Employer bias against the long-term unemployed?

Learned today about an NBER study done by three economists (one at Chicago Booth) on how employers react to job applications from the unemployed.  The researchers sent out identical resumes to employers with online postings in 100 cities that varied only on one critical dimension: number of months since the applicant's last job, which randomly varied between 1 and 36.  The key result: the odds of getting a callback dropped with the amount of time unemployed.  Persons reporting 8 months of joblessness had a 45% lower probability of getting a call than those reporting 1 month.  After 8 months, additional time unemployed had no effect on the odds of getting a call. 

The study took a more careful look at how local labor market conditions influenced the results.  They found that the relationship between months unemployed and callback odds was strongest in cities with tight labor markets (relatively more vacancies and low unemployment).  In cities with few vacancies and high unemployment, there was no strong relationship between months unemployed and callback odds.

These results indicate that employers use time out of work as a signal of productivity and motivation.  In tight labor markets, employers seem to think something must be wrong with the applicant if they have been out of work 6 months or more; in contrast, time unemployed does not seem to be an issue in areas where there are very few jobs.  The lesson: holding out for better job offers can be a self-defeating strategy.