“Blind Hiring” to Avoid Bias: Wave of the HR Future, or “Blind Alley”?
Is a hiring manager named “David” more likely to give an edge to a job candidate also named “David”?
Recent academic research suggests that the answer is “yes,” even for other names with similar sounds, like “Dan” or “Dustin.”
What does this say about a company’s ability to develop a merit-based, non-discriminatory hiring model?
In it, Hoffman discusses the provocative conclusions of a Wharton School of Business study and other research showing that factors having nothing whatsoever to do with “merit” can influence some pretty profound life decisions: how we vote, where we live, the profession we aspire to; and — last but not least — whom we hire — in statistically significant ways.
These are factors as seemingly inconsequential as the letters of our given name. The research involves the concept of “priming.”
The Wharton Study: Where We Vote Can Influence How we Vote
A study just published on July 1, 2008, found that voters were more likely to support a school funding initiative on the ballot, simply by virtue of having been assigned to vote in a school!
The study, “Contextual Priming: Where People Vote Affects How They Vote,” used regression analysis and other statistical means to control for such factors as the liberal voting patterns of certain regions, and the possibility that people living near schools (hence, assigned to vote there) were more likely to vote pro-education.
In other words, just being in a school environment for the purpose of voting influenced voters’ decisions positively towards funding education. The study authors — from the Wharton School at the University of Pennsylvania; MIT; and the Stanford Graduate School of Business — concluded:
These results illustrate the dramatic and unexpected influence that the environment can have on behavior. Seemingly trivial environmental contexts were found to have significant effects on consequential real-world decision making. The type of polling location where people were assigned to vote influenced how they ended up casting their ballot. . . . [T]hese effects can occur outside of conscious awareness.
Priming research has primarily been conducted in carefully controlled laboratory settings, but this paper extends this literature to demonstrate how priming plays out in complex, messy real-world environments. Everyday environments contain numerous cues that can activate various conflicting constructs in memory.
What is “Priming”?
While “priming” has somewhat varying definitions in psychology, the online glossary of the American Psychological Association defines it as “the advantage conferred by prior exposure to a word or situation.”
In less academic jargon, we can find ourselves more or less receptive to just about anything or anyone based on prior exposure having nothing to do with the current situation.
In the Wharton research cited above, for example, simply by walking into a school environment, voters were more likely to vote for an education funding initiative.
Wikipedia gives a further example of priming:
[W]hen a subject reads a list of words including the word “table,” and is later asked to complete a word that starts with “tab . . . ” , the list “primes” the subject to answer “table,” as shown by evidence the probability that the “primed” subject answers “table” is higher than for non-“primed” subjects.
What’s in a Name? Everything, it Seems
The stimuli that can lead to a priming reaction can be amazingly simple.
Auren Hoffman also points to “a groundbreaking study” by the University of Buffalo’s Associate Professor and Psychologist Brett Pelham concluding that “some of the biggest decisions of our life — where we live, what we do, and who we marry — are influenced by our first name.”
Pelham’s summary of his research states that people are disproportionately likely to:
- live in states or cities resembling their names (e.g., people named Louis are especially likely to live in St. Louis);
- have careers that resemble their names (e.g., people named Dennis, Denis, Denise and Dena are all especially likely to be dentists);
- marry other people whose last names begin with the same letter as their own. A similar, but weaker, matching effect also occurs for people’s first names.
A January 2007 UK Telegraph article pointed up even more incredible linkages in the power of names as revealed in Pelham’s work:
According to Pelham, the same effect even extends into politics. During the 2000 US presidential campaign, people whose surnames began with B were especially likely to make contributions to the Bush campaign, whereas those whose surnames began with G were more likely to help Gore.
Hoffman also cites further evidence of the name effect from the book The Happiness Hypothesis by Jonathan Haidt, as Haidt — referring to Pelham’s research — writes: “Men named Lawrence and women named Laurie are more likely to become lawyers.”
Haidt suggests that name similarity causes a “spark of happiness” that unconsciously affects decisionmaking.
As another apt example of name “priming,” Hoffman suggests that we might actually decide to see a movie simply because the star has our same surname. Hoffman notes that he might be primed to see a movie starring Dustin Hoffman: not only does he share an identical last name with the actor (“Hoffman”), but both his first name and the star’s have a similar syllabic ring (Auren, Dustin).
The “Gut Reaction”
Hoffman — in summarizing the effect of “priming” in relation to that ever-mysterious concept –- the “gut reaction” — goes on to state:
A gut reaction is generally a collection of biases and can be easily primed. While it can be right (the brain can often analyze information implicitly faster than it can explicitly), it can also be dangerously wrong. It would be a really bad idea to hire someone to watch over your child just because you got a “good feeling” about the person.
Your gut might be much better at telling you what not to do than giving you good direction on what to do. If your gut tells you something is wrong with someone, than you probably do not want to entrust your kid with her. But a positive gut-check often does little good (at least for me).
When thinking about how this affects hiring, our goal at Rapleaf [where Hoffman is CEO] is to attempt to remove primed biases from hiring decisions. While you’ll never be able to remove all bias, removing just a few of them can give you a dramatically large advantage over a competitor. …
As advanced decision makers, we need to make sure we are making important decisions for the right reasons and not just because of being primed. …
In hiring, positive gut feelings can lead to decisions based on superficial similarities to the decision maker, including those that may not be lawfully considered, such as age, race, national origin. Equally importantly, they can lead to decisions based on a host of other irrelevant factors, such as hair color.
Either way, Hoffman is absolutely correct that such decisions are bad business practices that can confer a competitive disadvantage.
The symphony screen
Hoffman goes on to discuss a third empirical example of unconscious bias:
Malcolm Gladwell has a great anecdote … in Blink: The Power of Thinking Without Thinking, where a metropolitan symphony decides to change its hiring by listening to someone play (person was behind a screen) rather than seeing them play.
It turned out that the symphony in question massively increased the number of women they hired when they stopped watching people play and instead just listened to them. And, of course, the quality of the music got much better too.
So What About Blind Hiring?
The logical extension of all this intriguing information is to ask what kind of “blind” hiring process would be analogous to the symphony screen — deliberately shutting one’s eyes to irrelevant information so as to focus only on the relevant.
For instance, how many of us have been interviewed by someone with the same first name? Can we deny the subtle yet unmistakable note of warmth that gave us a slight but crucial advantage in those important first few seconds?
One could set up interviews so names are not provided to interviewers, but it would feel awkward to try and skirt the ingrained social pleasantry of greeting someone by name.
And in any event, how can one interview someone without noticing that their mannerisms remind one, for instance, of one’s least favorite uncle “Edward”? Or that they sound uncannily like one’s favorite aunt “Nia”?
There are simply way too many factors that may allow unconscious bias to creep into an in-person interview to be able to successfully eliminate them all.
Sally Ride’s Opinion
Sally Ride, the first American woman in space, was asked by U.S.A. Today about why girls don’t pursue degrees in math and science fields at the same rate as men, and whether businesses should consider “blind hiring.”
She noted that “it’s really difficult to do gender-blind hiring. You can do gender-blind initial screening, but the most effective thing would be to make managers aware that they have preconceptions.”
Ride noted that the same “symphony effect” noted in the Malcolm Gladwell anecdote, above, occurs with science articles reviewed for publication. “When authors’ names were removed, the number that were accepted from women rose. Well-intentioned, educated people can have biases that they don’t know they have,” Ride said.
(Speaking of names and choice of careers, anyone else find it interesting that choosing to be an astronaut allowed Sally Ride to take the ultimate “ride” — a space mission?)
The “Anonymous Resume” in France
In France, a corporate takeover artist, Claude Bebear, headed up a commission to give recommendations on how to end job discrimination in the country. The report – “Visible minorities: Addressing the challenge of access to employment and integration in the workplace” – was completed in 2004.
Bebear advocates the “anonymous resume,” according to a January 29, 2006 Washington Post article:
Typically, in France, “they throw away the résumés of people who are from bad parts of town which are supposed to have Arabs or blacks,” Bebear, 70, said in an interview. “When you have somebody whose name is Mohammed and he lives in St. Denis,” a low-income community outside Paris, “you say, ‘I won’t bother with that one,’ and so they don’t even answer them.”
The solution, Bebear said, is to strip résumés of anything that could tip off recruiters to a person’s racial, ethnic and national background or other information that could be used to discriminate — name, age, sex, even residential postal code.
Bebear has made his company, AXA (“a 112,000-employee behemoth that receives 40,000 resumes a year in France alone”) “a testing ground for anonymous resumes.”
Is it Possible to Have a “Blind” Hiring System That Still Allows Assessment of Intelligence, Creativity, and Personality?
Barring consideration of certain categories of information raises tough issues of what is and is not relevant to hiring.
A truly blind resume, for instance, might eliminate the names of colleges and schools attended by a candidate. This could reduce the “clubbiness” of certain firms and encourage fair consideration of candidates who may not have had the financial means to attend a top tier school. But academic degrees robbed of the name of the institution that awarded them arguably have less meaning.
Between a totally “blind” resume and a totally subjective process with no safeguards to ensure objective hiring, there are many ways to improve a hiring process so as to increase the chances of making merit-based, non-discriminatory hiring decisions.
Screening Online Applications
The online application can be a tool for increasing objectivity at the initial screening stage. It can be used so as to limit such screening to specified categories of relevant data about the applicants. The employer can conceal from ther screener, for instance, the “name” and “address” fields.
So used, the online application ensures that the screener receives information in uniform fields that can be easily compared among applicants, and it also puts into place a few more layers of objectivity.
A benefit to employers of using such a “blind” screening process is that it makes hiring decisions much more defensible legally. The more “blind” the process is in the initial screening stages, the more difficult it will be for an applicant rejected at this stage to prove intentional discrimination.
All applicants eliminated during blind screening will be subject to an absolute defense: that the employer didn’t know about age, sex, nationality, or other prohibited characteristics. Without such knowledge, obviously, there cannot have been intentional discrimination.
After a candidate passes intial screening, they could progress to an intermediate stage of a telephone interview, providing yet another layer of objectivity by avoiding visual cues that could stimulate positive or negative reactions based on primed responses not relevant to job suitability.
Even at the in-person interview stage, interviewers’ idiosyncratic biases (positive and negative) can be neutralized in a few different ways.
For one, the applicant might be interviewed by multiple interviewers of diverse backgrounds, whose biases may be expected to cancel each other out.
In addition, a “structured interview” format might be used. During a structured interview, all applicants are asked the same questions, and are scored on the same scale. In addition, the interviewer usually has a description or analysis of the education, interpersonal skills, technical skills, etc., that an “ideal candidate” would have. There are many advantages to use of structured interviews.
It seems that ultimately the best “blind hiring” system will actively strive not merely to eliminate all markers of individuality from resumes but to emphasize diversity among hiring decision makers, as well as diversity training. The goals should be to reduce opportunities for bias; to make the unconscious conscious, as much as possible; and to try to neutralize bias with diverse input.