4 ms·

This progressive-SV obsession with diversity is a pretty odd shibboleth from an outsider's perspective. Why do companies set this as a goal for themselves, or

This progressive-SV obsession with diversity is a pretty odd shibboleth from an outsider's perspective. Why do companies set this as a goal for themselves, or brag about it? There are only three ways to have a workforce significantly more diverse than your applicant pool: 1) Get lucky, and have the best candidates just happen to have the particular immutable characteristics of birth you think qualify them for a job. In which case, congratulations? But it's nothing to brag about. And it's only possible for low-n. The law of averages will wipe it out sooner or later. 2) Intentionally hire suboptimal candidates because they have the immutable characteristics of birth you think qualify them for a job. Congratulations, you made your company worse. 3) Spend an inordinate amount of time interviewing candidates in order to find ones who are both excellent and have the immutable characteristics of birth you think qualify them for a job. Congratulations, you made your company worse. Time and effort are finite resources and you should have been spending yours building better products and connecting with your customers. All this, for what? Plaudits in the tech press? Maybe attracting investment from VC firms that say they want their companies to employ people with a wide range of immutable characteristics of birth? You haven't made your products better, and you haven't connected with your customers more. You've either hired a worse team or wasted your time. If, for sufficiently large n, your workforce does not look like your applicant pool, you have broken hiring practices.

noobermin 10y ago
I think a large assumption you've made is what the make up of their applicant pool is, which I too have no idea what that looks like.
- redthrowaway 10y ago
  It's not an assumption I'm making at all. It's an independent variable. I'm assuming that talent is more or less uniformly distributed among applicants. If that assumption is true, then an optimal hiring process will, with sufficient hires, produce a workforce that broadly resembles the pool of applicants. More formally, under an optimal recruitment system, as the number of hires approaches infinity, the Levenshtein distance between the anatomical characteristics of the workforce and applicant pool approaches zero.
x1798DE 10y ago
I can imagine many scenarios where even without biasing your hiring that your company doesn't look, demographically, like your hiring pool. For example, if your candidate pool is 80% foreign and you have a limited number of visas to offer. One can also imagine that if you are a particularly "female-friendly" company, word of mouth and primarily-female contact networks (e.g. women who know each other through female-only meetups or "girls who code" style events) would naturally bias your most qualified candidates towards being women, particularly if almost all your competitors are either seen as "female unfriendly" and/or don't have access to the primarily-female contact network.
- drawnwren 10y ago
  Your first argument is satisfied by 2 and 3 of the post you replied to. Your second argument boils down to, 'perhaps they have a larger female applicant pool than you suspect.' However, it doesn't really seem support an argument that the company will look different than the applicant pool.
  - x1798DE 10y ago
    Well, I'm assuming "applicant pool" means the pool of all people applying. My point was that even if women are a small portion of the applicant pool, they may represent a larger quality-weighted portion of the total applicant pool. I agree that if as your population becomes large, you should asymptotically be approaching the demographics of your quality-weighted applicant pool. Note: By "quality-weighted" I really mean something more complicated like "ROI-weighted", since it could also be that because of your reputation as a "female-friendly" employer, high-quality female employees will be willing to work at a lower salary than they would otherwise demand at a "male-dominated" workplace (e.g. "Well, company X pays better, but I really like the culture/people/policies at company Y...")
- redthrowaway 10y ago
  How does one gain a reputation for being "female-friendly" (or hispanic-firendly, or gay-friendly, or whatever) without going through one of the three steps above? Personal networks are huge in hiring, but you have to have a lot of women working for you before you'll have access to a lot of female personal networks. And given how many companies in SV are scrambling to hire women, you have to work pretty damned hard to get them. If you were running a company in SV and wanted to get the best ROI on your time and money, you'd probably be better off hiring 50 year old straight white men who vote Republican and have been at it since the days of Netscape. Massive market inefficiencies exist in hiring. Zigging where your competitors zag is likely to pay off more than following the trend. Or, just throwing this out there: you could decide what you're really looking for in an employee and hire people who satisfy those criteria and ignore everything else. Perish the thought.
Klathmon 10y ago
Well there are a few other options. Like that some people won't work in a place where there is only one demographic, or there is a VERY large imbalance. Anecdotally I know a woman who refuses to work in a place where there aren't at least 2 other women that she will interact with on a daily basis. She's an amazing developer, but has been brought into male-only teams or teams with only one other female and it becomes an issue. And she's at a point in her life where she can require that of potential employers. So by striving for a more diverse workforce, you could be opening up doors to exceptionally good employees who have the ability to set the bar for what they want in a company. To put it more concisely, diversifying your employees can increase your hiring pool.
- nichochar 10y ago
  This example is good but surely an edge case. Optimizing for edge cases is typically wasted effort. I think a big question is why companies do it (there is a comment thread above tackling this). A company's goal, in a capitalistic country, is to maximize profits, if you can prove that a 50/50 split does this (and getting there should be included in the cost), then great.
  - Klathmon 10y ago
    I wouldn't be so sure it's an edge case. Would you want to work somewhere that nobody shared any interests? Maybe a language barrier? Perhaps a place where you would be bullied? or be talked down to constantly? If you had to choose between a place like that, or a place that you would "fit in" better, which would you choose? It stands to reason that the "best" candidates are going to have a choice in where they work. Hell, even "slightly above average" candidates can often have a choice between multiple offers. So when you are looking for talented individuals, you are going to either need to pay more, offer some other benefits, or will need to ensure your workplace doesn't make them feel unwelcome. Personally I think of diversifying your employees like buying a pingpong table for the break room, or paying for lunch a day or 2 a week. It's a small price to pay, that can pay off big time when hiring. And that seems like a big enough "reason" for me to want to do it.
  - geofft 10y ago
warcher 10y ago
Everybody's hiring practices are pretty broken. As a white male engineer, I've never really had any success even getting an interview with ordinary tech companies out of engineering school. Had to start my own. Going ok, has been for many years now, which leads me to think all those guys throwing my resume in the trash may have some false negatives in their system somewhere. I think there are objective benefits to fixing some of the weird culture in tech, including but not limited to (at least) doubling the available talent pool. And the objective truth that having more ideas and more viewpoints makes for much more effective teams. I'm not so stupid as to think there aren't costs associated with bringing populations that aren't well represented in. There's gonna be catchup. You're going to have to develop talent. It is going to take a while. Engineering and entrepreneurship is absolutely a hard job, and most folks aren't cut out for it, not really. So widening the net is going to take work. And bigger, more established companies are probably better equipped to shoulder more of that investment, but smaller, hungrier companies can ALSO afford to take risks on talent that has potential, but isn't a 100% safe bet.
- warcher 10y ago
  I'll say as somebody low enough on the food chain to need to work harder at hiring, but high enough that I do make those decisions, I actively seek out folks outside the usual boxes. I look real hard for people that aren't guaranteed first-round draft picks, and who may be overlooked by the system. There are tons of great hires who aren't getting any love, you just have to know what you're really looking for, do the work, and trust your instincts. I've seen a lot of sloppy, lazy, highly risk-averse work in hiring and talent development over the years. Everybody's trying to not look stupid, not to win.
- redthrowaway 10y ago
  >And the objective truth that having more ideas and more viewpoints makes for much more effective teams. This seems to be an article of faith among the diversophiles, and it seems to me to require a lot more supporting evidence than is typically proffered. First, the notion that a diversity of anatomical characteristics leads to a diversity of thought seems specious. College campuses are wonderfully diverse in ethnicity and gender and sexuality, but strikingly uniform in belief systems. Whereas the American Founding Fathers were all straight white male aristocrats, and yet had vast and meaningful ideological differences. If diversity of thought is what you seek, why optimize for the dubious proxy of race and sex? Why not just seek out candidates with vastly different belief systems to your own? Hire conservatives and progressives and liberals and socialists and libertarians. Second, the notion that diversity of belief systems leads to success in engineering endeavours also strikes me as needing far more support than its proponents provide. Nazi Germany had a brutally enforced ideological homogeneity, and yet produced a rocketry program that was decades ahead of anyone else. Japan has a very rigid corporate culture, and yet has a well-earned reputation for engineering excellence. And the most successful companies in Silicon Valley have a very clear and well defined corporate identity, vision, and mission to the point where jokes are made about the odd ways their employees all think alike. It would seem that ideological unity and cohesion is actually more correlated with engineering success than diversity of thought. Again, this is all from the perspective of an outsider. I don't live in Silicon Valley (or even the United States) and I don't identify with the vast majority of its residents or companies. My critiques are made from the perspective of a disinterested observer at 10,000 ft.
RoboticFoot 10y ago
> Why do companies set this as a goal for themselves, or brag about it? A lot of tech people I know feel a lot of guilt because of their success. They feel as though they don't deserve it, so they look for ways to "blame" their success on external factors. One of the ways this manifests is a strong desire to politic against anything that might ostensibly be perceived as unfair. In the strongly left-leaning SV community, things like institutional sexism are very obvious targets to project internal feelings of guilt onto. I'm obviously not a psychologist or anything, but this is just my personal explanation I've built up after a few years living in the Bay Area.
geofft 10y ago
> If, for sufficiently large n, your workforce does not look like your applicant pool, you have broken hiring practices. Well, one, to first order everyone in the Valley has badly broken hiring practices. Two, I don't think that's true. It's extremely common for, say, top universities' freshman classes to look nothing like their applicant pools, because their applicant pools are full of people who want to go to that university but genuinely wouldn't do well there. There are a lot of people who optimize for making good grades and a perfect SAT score and have nothing to characterize them otherwise, and I would expect there are way fewer of those people, by percentage, in the freshman classes. (And yes, the modern university application process came out of anti-Semitism and there are a lot of other indefensible things about it, but this effect, I think, is defensible.) My personal hypothesis is that technical aptitude is pretty uncorrelated with demographics, but that due to (for lack of a better term) "the pipeline", certain demographics are more encouraged to stay with technical studies and apply to these roles than others. So you have demographic A, which consists of a large number of people of varying abilities, and demographic B, which is a smaller number of people who have done particularly well throughout the pipeline. A fair hiring process, that selects for people with high aptitude, would naturally reject many more people from group A (who weren't filtered out from the pipeline) than from group B.

noobermin 10y ago

I think a large assumption you've made is what the make up of their applicant pool is, which I too have no idea what that looks like.

redthrowaway 10y ago
It's not an assumption I'm making at all. It's an independent variable. I'm assuming that talent is more or less uniformly distributed among applicants. If that assumption is true, then an optimal hiring process will, with sufficient hires, produce a workforce that broadly resembles the pool of applicants. More formally, under an optimal recruitment system, as the number of hires approaches infinity, the Levenshtein distance between the anatomical characteristics of the workforce and applicant pool approaches zero.

x1798DE 10y ago

I can imagine many scenarios where even without biasing your hiring that your company doesn't look, demographically, like your hiring pool. For example, if your candidate pool is 80% foreign and you have a limited number of visas to offer. One can also imagine that if you are a particularly "female-friendly" company, word of mouth and primarily-female contact networks (e.g. women who know each other through female-only meetups or "girls who code" style events) would naturally bias your most qualified candidates towards being women, particularly if almost all your competitors are either seen as "female unfriendly" and/or don't have access to the primarily-female contact network.

drawnwren 10y ago
Your first argument is satisfied by 2 and 3 of the post you replied to. Your second argument boils down to, 'perhaps they have a larger female applicant pool than you suspect.' However, it doesn't really seem support an argument that the company will look different than the applicant pool.
- x1798DE 10y ago
  Well, I'm assuming "applicant pool" means the pool of all people applying. My point was that even if women are a small portion of the applicant pool, they may represent a larger quality-weighted portion of the total applicant pool. I agree that if as your population becomes large, you should asymptotically be approaching the demographics of your quality-weighted applicant pool. Note: By "quality-weighted" I really mean something more complicated like "ROI-weighted", since it could also be that because of your reputation as a "female-friendly" employer, high-quality female employees will be willing to work at a lower salary than they would otherwise demand at a "male-dominated" workplace (e.g. "Well, company X pays better, but I really like the culture/people/policies at company Y...")
redthrowaway 10y ago
How does one gain a reputation for being "female-friendly" (or hispanic-firendly, or gay-friendly, or whatever) without going through one of the three steps above? Personal networks are huge in hiring, but you have to have a lot of women working for you before you'll have access to a lot of female personal networks. And given how many companies in SV are scrambling to hire women, you have to work pretty damned hard to get them. If you were running a company in SV and wanted to get the best ROI on your time and money, you'd probably be better off hiring 50 year old straight white men who vote Republican and have been at it since the days of Netscape. Massive market inefficiencies exist in hiring. Zigging where your competitors zag is likely to pay off more than following the trend. Or, just throwing this out there: you could decide what you're really looking for in an employee and hire people who satisfy those criteria and ignore everything else. Perish the thought.

Klathmon 10y ago

Well there are a few other options. Like that some people won't work in a place where there is only one demographic, or there is a VERY large imbalance. Anecdotally I know a woman who refuses to work in a place where there aren't at least 2 other women that she will interact with on a daily basis. She's an amazing developer, but has been brought into male-only teams or teams with only one other female and it becomes an issue. And she's at a point in her life where she can require that of potential employers. So by striving for a more diverse workforce, you could be opening up doors to exceptionally good employees who have the ability to set the bar for what they want in a company. To put it more concisely, diversifying your employees can increase your hiring pool.

nichochar 10y ago
This example is good but surely an edge case. Optimizing for edge cases is typically wasted effort. I think a big question is why companies do it (there is a comment thread above tackling this). A company's goal, in a capitalistic country, is to maximize profits, if you can prove that a 50/50 split does this (and getting there should be included in the cost), then great.
- Klathmon 10y ago
  I wouldn't be so sure it's an edge case. Would you want to work somewhere that nobody shared any interests? Maybe a language barrier? Perhaps a place where you would be bullied? or be talked down to constantly? If you had to choose between a place like that, or a place that you would "fit in" better, which would you choose? It stands to reason that the "best" candidates are going to have a choice in where they work. Hell, even "slightly above average" candidates can often have a choice between multiple offers. So when you are looking for talented individuals, you are going to either need to pay more, offer some other benefits, or will need to ensure your workplace doesn't make them feel unwelcome. Personally I think of diversifying your employees like buying a pingpong table for the break room, or paying for lunch a day or 2 a week. It's a small price to pay, that can pay off big time when hiring. And that seems like a big enough "reason" for me to want to do it.
- geofft 10y ago

warcher 10y ago

Everybody's hiring practices are pretty broken. As a white male engineer, I've never really had any success even getting an interview with ordinary tech companies out of engineering school. Had to start my own. Going ok, has been for many years now, which leads me to think all those guys throwing my resume in the trash may have some false negatives in their system somewhere. I think there are objective benefits to fixing some of the weird culture in tech, including but not limited to (at least) doubling the available talent pool. And the objective truth that having more ideas and more viewpoints makes for much more effective teams. I'm not so stupid as to think there aren't costs associated with bringing populations that aren't well represented in. There's gonna be catchup. You're going to have to develop talent. It is going to take a while. Engineering and entrepreneurship is absolutely a hard job, and most folks aren't cut out for it, not really. So widening the net is going to take work. And bigger, more established companies are probably better equipped to shoulder more of that investment, but smaller, hungrier companies can ALSO afford to take risks on talent that has potential, but isn't a 100% safe bet.

warcher 10y ago
I'll say as somebody low enough on the food chain to need to work harder at hiring, but high enough that I do make those decisions, I actively seek out folks outside the usual boxes. I look real hard for people that aren't guaranteed first-round draft picks, and who may be overlooked by the system. There are tons of great hires who aren't getting any love, you just have to know what you're really looking for, do the work, and trust your instincts. I've seen a lot of sloppy, lazy, highly risk-averse work in hiring and talent development over the years. Everybody's trying to not look stupid, not to win.
redthrowaway 10y ago
>And the objective truth that having more ideas and more viewpoints makes for much more effective teams. This seems to be an article of faith among the diversophiles, and it seems to me to require a lot more supporting evidence than is typically proffered. First, the notion that a diversity of anatomical characteristics leads to a diversity of thought seems specious. College campuses are wonderfully diverse in ethnicity and gender and sexuality, but strikingly uniform in belief systems. Whereas the American Founding Fathers were all straight white male aristocrats, and yet had vast and meaningful ideological differences. If diversity of thought is what you seek, why optimize for the dubious proxy of race and sex? Why not just seek out candidates with vastly different belief systems to your own? Hire conservatives and progressives and liberals and socialists and libertarians. Second, the notion that diversity of belief systems leads to success in engineering endeavours also strikes me as needing far more support than its proponents provide. Nazi Germany had a brutally enforced ideological homogeneity, and yet produced a rocketry program that was decades ahead of anyone else. Japan has a very rigid corporate culture, and yet has a well-earned reputation for engineering excellence. And the most successful companies in Silicon Valley have a very clear and well defined corporate identity, vision, and mission to the point where jokes are made about the odd ways their employees all think alike. It would seem that ideological unity and cohesion is actually more correlated with engineering success than diversity of thought. Again, this is all from the perspective of an outsider. I don't live in Silicon Valley (or even the United States) and I don't identify with the vast majority of its residents or companies. My critiques are made from the perspective of a disinterested observer at 10,000 ft.

RoboticFoot 10y ago

> Why do companies set this as a goal for themselves, or brag about it? A lot of tech people I know feel a lot of guilt because of their success. They feel as though they don't deserve it, so they look for ways to "blame" their success on external factors. One of the ways this manifests is a strong desire to politic against anything that might ostensibly be perceived as unfair. In the strongly left-leaning SV community, things like institutional sexism are very obvious targets to project internal feelings of guilt onto. I'm obviously not a psychologist or anything, but this is just my personal explanation I've built up after a few years living in the Bay Area.

geofft 10y ago

> If, for sufficiently large n, your workforce does not look like your applicant pool, you have broken hiring practices. Well, one, to first order everyone in the Valley has badly broken hiring practices. Two, I don't think that's true. It's extremely common for, say, top universities' freshman classes to look nothing like their applicant pools, because their applicant pools are full of people who want to go to that university but genuinely wouldn't do well there. There are a lot of people who optimize for making good grades and a perfect SAT score and have nothing to characterize them otherwise, and I would expect there are way fewer of those people, by percentage, in the freshman classes. (And yes, the modern university application process came out of anti-Semitism and there are a lot of other indefensible things about it, but this effect, I think, is defensible.) My personal hypothesis is that technical aptitude is pretty uncorrelated with demographics, but that due to (for lack of a better term) "the pipeline", certain demographics are more encouraged to stay with technical studies and apply to these roles than others. So you have demographic A, which consists of a large number of people of varying abilities, and demographic B, which is a smaller number of people who have done particularly well throughout the pipeline. A fair hiring process, that selects for people with high aptitude, would naturally reject many more people from group A (who weren't filtered out from the pipeline) than from group B.