerted effort to transfer people from state-funded assistance programs to the federally funded SSI program.

The article has no comment at all on how or whether unemployment insurance — which is very different from welfare  – comes into the picture. It has no comment on deficits, either.

In the absence of any literature , the burden of proof certainly lies with anyone who wants to jump from the limited conclusions of this study to further conclusions about what happens if unemployment benefits are cut back or extended. If anyone has tried to carry that burden, Murray does not mention their efforts.

Moreover, whatever you think of the “fuzzy feeling” hypothesis, it does not imply or entail that a large percentage of today’s unemployed are actually eligible for disability. As part of the explanation for the possible link between welfare and disability benefits, the article cites a 1993 study for “evidence that a substantial share of AFDC (welfare) mothers have disabilities.” This is presumably less likely among those who were recently working and now receive unemployment benefits — and again, the burden of proof is with those who would claim otherwise.

Then there is the deficit question. It is very difficult to apply for disability — the waiting period alone poses a major obstacle — and the Social Security Administration denies a large majority of claims. For every person who, facing a cut in temporary unemployment benefits, moves successfully to permanent disability benefits, there may well be dozens who will simply stop collecting unemployment or who apply for disability unsuccessfully, offsetting any added costs to the system. Anyone who wants to assert that a cut in unemployment would increase deficits must show at least some proof to the contrary, and no such proof has been offered.

And the average monthly disability benefit ($1,065, according to SSA) is actually a bit less than the unemployment benefits paid in most states. So for the first 99 weeks at least — under any circumstances — deficits should decrease, even if the rest of the hypothesis is true.

And here’s another problem with Murray’s treatment of the study she cites: She sums up the study’s detailed survey of other studies on the link between unemployment and disability claims (on page 10 of the pdf) thus:

Other studies have shown that for each percentage point increase in the unemployment rate, disability rolls increase anywhere from 2% to 7%.

This is incorrect. The numbers she cites are for the effect of unemployment on disability applications, not on disability awards or beneficiaries. It’s an important distinction if we’re talking about deficits. In recent quarters, applications have spiked, and the agency has been far more stingy in approving them than before.

Also, you wouldn’t know it from reading Murray’s blog post, but four of the eight studies mentioned found a negligible statistical relationship between unemployment rate and disability applications.

When it comes to disability awards, the story is similar. Three of the six studies surveyed on the link between unemployment and disability awards found no statistical link. Of the other three, one found a 2 to 4 percent increase in awards for each percentage point increase in unemployment, one (the most recent) found a 3 percent increase per point, and one found a 5 to 6 percent increase per point.

Of the two studies that looked directly at the link between the unemployment rate and the number of disability beneficiaries, one found no link, the other found a 1 percent increase in disability beneficiaries for every point increase in the unemployment rate.

And again, nothing here about unemployment benefits at all. What we have here is a very limited study being stretched well beyond its ability to prove anything.