July 25, 2010
Against Extending Unemployment Benefits
Against Extending Unemployment Benefits—Posner
I do not favor extending unemployment benefits. The bill signed into law by the President last Wednesday would extend unemployment benefits by up to 13 weeks (depending on a state's unemployment rate) for persons, estimated to number some 2.3 million, who have been unemployed for six months or longer, up to a maximum of 99 weeks (almost two years), at a total cost to the government estimated at $34 billion.
I would not object to the government's giving the states some fraction of $34 billion to enable them to make additional welfare payments to persons experiencing serious economic hardship. But extending unemployment benefits beyond the standard six-month limit is a bad idea.
As Becker points out, the total amount of the additional benefits is too small to have a significant positive effect on employment—and for the further reason that a transfer payment cannot be assumed to have a positive effect on consumption or investment. The effect will depend on what the recipient of the transfer does with it. If he saves it, at least in an inert form such as cash or a federally insured demand deposit that the bank in which the money is deposited uses to buy a Treasury security, then it will not stimulate production and hence employment. Even if he uses it to buy something, it may turn out that the seller is selling from inventory and doesn't plan to restock, in which event the effect on production will depend on what the seller does with the money he receives for the sale—maybe he will save it in some inert form!
Far from being effective as stimulus, the extension of unemployment benefits will have two negative effects on employment. First, it will increase the opportunity cost of the recipient's rejoining the labor force. Unemployment benefits are set lower than earnings to reduce the moral hazard that Becker discusses, but the gap between benefits and earnings is narrowed by the costs of work (such as commuting, and any disutility associated with work, such as fatigue and boredom) and by the benefits of household production and of leisure—and those benefits, unlike earnings, are not taxed. The gap is so small for many unemployed people that studies show that they do not begin a serious job hunt until their unemployment benefits are about to expire.
So extending or otherwise enhancing unemployment benefits, far from stimulating employment, is likely to reduce employment and so slow the pace of economic recovery.
Second, extending unemployment benefits has a negative long-term effect on employment. The longer a person is unemployed, the less likely he is ever to return to the labor force, at least in a job comparable to the one he held before becoming unemployed. Apart from erosion of skills and of the habit of working, persons unemployed for a long time are unattractive hires because employers are suspicious of these persons' attachment to or aptitude for work.
Thus the net effect of the just-enacted extension of unemployment benefits is likely to be to reduce employment and output. And while the amount of the benefits involved in the extension is not large, the negative effect on the economy may be if a substantial number of the 2.3 million persons who will be receiving the benefits are unemployed for a longer time, as a result, than they would otherwise be.
So the extension has to be defended if at all as a welfare measure. It is a bad measure because it is not means-tested. To be counted as unemployed you have to be looking for a job, so one can assume that the unemployed are involuntarily unemployed and so would prefer to work. But that doesn't mean that they necessarily face economic hardship as a result of being unemployed. They may be unemployed because their attachment to the labor force is actually quite week; they do not try very hard to meet their employer's expectations and so are quick to be laid off in bad times, or their search for employment is lackluster. They may have large savings or a high-income spouse. No doubt the bulk of the long-term unemployed are hardship cases, but it is only hardship cases that make a strong claim on a welfare program.
The fact that not all persons who have been unemployed for a substantial period of time are hardship cases reinforces my concern that extending unemployment benefits will cause disemployment. The persons who can afford, as it were, to be unemployed are likely to pocket any additional unemployment benefits that they receive and slow their search for a job.
Should Unemployment Compensation be Extended? Becker
President Obama during this past week signed into law extending unemployment benefits to a maximum of 99 weeks, or almost two years, for persons who have been unemployed for over half a year and have exhausted their state benefits. The degree of extension varies among states depending on a state's unemployment rate, with higher unemployment states getting longer durations of coverage. The bill that became law is highly partisan, passing with almost all Democrats, 31 House Republicans, and only 2 Republican senator s. I believe the law extends unemployment benefits for too long, although the economics of optimal unemployment insurance gives a less than certain answer.
Unemployment insurance tries to balance two conflicting goals. One is to protect at least some of the earnings of workers laid off from their jobs through no fault of their own, while the conflicting goal is not to make unemployment status so comfortable that workers try to get laid off, and do not look seriously for jobs when they are unemployed. The first aim is a typical goal of insurance against bad outcomes, while the second goal is to reduce the degree of "moral hazard"; that is, to reduce the incentive of persons to reduce their efforts to remain employed and look seriously for work when unemployed because they have insurance against the cost of being unemployed.
One typical way insurance tries to reach a compromise between these conflicting goals is to have a deductible that is paid by insured persons, such as the $500 deductible many car owners have on their automobile insurance. The trouble with practically all the state-run unemployment insurance plans is that they typically have no or a minimal deductible because they provide coverage essentially from the first week of unemployment. In addition, they usually limit coverage to a fixed number of weeks, such as 26 weeks. This is an inefficient and costly approach since practically all the unemployed can readily cover their first several weeks of unemployment from savings, spouses' earnings, or borrowing on credit cards and in other ways. Unemployed workers usually run into financial trouble only when they have been unemployed for an extended time. An optimal unemployment insurance plan would make unemployed workers responsible for their first month or two of unemployment, and mainly spend unemployment insurance resources on the longer-term unemployed.
A second insurance approach to the moral hazard problem is to require significant co-payments, so that the insured have to pay a portion of any additional losses they experience after they exhaust the deductible. American unemployment insurance plans usually do pretty well on this by only paying about half or so of the earnings the unemployed had received when they were employed. Many European plans had usually replaced most or all of the earnings of the unemployed, and covered unemployed for many years. After they learned the hard way the prediction of economic theory that this encouraged significant increases in unemployment, several countries greatly cut back both the duration and payment (i.e., replacement) rates for the unemployed. Interestingly, In Germany this was done under a center-left Social Democratic government.
This analysis of insurance provides background for evaluating the new extension of unemployment benefits for the longer-term unemployed. The approximately $30 billion committed to this extension has been partly justified as a stimulus to what may be a slowing US economy. However, since any "stimulus" from $30 billion would be paltry even to the most optimistic stimulus calculations, such a stimulus can hardly be a serious justification for this extension of benefits. This is especially the case when federal budget deficits have been so large during the past couple of years, and there is no serious evidence that the $800 billion stimulus package passed over a year ago has had much stimulating effect on unemployment or GDP. Much of the evidence usually cited about number of jobs created by the stimulus package is based on terrible analysis. Any new hires under stimulus money is assumed to be net jobs created by the stimulus rather than a transfer of employment from non-stimulus activities to stimulus-supported activities.
So the case for the new law rests on its insurance provisions rather than on its stimulus capabilities. I argued earlier that covering the longer-term unemployed is the more optimal way to approach unemployment insurance since the long-term unemployed face the largest economic hardship. From that viewpoint, extending the duration of coverage beyond 6 months makes sense in an environment where the unemployment rate remains excessively high at 9.5%, especially if the extension is accompanied by the elimination of unemployment coverage for the first 6-8 weeks of unemployment.
However, the actual large extension poses a major risk of creating an unemployment culture where men and women remain " unemployed" for years. Once the period of unemployment becomes long enough, people begin to get the habits from being unemployed for a long time: they sleep late, develop various leisure interests, and at the same time their work skills depreciate from not using them for an extended period. Studies have shown that skill depreciation is a serious effect of being unemployed for a long time.
Some might retort that this argument is persuasive during periods of normal unemployment rates, perhaps 7% and under, but not when jobs are scarce, the unemployment rate is over 9%, and it is coming down slowly. There is merit to this response, but on the other hand, the JOLTS data show that even with the current high unemployment rates, about 4.5 4 million persons were hired in May 2010 (and about the same number are either being laid off or quitting their jobs). So for the most part, even the long term unemployed can find jobs if they are willing to take a cut in their earnings, and/or move to other industries and occupations.
This analysis leads me to the following conclusions. During bad times, 6 months of unemployment compensation may not be long enough, but the 2 years in the new law is too long. About 9 months of unemployment compensation would be the right length. Anyone unemployed longer than that would lose these benefits. If they want to work they should be forced to adjust, at least temporarily, to the bad economic environment, and accept jobs that they would turn down during good economic times.
[]
[K1]"The vote in the House was 272 to 152, with 31 Republicans joining 241 Democrats in supporting the measure. Voting against were 142 Republicans and 10 Democrats."
Fewer senators crossed partisan lines (2 Republicans and 1 Democrat) on this issue.