One of the most troubling features of America’s economic stagnation is the ever-increasing ranks of the long-term unemployed. Almost half of current job seekers have been out of work for half a year or more. Beyond serving as an indication of the brutal job market nationwide, it’s also raised concerns that millions of Americans could slide into a cycle of poverty and unemployment: After a few months on the unemployment rolls, skills stagnate and employers aren’t willing to take on those stigmatized by joblessness. If he’d be such a good employee, why can’t he get a job?
Without downplaying the significance of the uptick in long-term unemployment—job loss has a substantial impact on lifetime earnings and indeed even life expectancy—there’s at least some good news for the employment prospects of the Great Recession’s unemployed. A field study run by economists at the University of Chicago, McGill, and the University of Toronto finds that employers are far less likely to penalize the unemployed for being out of work in communities hammered by recession, since there’s less of a sense that job loss signals that an applicant is “damaged goods.”
Given the many reports of the brutal job market facing the unemployed, you wouldn’t think we’d need a trio of economics professors to prove that it’s harder to find a job if you don’t already have one. But among job seekers, there are all sorts of differences between the recently unemployed and the long-term unemployed that make a comparison of their job searches problematic. The latter group may fare poorly in the job search not because employers automatically discard applications from anyone who hasn’t held a job in a while, but because the long-term unemployed may not have the qualifications or skills that employers are looking for, or may not exert the same effort in their job search, sending out fewer résumés and showing less initiative in following up on leads than the recently unemployed. If researchers don’t properly account for these underlying differences among job seekers—something that’s nearly impossible to do—they may incorrectly conclude that unemployment has a direct impact on job search success, when in fact none may exist. Because of the difficulty in tracking these differences, the evidence of the effect of unemployment length on the chance of finding a job has been described as “mixed and controversial” by Nobel laureate Thomas Sargent.
This basic difficulty is why the researchers—Kory Kroft, Fabian Lange, Matthew Notowidigdo—didn’t look at real job applicants at all. Instead, they sent fictitious applications in response to job postings around the country, and recorded which ones triggered call backs for an interview. Such “résumé audits” have been used to examine the effects of age, gender, and race on employment decisions. The approach allows experimenters to control all aspects of applicants’ résumés, keeping every qualification the same, then randomly assigning some people to appear young and others old (based on year of high school graduation); black or white (by setting the applicant’s name as, Lakisha or Emily); male or female. By comparing the call-back rates of otherwise identical résumés, audit studies can isolate the effect that race, gender, or anything else has in securing a job interview. (This research methodology isn’t without controversy. It wastes the time of employers looking to fill job postings. There is, more generally, an active debate on if and when it is appropriate for social scientists to use deception in research.)
Kroft and his coauthors’ assembled 10 résumé templates that would be appropriate for applicants to entry-level positions in administrative, service, and sales positions, using material drawn from more than 1,000 résumés pulled from online job boards. Each résumé was randomly assigned a length of unemployment, which appeared on the résumé in the form of an end date for the applicant’s most recent job. One-quarter were assigned to be currently employed, and the remaining 75 percent received an unemployment spell of one to 36 months. The researchers sent more than 12,000 résumés to 3,040 job postings in the 100 largest metropolitan areas in the country. In each region, they bought several phone lines to field responses from prospective employers, and waited to see which résumés elicited call-backs.
Consistent with conventional wisdom, the call-back rate for applicants with six months’ unemployment was 4 percent—a bit more than half of that of applicants who had just lost their jobs. After six months out of work the effect flattened out: Even after three years of unemployment, the call back rate was still 4 percent. (Curiously, holding a job was a liability in this job market—only 5.4 percent got calls, 2 percent less than newly unemployed applicants. Employers probably didn’t want to waste their time on applicants that were just playing the field.)
More interestingly, the call-back rates in cities that have been hard hit by recession—places like Las Vegas, Miami, and Detroit—were much less affected by unemployment status than in cities that prospered, relatively speaking, in recent years, for example, Boston, Washington D.C., and Minneapolis. And for high unemployment cities like Las Vegas, there’s scarcely any difference in the callback rate for an applicant whether he’s been out of work for just a month, or for several years. That is, employers in dismal labor markets recognize that long-term unemployment isn’t very revealing about a prospective worker’s abilities.
Of course, this is only mildly encouraging news for the millions of jobless Americans in stagnant and decaying cities—it doesn’t much help to be on even footing in the job market if there’s still no one hiring. But the study may underscore the benefits of encouraging jobseekers to leave high unemployment regions to look for work in places where the economy is doing better. While the study never examined this question directly, one could imagine that employers in Boston wouldn’t discriminate against applicants from Detroit based on their employment records–they would understand that no one can find a job in the Motor City.
And for people who are out of work in the parts of the country that are doing relatively well? One implication that one of the authors, University of Chicago professor Matthew Notowidigdo, takes from the study is that unemployment insurance should start low then increase after a few months of suffering without a job. The reason is that, if unemployment serves as a negative signal to employers, you want to give the recently unemployed every motivation to get back to work as soon as possible.
A perhaps more humane implication that Notowidigdo sees in the study is that President Obama may not have been so far out in left field when he argued that discrimination based on recent employment history “makes absolutely no sense” and should be prohibited. Think of an otherwise qualified worker who loses his job when his company downsizes. He gets a few unlucky breaks in his first few job interviews. All of a sudden he finds himself out of work for half a year, and no one will even let him in the door. This sort of “rational herding,” while sensible from the perspective of each individual employer, can potentially leave a lot of well-qualified employees out of work for years. And suggesting they move to Detroit to get a fair shake from employers probably isn’t the answer.