Women Less Likely

On sexism in the tech industry, but also some frustration

Imran H
Imran H
Aug 8, 2017 · 11 min read

I.

Gizmodo publishes a Googler’s internal memo named “Google’s Ideological Echo Chamber”. Gizmodo and most of the media reporting on this name it an “anti-diversity screed/manifesto/rant/whatever” because, you know, sensationalizing.

My initial reaction upon reading this was that it has, while somewhat good-intentioned, a few flaws and misinterpretations in its argumentation and does a bit too much handwaving in getting to its conclusion but overall carries across a good point. I don’t think it’s surprising that, given a few choice quotes, some have chosen to misrepresent this as anti-diversity, because I think it’s quite the opposite. And frankly, I think that calling this an “anti-diversity screed”, like a lot of things the media does to generate clicks, lends itself to letting people get outraged by it without reading it first, reading the article and maaaaaaaaybe the actual memo with that outrage, then letting the outrage snowball into God knows what.

(but hey, it works, doesn’t it?)

The memo goes something like this (paraphrasing the Googler’s words here):

  1. People generally are well-intentioned, but can be subject to biases. Google and, more to the point, the political left and right have biases that, while aren’t 100% good, are also both necessary.
  2. We’re told implicit biases exist, but accounting for universal (i.e. not cultural) physiological differences between men and women might paint a better picture. Women, on average, are more people- and feeling-oriented and more cooperative, whereas men are more thing-oriented, systematizing, and competitive. The gender gap does not necessarily imply sexism, but might be the result of these differences.
  3. Google has made strides to increase women in tech, but as long as tech remains high-status, men will want them more. While feminism has made strides in freeing women from the female gender role, men are much tied to their own.
  4. While believing in gender and racial diversity, Google’s practices to create more equal representation are discriminatory, prioritizing some groups over others. Seniors might say this is morally and economically good, but it’s leftist bias at work.
  5. Humanities and the social sciences very overwhelmingly lean left, and that creates confirmation bias, resulting in some myths being maintained in those fields. Google’s left bias makes it blind to this and uses findings from these sciences to justify politicized programs. While the left has an affinity for the “weak”, this same compassion creates an environment for political correctness, stifling discourse.
  6. Google needs to de-moralize diversity (i.e. stop treating critics of diversity programs as immoral), stop alienating conservatives in the interest of viewpoint diversity, confront Google’s biases, stop restricting programs to certain genders or races and focus more on psychological safety, have an open discussion about diversity programs, de-emphasize empathy (i.e. decrease reliance on anecdotes), prioritize intention (microaggression training leads to authoritarian tendencies without being backed by evidence), be open about human nature, and reconsider mandatory implicit bias training (might introduce overcorrecting and is haunted by political bias).

(read the entire thing though, this probably isn’t the best tl;dr)

I think points 2 and 3 are the ones that people most take issue with, and point 4 is the one that leads the media to label this as “anti-diversity”. But I think doing that kind of proves the memo’s point about the conversation being dominated by leftist bias while attempting to dismiss contrary viewpoints as bigoted. I wouldn’t be surprised if people called this memo sexist. People already throw words like that around lightly enough as it is.

Mainly I think it was primarily about how inequality in outcome does not necessarily imply discrimination, but might be a result of people making choices following their own character. This is also related to the gender wage gap (at least in the US), which can partially be explained by people making their own choices (more on that later).

I don’t think these points should lie outside the realm of Things That Are Appropriate To Talk About. Either they aren’t true, and counter-evidence exists to prove as much, or they are true, or at least partially so, and Google needs to adjust their efforts. Either way, this can’t really go anywhere but underground by being shoved under the rug, can it?

It’s important to note that the author mentioned that:

I value diversity and inclusion, am not denying that sexism exists, and don’t endorse using stereotypes.

I don’t think the author is sexist, and I don’t think they were trying to propagate sexism in publishing this memo. I think this was a good faith effort to get at least something started, but they just lacked tact in trying, which led to the reactions that we have today.

I agree with some points that it makes, and disagree with others. I can appreciate the memo’s point about sexism not explaining everything. I don’t really get the impression from this memo that physiology explains everything, however, and the author acknowledges workplace sexism. So it has to be somewhere in between.

The problem is: where? How much of the gap does sexism account for? 100%? 75%? 50%? Given how long we’ve had this debate over sexism in the wage gap, I don’t think that this is 1) an easy question to answer and 2) something a single number will solve. This is where I think the handwaving comes in. The author doesn’t really explain how much of the gap can be explained by this, though probably only because they were just pointing out the possibility that it might not all be sexism. This miiiight have been something the author was trying to address.

Not only that, but what are the objectives of diversity programs like the ones the author mentioned Google creating? Obviously to increase representation, but to what extent? Do they want absolutely equal representation (50–50 for male and female) in the industry? Are they trying to eliminate bias, or enforce equality? How do you measure for success with programs like this? And how much of this can you chalk up to political bias?

II.

While reading this, my model of a leftist blogger/activist cringed somewhat at the points about physiological differences between the sexes. I’ve also seen reactions to this memo such as “women aren’t fit for tech because women”. The phrases “red pill” and “pseudoscience” also came up. This memo probably is more tribalistic than it intended to be, or maybe people thought of it as such. Sure, on average women tend to be more people-oriented and men thing-oriented, but this doesn’t map to all men and all women. People are individuals, they have their own identity and character. This memo (and I can’t believe this has to be said) is not saying that women are unfit for tech. The author is all for inclusion, their point being I don’t think they want to force it the way they see Google doing it. There’s a big difference between “women are unfit for tech” and “maybe it’s not all sexism”.

Gizmodo removed “several” hyperlinks contained within the memo when publishing it (I wonder why). One article it might have linked to is from Psychology Today: Why Brilliant Girls Tend to Favor Non-STEM Careers. Another related post from Scott Alexander in Slate Star Codex: Gender Imbalances Are Mostly Not Due To Offensive Attitudes.

Both link to meta-analyses and studies that find a large difference between men and women on a people-things axis. Specifically, men preferred working with things, and women prefer working with people, with d=0.93 and 1.18 respectively. For those who don’t know, Cohen’s d is a measure of the difference in two means, usually residing between 0 and 1. From SSC:

All I can say is that I spent several years believing that the d statistic was a scale from 0 to 1, because I’d never seen a d go outside that range before.

And from Psych Today:

But the effect size was d= .93, and even if you are not familiar with effect sizes, this would make it one of the largest effects in social psychology; it is gigantic.

Basically, those are significant effect sizes for studies like this. But a wonderful thing that Scott does is not completely map systematizing-types and empathizing-types to men and women. That’s where I think a lot of controversy surrounding things like this is generated from. Going from “women on average are empathizer-types” to seeing a woman and thinking “she’s probably more fit for people-oriented tasks”. If someone applies for a job in STEM, chances are they’re probably a good fit for a job in STEM, regardless of the average placement of their gender on the people-things axis.

This is somewhat of a problem I see in trying to be both an individualist and confronting studies that deal with people as groups. If feminism is about promoting choice, then, having seen these studies, how do you update your beliefs accordingly while still treating people primarily as individuals? I don’t know if I have a good answer other than this: you can let those studies shift your beliefs, but you have to let the fact that people doing their jobs are probably good fits shift them too.

Slight tangent but not unrelated: When controlling for several factors pertaining to the ~20% gap between men and women, a bit over half of the gap disappears, leaving about a 8% disparity in pay between the two (see page 20). (Counterpoint: Some controls reveal how the gap works. Countercounterpoint: Shouting ‘sexism’ doesn’t lead to good solutions.)

This doesn’t perfectly map to representation of women in tech, but it’s related enough to say that the disparity is probably not all a result of choices that women make, but it’s also probably not all sexism either.

III.

That isn’t to say sexism doesn’t exist.

In February, Susan Fowler published a post detailing her experience at Uber. Spoiler: she encountered a bunch of sexist harassment and HR refused to do much about it. It’s probably the most pivotal post dealing with sexism in the tech industry. It prompted Uber’s CEO Travis Kalanick ultimately to resign in June. There’s definitely a culture in the tech industry, and it appears to be hostile to women trying to join it.

The reaction to the Googler’s memo has been mostly negative. Some think it's yet another affirmation of how the tech industry is rotten to its core. One really telling quote from that Atlantic piece:

Women used to avoid computer science because they didn’t know what it is. Now they avoid it because they know exactly what it is.

This is really unfortunate. I would love for everyone who is good at STEM (including women and conservatives) to feel welcome and like they can be themselves in a STEM career. If that involves getting rid of aspects of the industry’s culture that prove hostile to certain groups of people, so be it.

Yonatan Zunger, a former senior Googler, writes on Medium about this manifesto. Right off the bat, he starts off by strawmanning the article:

You have probably heard about the manifesto a Googler (not someone senior) published internally about, essentially, how women and men are intrinsically different and we should stop trying to make it possible for women to be engineers, it’s just not worth it.

What? If that’s the message you got from this memo, then I think you’re injecting a lot more information into it than you’re getting out. The author of this memo did not say we should stop trying to make it possible for women to be engineers. Their point was that we should reduce the gender gap in non-discriminatory ways.

He goes on to make three main points:

(1) Despite speaking very authoritatively, the author does not appear to understand gender.

(2) Perhaps more interestingly, the author does not appear to understand engineering.

(3) And most seriously, the author does not appear to understand the consequences of what he wrote, either for others or himself.

Zunger kind of sidesteps (1) by saying he isn’t an expert so he’s not going to go into it, which kind of invalidates that point. I think (2) is as constructive a point as we’re going to get here, and (3) is just bullying by saying “if I were your superior I’d fire you for this stupid memo.” I would not want to have a boss like that, personally.

In (2), Zunger’s point is that you can’t possibly avoid working with people in engineering:

Essentially, engineering is all about cooperation, collaboration, and empathy for both your colleagues and your customers. If someone told you that engineering was a field where you could get away with not dealing with people or feelings, then I’m very sorry to tell you that you have been lied to.

I would argue that you could replace “engineering” with just about any field and produce just as correct a statement. But maybe that’s his point. Engineering, like all fields, requires empathy. The memo does mention de-emphasizing empathy, but only because relying on it causes us to focus too much on anecdotal information and favor those similar to us. Obviously you won’t survive much if you don’t work with others, that’s true of any profession. But the technical side matters too.

IV.

So what would a good solution to all this be? Well, what exactly are we trying to optimize for? I’d argue we’re optimizing not for diversity, but for freedom. We as a society (I hope) want people to be whoever they want to be, something something the American dream. I don’t have a good solution, and I fully recognize that we aren’t quite there yet in terms of equality of opportunity, but I do have some closing thoughts. If any of these seem obvious to you, that’s probably a good sign.

  1. Sexism and harassment are bad. They don’t belong in the workforce. We have to get rid of that aspect of the culture in the tech industry. In retrospect, this is probably what feminists have been saying all along.
  2. It’s really hard to measure for success in diversity programs, and I don’t think comparing # of men to women is going to yield an accurate measure of that, unless the objective all along was to enforce a 50–50 divide.
  3. Women, on average, tend to be empathizing-humanities-types, and men, on average, tend to be more systematizing-utilitarian-types. But this doesn’t mean all women or all men fit in these groups. If we’re looking to optimize freedom, we need to maximize opportunity.
  4. I agree with some points of this memo and disagree with others. Diversity of viewpoints should be valued, because that actually gets people places. But diversity of gender/race/etc (demographics? identity?) is important insofar as everyone is given an equal opportunity to work in tech. If the group of people who aren’t up for it solely because of the nature of the work happen to be women, then we can either a) find a position that would be a good fit for them, b) introduce new concepts in an attempt to be more inclusive (“pair programming” was mentioned in the memo) or c) find a different career. Beyond equalizing opportunity, we’d be optimizing against freedom, we’d be losing.

V.

Vice reports that the internal reaction to this memo “show[s] anti-diversity views have support” (yay, more sensationalizing). One employee divides the reaction into three groups:

“From what I’ve seen it’s been a mix of women saying, ‘This is terrible and it’s been distracting me from my work and it shouldn’t be allowed;’ Men and women saying ‘this is horrible but we need to let him have a voice;’ and men saying ‘This is so brave, I agree,’” the employee said.

A good amount of the reaction on the internet agrees with the first group, but I’ve also seen what I’d make to be a fourth group: “The author makes some good points, slips up on a few more, but we should let him speak.” Clearly, if the reaction is so mixed, just shoving it under the rug by firing the author doesn’t solve anything. It’ll just drive the sentiment further underground. It would also ironically prove the memo’s point that there is “an ideological echo chamber where some ideas are too sacred to be honestly discussed,” and that “the lack of discussion fosters the most extreme and authoritarian elements of this ideology.”

I sincerely hope one of the largest tech companies on Earth does not fire this person because of criticism leveled at their programs. I can understand why, under various outlets’ misinterpretations, this could have been considered harmful, but I do hope that the memo prompts a good discussion about the intended and actual effects of those programs.

From the SSC article I linked to in section II:

So yes, let’s try to build a better world. But let’s do it because it’s a good thing to do, not because we expect it to single-handedly normalize gender ratios.


Update: Google has fired the author of the internal memo. If ever there were any closet conservatives in Silicon Valley, their doors just got a bit tighter.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade