For an industry that prides itself on its analytical ability and abstract mental processing, we often don't do a great job applying that mental skill to the most important element of the programmer's tool chest - ourselves.
It's a staple of the modern software development interview: the “culture fit” portion.
In many ways, it remains the last real “interview” part of the interview. Unlike a number of other positions (sales, marketing, management, and so on), a software development interview is usually chock full of technical parts that have far more binary elements to them: You either know how to use MongoDB or you don't; you either know how to implement a quicksort or you don't; you either know how to code a binary tree at a whiteboard with no help from Google or StackOverflow or even your other teammates when you will never do this again in your life unless you interview again… or you don't.
But then, once your technical skill has been determined, there's still the notion of “culture fit.” Will you get along with the other team members? Will you be able to share in the jokes? Will you be able to be a positive force within the team? Or will you be the prima donna, the annoying one, the one that just doesn't seem to fit in with the rest of the crowd? This is an important part of team chemistry, and so therefore a necessary part of the interview. Are you a good culture fit for this position?
Apparently, it's the last great opportunity for discrimination and exclusion to happen.
Building an Orchestra
As industries go, not much really seems to overlap between the software development industry and the orchestral arm of the music industry. One is so very much bigger than the other, one is so clearly an “entertainment art” where the other is a staple of industries all across the spectrum, and one has a strict dress code (tuxedos and cocktail gowns) while the other… well, it also has a strict dress code, but at the other end of the spectrum (T-shirts and jeans).
Even the very nature of delivery is radically different. An orchestra practices the same symphony over and over again for months in hopes that they will deliver an amazing performance, when they will start the symphony all over again, and begin from the very first note, and a software development team spends months writing lines of code that collectively and cumulatively create a deterministic (we hope!) result on the day when we run the final demo and deliver the software into production where it will operate entirely without any additional input from us.
Having said that, though, the nature of the interview for a position in the orchestra is not all that different from that of interviewing a software developer. The orchestra member clearly must have a level of technical skill - can you reach a high C on your instrument, can you demonstrate an ability to sight-read (that is, play music having never seen it before), can you adjust your style and interpretation of the music to match the direction being given by the conductor, and so on. And for years, orchestra managers have maintained that a culture fit is a necessary component of an orchestra - that the various members of each section must have a certain emotional connection to each other in order to fit in well as a whole.
All of this is why this recent study, as described by a New York Times article on interviews, is so important - and disturbing - to read:
In the 1970s, symphony orchestras were still made up almost exclusively of white men - directors claimed they were the only ones qualified. Around that time, many began to use a new method of hiring musicians: blind auditions. Musicians auditioned behind screens so the judges couldn't see what they looked like, and walked on carpeted floors so the judges couldn't determine if they were women or men - the women often wore heels. The Boston Symphony Orchestra pioneered the practice in 1952, and more orchestras began using it after a high-profile racial discrimination case was brought by two black musicians against the New York Philharmonic in 1969. Researchers from Harvard and Princeton took notice and studied the results; they found that blind auditions increased the likelihood that a woman would be hired by between 25 and 46 percent. In fact, with blind auditions, women became slightly more likely to be hired than men. (http://www.nytimes.com/2016/02/28/magazine/is-blind-hiring-the-best-hiring.html)
If you're a software developer, and you're not starting to see the parallels with our own industry, it's because you haven't been paying attention to the news more recently.
Still not convinced?
Researchers from the Paris School of Economics and Stanford University sent out fake resumes to apply for real jobs in Paris. All six resumes detailed identical work experience. The only differentiator was language skills on two of the resumes. The two French-sounding names received 70% more callbacks than the other four names - two of North African origin, and two that sounded foreign, but were hard to place. “Foreign applicants, whether their speci?c minority group is identi?ed or not, are equally disadvantaged as compared to French applicants across all dimensions under study - for both genders, and whether or not more information is available in the application,” the paper found. The reason? Homophily ? a preference for people who are more like you.
Unfortunately, findings like these are not limited to our French counterparts. American researchers had similar conclusions in a study: “Are Emily and Brendan More Employable than Lakisha and Jamal?” Applicants with white-sounding names were 50% more likely to get called for an initial interview than applicants with black-sounding names. Applicants with white names need to send about 10 resumes to get one callback. Applicants with black names need to send about 15 resumes to achieve the same result. (http://www.forbes.com/sites/ruchikatulshyan/2014/06/13/have-a-foreign-sounding-name-change-it-to-get-a-job/)
That's pretty disturbing, but the deeper question, for me, is this: Am I guilty of the same thing?
At my consulting company (full disclosure: I am no longer the CTO there, for unrelated reasons), we went to great lengths to ensure that we weren't falling into this “single-culturism” trap. There was even an effort in our spoken conversation, when talking about candidates, to attempt to offset any sort of hidden bias. And yet…
When the company reached 20 people in total, I looked around and took a small inventory of our staff. Nothing formal, nothing HR-sensitive, just sitting back in my chair one day and starting looking around. What I found was that our teams were made up principally of white men. Granted, we had two women programmers and one woman business analyst (all three of Caucasian heritage), and we had one Asian male programmer, so we weren't entirely uniform, but it was pretty close. Our age range was a little more spread out (four programmers in their forties or fifties, by my estimation, not including the founders), but still, a pretty dominant cluster in the late-twentysomething/early-thirtysomething range.
We were, in essence, a pretty non-diverse company, when the final accounting came in. I have to wonder, how much of that was “just the luck of the draw” in terms of who applied, and how much was homophily at work?
Mediation
In many ways, one can argue that the industry is made up predominantly of white men because that's the only ones who are qualified - but the Github study seems to disprove that pretty easily. We can argue that cultural fit is an important part of employed life, but if that culture is a predominantly white male one, the chances of bringing additional cultures in is pretty small. Why should that matter? Because having a diverse workplace yields better results:
Firms with diversity tend to innovate and out-perform others, a Center for Talent Innovation study (http://www.talentinnovation.org/assets/IDMG-ExecSummFINAL-CTI.pdf) finds. (Editor's Note: The original link no longer works. The site is now Coqual.org and research is at https://coqual.org/our-research/.) Employees at these companies are 45% likelier to report that firm's market share grew over the previous year. They are 70% likelier to report that the firm captured a new market. Plus, a team with a member who shares a client's ethnicity is 152% likelier than another team to understand that client.
Maybe… just maybe… it's time to rethink our interviews.
Summary
This all begs the question: How do we do this?
I think maybe it's time to start considering dropping cultural fit as a candidate criteria point. It's of dubious nature to begin with - after all, if diversity is a corporate-endorsed value, then cultural fit is something of a contradiction in terms anyway. Let's be honest; most of the time, what most companies imply via cultural fit is really two things: work ethic and “passion” for the business. Work ethic is something an employer won't be able to ascertain for several months on the job at minimum, and “passion”...well, I've made my feelings regarding employee “passion” quite clear in previous articles. So let's put in a clause that states that an offer is conditional for the first three to six months pending a review (in order to give the company an easy out if they find that the work ethic isn't there), drop cultural fit from the list of candidate criteria, and start hiring blind.
Seriously. Drop names and any personal information off the resume, test them on their skills, and hire based on that and nothing else. Because isn't that what we're claiming we're doing already?
It's unlikely that this will change overnight. HR departments are notoriously slow to adopt to change, and it will take several iterations of the studies mentioned above before this starts to sink in to the business world's collective unconscious. But for an industry that constantly considers itself a meritocracy, I think it's high time we stop waiting, and start leading.
Give it a shot.
