Resumes and Bad Judgment: More bias than you think

by cv harquail on April 20, 2010

Every resume should come with a warning label:

Don’t trust your judgment of this person. It’s probably biased.

Study after study shows us that the assessments we make of people based on their resumes are biased. These biases are predictable, and they skew our judgments of job candidates.

Judging personality from resumes? You’re probably wrong.

When people look at resumes, they think that they can assess the general personality of candidates. Except that we can’t. Studies show that we correctly assess someone’s correct personality traits only 20 % of the time, and even then we can get only a few kinds of traits close to right.

We think we’re picking the extrovert, the detail oriented, the get it done guy… but we are not.

Judging the “better candidate” from resumes? You’re probably wrong.

Think that you’ve got good judgment about which candidate is better-qualified? Think again.

Studies have shown us that given two resumes that list identical job experience and qualifications, and differ only in that one has a male name and the other has a female name, more often we will judge the man as being the better candidate.

We pick James over Jane. Then, we offer James a higher salary.201004201517.jpg

A now famous study demonstrated that, when presented with otherwise identical resumes, except that one has an obviously African American name (e.g., Jamal) and the other a more ambiguous name (e.g., James) we’ll pick James over Jamal.

We’ll also offer James more money.

Both of these streams of resume research show us that we discriminate, unconsciously, when we evaluate resumes. We express a preference for men over women, and we shy away from ‘obviously’ African – American candidates in favor of candidates of ambiguous race.

Judging future career commitment from resume? Surprise — Likely you’re wrong.

A recent resume study has shown that raters prefer a married female candidate who has kept her birth name (not her “maiden name”, please) over the female candidate who has taken her husband’s name.

Why? Think of it as a new kind of marriage penalty – a woman who takes her husband’s name is seen as less committed to her career.

(And, she is seen as having a warmer personality. In other words, if you refuse to take your husband’s name you’re a career oriented, cold-hearted bitch. But you knew that already.)

For the 7% of us married women who keep the names we were born with, this sort of discrimination might be good. While it may not balance out the discrimination for being female, for not being obviously white, for having children, and for not having played football in college, when discrimination works in our favor, why complain? Except that we may care about having an unfair prejudicial advantage over the other 93 – 94% of married women job candidates.

How valid are resume studies?

Resume studies are incredibly useful. Despite being ‘lab studies’ with predictable caveats about how well they apply in ‘real life’, lab studies let researchers control for competing explanations, to single out one particular difference that would be linked to differences in evaluation of a candidate.

Resume studies are also valid in a different way—resume studies mimic pretty well what hiring managers actually do.

Hiring managers sort through resumes making quick judgments about which 5 candidates out of 250 applicants they’re going to interview. And, they establish expectations of these candidates (e.g., how outgoing, how career-oriented, how warm-hearted) and penalize candidates if in reality the candidate differs from expectations (even if the candidate is better than expected).

Why does it matter?

A BBA student asked me a few weeks ago how we should make sense out of all these studies of gender- and rae-based pay discrimination. Where does it start? She wanted to know, I think, because she wants to figure out where to focus efforts for change.

Think about the estimated difference in pay estimated in this study. The Dutch researchers who conducted the married last name study determined that a woman who changed her name might earn 361,708,20 Euros less over the course of her working life, due to this bit of discrimination alone.

The size of the impact of any one piece or type of discrimination is less important than the fact that this discrimination still exists, in all the nooks and crannies in business decision-making.

Discrimination is invidious. It is everywhere. Thank goodness it is less blatant, less overt, that it used to be. But that doesn’t mean discrimination is gone, and that we don’t have to worry about it now that we have put so many corrections into place.

There is so much hard work still to do. There are fewer big wins, like chances to change policies that required women to give up jobs with travel expectations once they had kids, or presumptions that put African-American sales reps in urban territories rather than less “diverse” territories with higher potential.

Instead, what we’ve got to change are a myriad of minor moments of discrimination – the ones that happen as we scan a page and sort people into the “maybe pile” and the “no pile.”

We shouldn’t be too quick to assume that gender bias doesn’t exist, and that race-based discrimination at work is a thing of the past.

We’re likely to be wrong.

See Women, Work and a Name Change By Catherine Rampell in The New York Times online for the story.

Image: Judgment day begins…. from DeepBluC


Janna Rust April 21, 2010 at 7:21 am

Great article. Discrimination will never be gone because we are people who’ve created our beliefs based on our experiences. Matter of fact, a friend of mine and I years ago used to say we wanted the “no baby bonus” (aka the same pay men got for the same work) because we didn’t have children.

Unfortunately, our biases when hiring, especially those based upon first impressions such as a resume or 45 minute interview don’t always yield the best candidate. That’s why I recommend the use of validated behavioral assessments as part of any hiring prociess.

Tom N. April 21, 2010 at 7:37 am

Interesting reading, and certainly something for managers to keep in mind when reviewing new resumes. Having been involved in the hiring process a few times, I’m sure my own biases were reflected in some of my decisions, regardless of whether I was conscious of them or not at the time. However, I find that many times in the business world the decision of who to hire is not up to a single individual. In my own current workplace, resumes are reviewed by at least two people (typically a manager and a senior employee), followed by 2-3 interviews conducted by different individuals (or groups of individuals) each time. Granted there is still opportunity for bias, but by involving more employees in the decision making process hopefully this bias is lower.

I have to disagree with your point about the chance of a woman being hired based on taking her partners name or not. I read through the study you linked, and while I found it to be interesting and informative, I do not feel it accurately reflects the real world. If I read the name “Jane Smith” on a resume, I do not know whether she is a “Smith” by birth or by choice. For that matter, legally I am fairly certain I’m not even allowed to ask that question (at least in the US… I’m not sure if the laws are different in the Netherlands where the study was conducted). I will concede that the appearance of a hyphenated name may lead to the conclusion that the candidate is married, but have not personally encounter them very often.

As a side note, I found your last post about your children’s plea for an iPad to be absolutely adorable. I can only hope my own children will be so persuasive (and tech savy!).

– Tom

cv April 21, 2010 at 12:19 pm

Tom, thanks so much for your comments — I think that your point that having several folks reviewing resumes might reduce the chances of unconscious bias ‘getting past’ the screen sounds right. Anytime we have to articulate our reasoning we have a chance to uncover implicit biases…. this would be an interesting twist to test empirically too. Sometimes conversation reinforces bias, other times reveals it, and still other times reveals and fixes it. An ideal study would help us figure out what needed to happen to uncover biases and address them…

Also, you found the Achilles heel in the study (are you a social scientist yourself?). Here in the US we’re not supposed to care, or to ask, about marital status (and how would we know a woman’s spouses’ last name anyway?). So the direct application of the study;s conditions to the real world is a bit weaker here.

Still, though, the question of what we assume about a woman’s career commitment, and how we make these assumptions, is important.

I just think about the women I know who have been asked to sign contracts before they go out on maternity leave promising that if they don’t come back full time they’ll need to repay their paid leave…or women who are asked what their husbands do so that the interviewers can assess how likely it is that the family will move for the husband’s job…and now that we know that women married to men who work 60+ hours a week are more likely to quit or downshift than women married to men who don’t ‘overwork’… it just goes on and on. And those conversations happen, sometimes, unofficially, when hiring managers play by their own rules.

I also can think of a few folks who would say, with strong reasoning, that it makes sense to suss out a woman candidate’s career commitment, since hiring decisions are critical and since “let’s face it” women are more likely than men to downshift due to spousal & parental roles demands. I guess this is all fodder for another post! thanks for nudging me to think more about it. cvh
.-= cv´s last blog ..Resumes and Bad Judgment: More bias than you think =-.

cv April 21, 2010 at 12:24 pm

Hi Janna– I think I was also answering your comment too as I was responding to Tom. Behavioral assessments *are* better, esp better than interviews (notoriously not reliable but still everyone’s favorite screening tool). I am often asked by students what you should do to assess candidates, and I think about the ‘realistic job preview’ things I did when hunting for my first job. Boy that was a lot of work, for everyone involved.
If we took more seriously the wisdom about ‘getting the right people on the bus”, and kept in mind just how hard it is to fire folks who aren’t really working out well, maybe we’d feel more comfortable about doing the work to hire well. cvh
.-= cv´s last blog ..Resumes and Bad Judgment: More bias than you think =-.

Tom N. April 21, 2010 at 3:40 pm

“are you a social scientist yourself?”

No, although I am hoping to steer my career more towards organizational behavior and theory. As an undergrad my favorite classes were those that dealt with human nature, sociology, and the humanities. Although my current Master’s studies are more IT based (Information Systems), I haven’t lost any interest in the human side of the equation.

– Tom

Brain April 22, 2010 at 6:43 pm

I have twelve years of experience as a Recruiter (USA) in the real world. I find the studies mentioned in this blog entertaining, but not scientifically valid.
The race name study says the resumes were “almost identical” with the names being the biggest difference…..or at least it is assumed the names are the biggest difference in the resumes. We don’t see the resumes, so we don’t know if, say, the schools are different as well as the names on SOME resumes used in the study.
I can tell you as a Recruiter, that hiring mangers will hire someone from their Alma Mater, if all other things are equal between the best candidates. Is this a bias? Absolutely! But not one based on race or hyphenated names.
Handsome males and beautiful females also get jobs faster, get better offers, and tend to get promoted faster than their less attractive counterparts. Is this a bias? Absolutely! But not one based on race or hyphenated names.

What about the so-called Diversity initiatives so many companies have in the USA? Abbott Labs and Rockwell Automation, (together they employ around 50,000 employees) have Diversity “goals,” including, but not limited to Recruiters must submit a candidate pool that’s at least 50% female for hiring mangers to interview. The problem with this is that the workforce is usually only 35% or so female in the areas where these companies are hiring. So how do they achieve a goal of 50% female candidates to interview when the workforce in the geographical area is only 35% female? You achieve the goal by eliminating better qualified male candidates from those that are interviewed. I never see any articles about this type of unethical, immoral, and illegal bias. There are hundreds of companies that have Diversity goals and incentives like this one.

So, to the point of the article, as a Recruiter with twelve years of experience, I can tell you that there are biases in hiring, but not necessarily the race and gender biases that this article would have us believe.

physician assistant June 7, 2010 at 3:40 am

I’ve recently started a blog, the information you provide on this site has helped me tremendously. Thank you for all of your time & work.

Comments on this entry are closed.