Do Employers Care About Potential Employees’ Long-term Unemployment Spells?

Short-term unemployment is below its prerecession level, but there is no such relief for the long-term unemployed, the New York Times recently reported. Long-term unemployment has tripled since 2007 according to the Bureau of Labor Statistics.

As Washington, DC grapples with the so-called “jobless recovery” in the United States, the 3.7 million people out of work for 26 weeks or more wrestle with the stigma employers place on long-term unemployment. According to new research in the American Economic Review, though, that stigma may be overstated.

In “Do Employers Use Unemployment as a Sorting Criterion When Hiring? Evidence from a Field Experiment,” authors Stefan Eriksson and Dan-Olof Rooth find that there is no negative stigma associated with jobless spells shorter than six months. After that, unemployment is used as a sorting criterion only in low- and mid-skill positions. Unemployment of any length appears to have no negative effect on applicants for high-skill positions.

These findings call into question the accuracy of previous research on the topic. The authors argue past papers did not pinpoint the effects of unemployment, all else equal, on employability. For instance, if a person with poor motivation loses a job and is unable to find a new one, is it because of the stigma associated with being unemployed or his own lack of drive? In this case, it is impossible to disentangle the effect of unemployment from the other unobservable, but relevant, characteristic.

Eriksson and Rooth draw a distinction by basing their findings on a field experiment where the only thing that made one applicant different from another is the fact that one was long-term unemployed and the other was not. The authors designed and sent 8,466 fictitious job applications to 3,786 employers in Sweden between March and November 2007. Eriksson and Rooth randomly varied several attributes in the applications, including past and contemporary spells of unemployment, professional experience, education, gender, and ethnicity. The occupations, which included both low- and high-skill positions, mirrored the mix of occupations among employed people in both Sweden and the United States.

For each type of applicant, the authors tracked the number of interview requests received. Ultimately, through the field experiment, they found:

• Employers do not attach a stigma to contemporary jobless spells that last six months or less.

• A stigma starts to attach to unemployment at the nine-month mark in the case of low- and mid-skill jobs, cutting the callback rate by about 20 percent.

• No such effect exists for high-skill jobs. Whatever the length of unemployment, the callback rate is essentially the same.

• Past long-term unemployment spells do not matter to employers for jobs of any skill level. One year of work experience is enough to reverse the effect of one year of past unemployment.

According to the Bureau of Labor Statistics, in the United States over one-third of the unemployed have been jobless six months or more, and about half of them are 35 or older. Eriksson and Rooth, though, limited the length of unemployment spells that they tested to 12 months and the age of applicants to 32 years. Policymakers should hesitate before generalizing the effects of long-term unemployment based strictly on this research.

Another weakness is its focus on the first stages of the hiring process. Eriksson and Rooth test the effect of unemployment on whether an applicant is invited to interview for a position, not whether they are offered the position. It is possible, and the authors admit as much, that long-term unemployment matters more later on in the process.

There are no obvious policy prescriptions from Eriksson and Rooth’s research other than the one (and only) thing economists appear to agree on: it is better to be a high-skilled job seeker than a low-skilled one. That said, the authors set a new standard for research on the topic based on randomization and control trials. Ultimately, this sort of research will benefit the design of innovative programs meant to help the long-term unemployed.

Article Source: Stefan Eriksson and Dan-Olof Rooth, “Do Employers Use Unemployment as a Sorting Criterion When Hiring? Evidence from a Field Experiment,” American Economic Review 104, No. 3 (2014): 1014-39.

Feature Photo: cc/(emdot)