About 1.3 million people have just had unemployment compensation benefits cut off in one fell swoop. Whether that is fair is debatable.
It will save the federal budget several billion dollars. And it will lower the “headline” unemployment rate, although not for the reasons that some poorly informed people think.
The net effects of this cut on the overall economy are hard to estimate, and economists will differ on this issue in accordance with their views of how changes in government spending affect total output. I think it is poor public policy, but with the current Congress, it may be the best we can expect.
Start with a few basics of unemployment compensation. In our country, it is a cooperative program between state and federal governments, with considerable state-to-state variation in details of eligibility and payment levels. In normal times, benefits are limited to 26 weeks.
However, when unemployment is high, existing law makes benefits available for an additional 13 weeks. The rationale is that it is much harder to find another job when the overall unemployment rate is high than when it is lower. Thus, even those people making extreme efforts to get a new job will, on average, spend more time before achieving success. Such extended benefits vary from state to state, with the total unemployment rate in a state having to exceed a threshold level before the longer benefits are available.
Beyond that, Congress has further extended benefits on an ad hoc basis during recessions. It did so with a complex, multi-tiered Emergency Unemployment Compensation act at the end of the Bush administration in 2008. This was further modified in 2009 and 2012. This program funded benefits for up to a total of 99 weeks in many states.
The recent budget compromise eliminated funding for any benefits beyond 26 weeks as a deficit reduction measure. That is what the current controversy is about.
There is much public confusion about the relationship between the unemployment rate and the number of people getting unemployment compensation. Many believe that when a person’s benefits run out, they are no longer counted as unemployed. That is not the case. There are many people tabulated as unemployed who either were not eligible for compensation in the first place or whose time has run out. So there is no automatic change in the unemployment rate from this cut in benefits.
However, over the next weeks and months, there will be some effects. Simple economic theory indicates that a government payment for being unemployed that must be given up when a new job is secured lowers the marginal benefit of getting a job and hence lowers willingness to search for new jobs or accept those that are not highly desirable. Labor economists recognize this as a legitimate problem, although the magnitude of this disincentive effect is a matter of perennial debate and is not as large as some in the general public think.
But to the extent that anyone on extended benefits was lackadaisical in their job search because of the payment they would have to give up, the cut should spur some activity. And it will motivate people to take lower-paying jobs than they would have before, since the gap between benefits at zero and whatever the job pays is now wider.
So the cut in extended benefits will lower the unemployment rate. I am most convinced by the labor economists who think this effect will be pretty small.
There is a second way in which the unemployment rate will drop. To get unemployment benefits, the law specifies that you must take certain actions to seek work, although this is not strictly audited in most cases. Similarly, to even be counted as unemployed by the monthly Current Population Survey, from which the unemployment rates are computed, you have to be actively seeking employment.
There always are some people who are eligible for benefits who look for work only because it is a condition of getting the payment. They might be on the verge of retirement, in any case, or have a pending application for disability benefits. They might be in an occupation or geographic area where the chances of getting another job are extremely low.
If their benefits are cut, they may simply drop out of the labor force. That will happen by default once they stop looking for work. If they really would like to have jobs, but stop searching because they judge that their chances of success are not worth wearing out shoe leather to make applications, they will become “discouraged workers,” omitted from the “headline” unemployment rate. But they’ll be included in the broader “Table U-6” numbers, also listed in the monthly Bureau of Labor Statistics employment report.
That effect, from people dropping out of the labor force in response to the cut benefits, is likely to be about as large as that resulting from slackers being jolted into taking jobs they were ho-hum about as long as they could get a check.
Note, however, that while the people prodded into taking jobs will increase total employment numbers, people dropping out will not. Their dropout will decrease the tabulated unemployment rate without increasing the number of total people having jobs.
To Keynesian economists, who think the government can manage the overall level of output and employment by varying spending and taxing, any cut in federal outlays right now, with the economy still sluggish, is counterproductive. Reducing spending on extended benefits is a bad idea, but so is cutting military procurement, federal funding of highway construction or food stamps. They see the job-seeking motivational effects of ended benefits as small.
Rational expectationists, who oppose Keynes’ ideas, don’t worry about the broader effects of a spending cut. Some see it as increasing general business and household confidence because long-standing deficits are being addressed, even if only infinitesimally. And they tend to see discontinuing benefits as motivating considerably more effort to get jobs.
We are conducting an experiment that will generate fascinating data economists will sift through for decades. Those families affected are likely to see things in starker terms.