Does Age Discrimination in Employment Exist?

If you ask many older workers if age discrimination exists, you will get a resounding yes.  Surveys show that at least some older workers think they’re being discriminated against .  However, self-reports are not the best measure of discrimination.  While some instances of discrimination are obvious, others are more subtle, and sometimes there’s a non-discriminatory reason behind perceived differential treatment.

If successful lawsuits are any indication, there’s at least some age discrimination.  In 2013, the EEOC reported 21,396 charges of discrimination under ADEA.  Of that number, 7,232 were resolved by the EEOC, 7 suits were filed, and 16 suits were resolved in court.  However, the number of charges filed in 2013 represents only 0.025 % of labor force participants over the age of 40 and 0.043% of labor force participants over the age of 50.  Of course, this number could also be underestimating the problem if people don’t realize they are being discriminated against or they want to avoid a costly lawsuit.

It is more difficult to study age discrimination than it is to study gender discrimination or race discrimination.  With race and gender, we can set up experimental situations in which job applicants are identical except for their race or gender, but we cannot do that with age because as you get older, you should also have more experience.  We may prefer a younger worker with just one year of work experience to an older worker with the same experience not because of age, but because that lack of experience with the older worker might suggest something about the older worker being unwilling or unable to keep a job.  Similarly, older workers who have been in the labor market really do have more experience and employers may genuinely be worried about “over-qualification” instead of just using that term as code for “being older.”

We know that older people take longer to get new jobs while unemployed and are less likely to get new jobs than younger people.  This difference by age is true even in cases where we might think the reasons for previous job separation have nothing to do with the quality of the worker, such as mass layoffs.  But maybe older job seekers are just picky and expect higher salaries, or are less willing to move, or maybe they’re more likely to be in dying occupations or dying industries.

Experiments using hypothetical resumes can separate out differential treatment by employers from different preferences by potential employees.  Laboratory experiments with simulated resumes suggest that older job applicants are less preferred, although the type of job matters in these experiments.  For example, Elissa Perry has several papers with coauthors demonstrating that young applicants are preferred for “young-type” jobs.  Similarly, while Phillip Young and coauthors found that younger PE teachers were preferred to older PE teachers, they found no difference in treatment by age for teachers in subjects that did not require physical activity.  Recent work by Barbara Fritzche and Justin Marcus notes that older “job changers” are especially disadvantaged.  There may be other interesting differences; some new work I am doing suggests that age discrimination in hiring in an entry-level position may have different patterns for blacks compared to whites.

In addition to laboratory experiments, researchers have done field experiments called “audit studies” that send simulated resumes to real employers and measure their response rates by age.  A study in the 1990s by Marc Bendick Jr. sent applications to Fortune 500 companies and found that 32 year olds were preferred to 57 year olds.  This study controlled for experience by giving previous experience as a teacher or in the armed forces, which may be introducing different biases other than age.  In a more recent labor market experiment, I sent resumes to entry-level job openings in Boston and St. Petersburg, and found that younger women applicants are 40% more likely to be called back for an interview than older women applicants.  This experiment, however, is only really externally valid for women applying to entry-level jobs and doesn’t say much about other labor markets or other job seekers.  Current work by David Neumark is looking at isolating the effect of experience on differential treatment by age for a wider range of jobs and positions.

Bottom Line:  Yes, age discrimination exists in hiring, but we don’t know its full extent.  We only have small pieces of the picture and there’s a lot more work that needs to be done to see the full picture.