3 ms·
As the article suggests, these "diversity and inclusion" initiatives have negative effects when they become institutional mandates. And yet we can all agree tha
by lkrubner 4y ago
As the article suggests, these "diversity and inclusion" initiatives have negative effects when they become institutional mandates. And yet we can all agree that there is a strong moral argument in favor of ensuring that each profession is open to everyone of real talent. My sense of the right way to approach this is to encourage every hiring manager to remember their moral obligations, but without formalizing such moral concerns into mandates.
I wrote a long bit about this in "One on one meetings are underrated:"
When hiring, I rely heavily on recommendations from people I know, or at least know of. To use some Silicon Valley jargon, I rely on "social proof." If a good engineer with whom I’ve worked recommends some other engineer or project manager or product manager, that counts for a lot with me.
I recently tweeted this idea on Twitter and James Youngman responded, “It seems to me that this kind of approach is what perpetuates the domination of the industry by an in-group (who know each other, directly or indirectly) at the expense of outsiders, to the detriment of both diversity and fairness.”
That’s a valid concern. We all have a professional and ethical obligation to be sure that we hire a diverse workforce; that is, the workforce must be open to anyone who has real talent. In my experience, there is no contradiction between that obligation and a requirement that a candidate be recommended by someone we trust. Assuming you have friends and colleagues who understand their professional and ethical obligations the same way you do, they should be able to give recommendations on people they know or put you in touch with friends of theirs who can offer such recommendations. Within your extended social penumbra of friends, friends of friends, and friends of friends of friends, you should be able to find someone who can vouch for most of the candidates that you need to hire. You just need to put in the effort, chasing down those recommendations through extended chains of acquaintances.
This approach very much works when you are hiring novices who are straight out of school. It helps to stay in touch with friends and colleagues who are teaching or mentoring at schools or the software developer bootcamps. Back in 2018, when I needed some junior-level frontend software developers, I hired several women from the Grace Hopper program for women that is run by Fullstack Academy. In that case I spoke with some experienced friends of mine who had either taught at the school or volunteered as mentors. As such, they could point me to those who were the best of the graduating class and I ended up with an unusually excellent team of novices.
[I then tell a story from 2009, when I go to a job interview, and they give me a technical test: to find the problems in their code. And yet, they didn't know what they were doing, and their code was horrible. Here is the conclusion of that story:]
...Needless to say, I didn’t get the job, nor did I want the job. He was looking to hire software developers who would be willing to follow his idiosyncratic and unprofessional style. I wasn’t interested.
I tell this story when I’m being critical of certain kinds of coding tests, and hiring managers respond, “Well, that is an extreme and ridiculous example.” It is indeed extreme and ridiculous, therefore, it’s easy to see the problem. Yet, even when tests are less idiosyncratic than that one, they will still reflect the values, skills, and aesthetics of the person doing the hiring. How could they not? Anyone who hires someone for their team will want someone who is at least somewhat consonant with the style and goals of the team. Wanting that consonance is reasonable, up to a point. Certainly, when I hire, I’m looking for someone who either shares my aesthetics or is willing to learn my aesthetics. That is, I explicitly recognize that there are subjective factors that shape hiring decisions.
Whenever I say this, I get pushback from people who say some variation of, “This is why tech remains exclusive, with a dominant in-group that never changes, because the people with the power to hire only hire those who are just like them.”
I, however, would say the opposite is true. The only way to make the profession of software development (or the profession of marketing or operations or sales or law or any other profession) more diverse is to explicitly recognize that all hiring contains a subjective element, and that one of the goals of hiring must be the creation of a diverse workforce, open to anyone with real talent. Only after you’ve explicitly recognized that there are subjective factors that shape hiring can you explicitly move to build a diverse workforce. By contrast, the pretense of objective tests has too often served as a smokescreen behind which lurk forces that conserve the status quo.
If the moral urgency of this issue leaves you unmoved, consider the practical element: everything you need to know about a candidate you can learn from recommendations and by talking to the candidate. If you ask them direct questions, and you ask follow up questions until you are sure you know everything that you need to know, then you will discover what you need without wasting time on tests. This is faster for you, and it is faster for the candidate, and therefore this is both more fair and more efficient for everyone.
This is not to say that I never use tests. I use tests all the time. Not because they are objective, but for the opposite reason: they often reveal some of the subjective factors that are essential for making good hiring decisions. If I am interviewing a junior level software developer, and I give them a challenge that only a senior level software developer could handle, I’m not actually looking to see if this developer can do the work. I know they can’t. I am looking to see if they panic, or if they remain calm and ask me all of the questions that they should be asking me.