Unconscious bias plays a bigger role than we like to admit in company decisions about hiring (pdf). Case in point: Multiple studies have found that having a name associated with a minority group leads to substantially fewer callbacks from potential employers. But which minority group?
Well that depends on what country you’re in.
The OECD’s most recent report on international migration compiles a wide variety of research on résumé discrimination in member-countries. All these studies use a particular metric to look at workplace discrimination. The key gauge is “callback ratio,” which is the average number of résumés that someone with a “recognizably minority name” submits before they get a callback for a job interview, compared to résumés with the exact same skills but from candidates whose names identify them with the majority population group in that country. Nearly every study has found at least some level of discrimination. (Though in some cases—Portuguese names in Switzerland, Italians in Australia, and Hispanics in the United States—it was quite small.)
A ratio of two means someone would need to make twice as many applications for a job to get a callback, compared to an applicant with a name associated with the country’s majority population. Here are some of the groups that have the highest ratios—which suggests higher levels of discrimination—where the results were statistically significant:
There’s a wide amount of variation in the studies, which makes direct comparison between countries very difficult. Everything from the skill level of the applicants to how long they’ve been in the country varies.
For example, the Swedish data focuses on young workers, and on a variety of medium- to low-skilled jobs. The Swiss data is for medium and low-skilled, foreign-born Albanian men in the German speaking part of the country. The Irish data focuses on medium-skilled jobs, like administration, accounting, and sales.
Discrimination rates generally tend to be higher in less-skilled professions. The experiments are designed and data collected in different ways. (There’s a full list of the studies the data came from in the OECD report that compiled them.)
That’s all to say that you shouldn’t generalize these results to a whole country or labor market, though some in-country comparison is reasonable.
But the ratios, which repeatedly approach 2, indicate discrimination. Having to get through a stack of applications doesn’t always encourage people to review their unconscious biases—they have to be reminded.
While not every organization is going to be able to devote resources to combat unconscious bias, at the very least they can make a point of informing people about them. Just prompting people to do a double take when they find themselves about to reject an application with an unfamiliar name can make a difference.
So what do these findings really mean? Well there are a few different ways of thinking about the economics of discrimination in the workplace. One, known as taste-based discrimination, simply suggests that some employers have a preference against hiring minorities, even if they’re just as productive as other workers. Another, implicit discrimination, is thought to reflect attitudes that the people making discriminatory decisions they are themselves unaware of. Finally, there’s the notion of statistical discrimination, in which the person making the decision is relying not on the characteristics—for example the job skills—of the person in question, but rather some other notion of the “the average characteristics of the group” to which that person belongs.
But really those are only elaborate ways of dressing up the obvious: discrimination is discrimination.