From a logical point of view …

by Daniel on August 11, 2017

I have now read that “google manifesto”. I read it more out of a desire to forestall people saying “but have you ACTUALLY READ IT?” than out of any expectation that it would contain new or unfamiliar information, and indeed it was your fairly standard evo-psych “just asking questions”, genus differences-in-tails-of-distributions. It’s a mulberry bush that was already pretty well circumnavigated when Larry Summers was still President of Harvard. But what really struck me was that I have changed in my old age; I used to be depressed at the generally very poor level of statistical education, now I’m depressed at the extent to which people with an excellent education in statistics still don’t really understand anything about the subject. I’m beginning to think that mathematical training in many cases is actually damaging; simple and robust metrics, usually drawn from the early days of industrial quality control, are what people need to understand. Let’s talk about distributions of programming ability.

My contention is that the true underlying distributions of computer programming ability for men, women, liberals, conservatives and any other demographic slices of the population are a) more or less totally unknown, and b) not worth the time and effort to estimate with any precision at all, because c) they are totally irrelevant to the practical questions which anyone interested in them might actually want to solve. This is true whether we’re interested in “get the best engineers for Google” or whether we’re interested in “get fair representation for minority groups and women in the workplace”.

The true underlying distributions would be useful if Google’s hiring process was to select people at random from the population, put them through a standard test of the single “quality” variable of interest, then take the ones who passed the test and discard the ones who failed. As a description of how recruitment processes don’t work, this is pretty spot on. Google (like any other company – I first started making this argument in the 1990s when McKinsey were publishing their incredibly influential, amazingly wrong and massively destructive “War For Talent” series) fills jobs by advertising for vacancies or encouraging through word of mouth and recruiters, using interview questions and tests which might have unknown biases, and recruiting people for their suitability for the roles currently vacant (which is not the same thing as “quality” because companies change all the time but keep the same employees. Each one of these stages is enough of a departure from the random sampling model to mean that the population distributions are not relevant.

The male/female ratio at Google is not the outcome of a neutral process; it’s a variable under Google’s control. And when you think of the male/female ratio as an input rather than an output, you can start thinking about recruitment as a quality control process and everything becomes much simpler.

Think of recruitment as a production process. It’s aim is to produce an output of employees of adequate quality, and its failure mode is to recruit inadequate employees[1]. Under very general assumptions, we can say that if the recruitment process is “fair” with respect to male versus female employees (or any other groups), the defect rate in the groups will be identical. It’s an argument similar to marginal cost optimisation.

To translate this into practical language, if the Google Manifesto was correct, then you would expect to see that Google was full of mediocre female employees, who had been hired by a process biased in their favour despite being inadequate to the task. Whatever the author of the manifesto thinks, Google does not believe this to be the case and as far as I can tell from industry blogs, it isn’t – female employees in tech are generally very good. This would, of course, be consistent with the hypothesis that the current selection process is biased against them.

I’d note that this argument could also be extended to one of the author’s other concerns about “ideological diversity” (mentioned in the context of Google, but most usually seen in discussions of university professors). If there were a genuine problem with a biased recruitment process, you would expect to see that the small minority of conservative professors were startlingly good and universally recognised as being so intelligent and productive of the best scholarship that they had got through the discriminatory process. One might call it the “Jackie Robinson Effect”.

If, on the other hand, one had a situation where the writers of windy conservative manifestoes about not getting fair treatment were in fact mediocre whiners who inflated their CVs, then that would be evidence that there wasn’t a bias in the recruitment and retention system, and that in fact there was probably an inefficiency caused by the extent to which mediocrities were able to bump along because their face fitted in a homogeneous techbro culture. The concentration on star engineers, senior executives and Sheryl Sandberg C-Suite geniuses is entirely wrong; the progress of gender equality in the workplace ought to be measured by the extent to which women can get into the ranks of time-serving dead-wood middle management roles.

Or, as the famous calypso had it …

[1] No, don’t make that ‘splainy comment about the Pareto distribution of engineers and how the best are ten times more productive. If it is important to recruit stars, then a non-star employee is a defect from the point of view of the star-recruiting process. That’s the great thing about robust methods, they’re robust.

{ 162 comments }

1

marcel proust 08.11.17 at 4:59 pm

On the veldt, all (surviving) programmers were mediocre because the star programmers couldn’t outrun the sabre-toothed tigers.

Welcome back Daniel. If I may speak for my fellow bent timbers, we have missed you.

2

Donald A. Coffin 08.11.17 at 5:32 pm

Nicely done, DD, and worthy of exposure in the wider world (although I would guess a number of readers around here will be sharing it).

3

CaptFamous 08.11.17 at 5:32 pm

I see this as a great example of the inherent subjectivity of statistics (and analysis in general) that often gets ignored (“numbers mean it’s real!”). Every analysis starts with a definition of your variables, scope and parameters. The author of the manifesto refuses to take responsibility for making any of these definitions; it’s like constructing a logic proof and refusing to acknowledge that you’ve made any assumptions.

4

Adam Hammond 08.11.17 at 5:40 pm

I wish the OP had been longer. (I am being honest, not ironic.)

5

orangeman 08.11.17 at 5:44 pm

If, after 50 or 100 years of teaching statistics, most mathematically inclinded people don’t get it , is that because
a) statistics is actually a *really* difficult subject
b) statistics teachers suck
c) some combo of A+B

I mean, people who teach stat have had almost 100 years to get their act together; if students are still misusing basic things like p values, that goes to the teachers , not the students

6

mojrim 08.11.17 at 5:48 pm

I hate starting with disclaimers, but this subject is contentious enough to demand one. I’m not in tech, I have zero interest in tech beyond the user level, and the hiring practises of the industry have no effect on me or anyone close to me. What I see is an interesting argument on the nature of sex-linked traits and neurodiversity, in which you seem to have made a logical error.

That is, if the median skill levels of men and women in the industry are similar, but women enter the field at 1/3 the rate of men, doesn’t that imply self selection more than anything else? Doesn’t it further imply that 1/3 as many women have the requisite combination of talent and interest to attain the requisite skill level? Or is there a metric I’m missing showing the median skill of women to be higher than that of men in the industry?

7

Cervantes 08.11.17 at 5:48 pm

Funny story. Almost all the physicians used to be men. Now half the medical school graduates are women. I guess the job description suddenly changed . . .

8

MrArt 08.11.17 at 6:08 pm

It’s a valid point and one that doesn’t seem to be made elsewhere, so thanks (and also welcome back!)

On measuring the ability of programmers and then comparing men and women: imagine for a moment Google did that, and it showed the women they’d hired were on average worse. Would that be taken, as you suggest, as evidence of biased recruitment in favour of women?

Also, it’s not necessarily distribution of ability that’s under debate, so much as distribution of inclination to work in software in the first place. That’s more easily measured and pretty obviously “favours” men.

9

Ian Maitland 08.11.17 at 6:27 pm

Wow! It is telling that, right off the bat, you are at such pains to defend your PC bona fides. But, of course, you would not have been caught dead reading this smut in the toilet but for the need to forestall people from saying “but have you ACTUALLY READ IT?” Rather ironic I think when Google has thrown intellectual diversity under the bus of gender diversity. Who says that intellectual terrorism isn’t alive and well?

I notice that you circumnavigate the mulberry bush of sex differences very gingerly. What I don’t find is any statement of your own view of the role of biology in shaping preferences, say, for math and computer science.

I think candor was called for because what drove James Damore’s professional suicide by memo was a “diversity” session (doubtless one of an unending series) in which he was instructed regarding what he could and couldn’t say. It is a premise of diversity policies that every disparity between men and women in Google’s workforce is due to discrimination/ harassment by men and their “techbro culture.” As Damore put it, “all disparities in representation are due to oppression.”

Now, if you are an earnest young man who favors diversity and inclusion (like Damore), it must be pretty demoralizing to be forced to submit to endless brainwashing or “re-education” sessions in which male employees are the designated bad guys and are forced to denounce themselves for their sexism, what Damore calls Google’s culture of shaming. Damore’s memo was intended to spark a debate over whether in part the “underrepresentation” of women might be due to “differences in distributions of traits between men and women,” rather than oppression by males. (It is hard for a tech company to achieve gender balance in its workforce when less than 18% of computer science graduates are women).

What makes Google’s behavior so contemptible is that, not only did it misrepresent what Damore said, but out of political expedience (I suspect) it has chosen not to challenge the charge that all gender disparities are due to discrimination. To have defended itself would only have invited more media abuse. But, by failing to defend itself, Google has failed to defend the honor of its male employees.

No matter what your views of the relative importance of nature and nurture in causing observed sex differences, the fact remains that the case for diversity is a fraud. It has a logical flaw at its heart. Either women are the same as men, in which case hiring more women won’t increase diversity. Or women are different from men, in which case it is no surprise that they make different career choices.

Diversity is a cloak for progressives’ agenda of imposing gender equality on every “male-dominated” occupation or profession without any regard to what actual men and women want. It would be refreshing if they admitted that truth.

10

Belle Waring 08.11.17 at 6:38 pm

MrArt: do you think the inclination to work in software is an immutable variable of some kind or might it be affected in some impossible-to-imagine fashion with the treatment one might get while working in software? I mean, what if Google polled your co-workers and superiors and it showed many of them believed that you were congenitally bad at your job because you had a vagina? Would that be taken as evidence that they were going to treat you like shit a lot of the time? I’m just asking questions.

11

nick s 08.11.17 at 6:47 pm

I’m depressed that this shite gets reheated every few years.

There’s a parallel process whereby a field or profession suffers a loss of relative prestige (and earning power) as women and minorities enter it, even though they often have better credentials and experience than the older generation because board licences now demand it. (Similarly, consider the position of nurse practitioners and physician associates as “not quite doctors.”)

12

PatinIowa 08.11.17 at 7:19 pm

One thing that’s manifestly clear: the writer of the memo has a set of biases that would make him very unlikely to evaluate other employees–especially women–fairly.

I don’t know that I would have fired him. I do know that I would have made clear to him that his future in my employment would not, under any circumstances, involve supervising other employees, and his compensation, references, and evaluations would reflect that.

13

deadpan 08.11.17 at 7:32 pm

The explanation for the gender disparity in tech proposed by the ‘manifesto’ was dimorphic preferences, not ability. ACTUALLY READ IT again perhaps?

14

parse 08.11.17 at 7:52 pm

Thanks for an insightful analysis. As someone deficient in statistical understanding, I learned a lot.

It seems like the author of the manifesto made a number of different claims:

1) An analysis of the reason for a lack of gender diversity at Google/among programmers and the effectiveness of affirmative action to change that

2) The assertion that the public expression of those views at Google was censored and would lead to ostracism or more dire consequences

3) The belief that the lack of a viable forum for those views within Google hurt the company

Given the way events played out, it looks as though #2 is a valid claim, probably because #1 and #3 are grossly invalid

15

dsquared 08.11.17 at 7:53 pm

imagine for a moment Google did that, and it showed the women they’d hired were on average worse. Would that be taken, as you suggest, as evidence of biased recruitment in favour of women?

Yes that’s what I would say, which is why I did say it. (I’d probably take it as evidence that failure in the outreach process was being foolishly compensated for by introducing biases to the recruitment process). But since that’s not what they have found, I find it hard to give a fuck about this hypothetical possibility. You are clearly much more resilient when it comes to giving such fucks so why not answer me a question of my own – how much human productivity might we have sacrificed over the course of the last 100 years simply because God awful annoying stupid tedious techbros of all sorts would rather go through literally any intellectual gymnastics than imagine that there could ever be such thing as a white dude from the right university who was just plain and simple f****** thick white dude from the right university who was just plain and simple f****** thick?

16

Alex K. 08.11.17 at 8:01 pm

This part is worth repeating:

…the true underlying distributions of computer programming ability for men, women, liberals, conservatives and any other demographic slices of the population are a) more or less totally unknown, and b) not worth the time and effort to estimate with any precision at all, because c) they are totally irrelevant to the practical questions which anyone interested in them might actually want to solve.

The true underlying distributions would be useful if Google’s hiring process was to select people at random from the population, put them through a standard test of the single “quality” variable of interest, then take the ones who passed the test and discard the ones who failed.

Then the reasoning gets a little opaque, I’m afraid. To check for possible discrimination, we should be looking at the male/female ratio both in the pool of applicants and in the pool of new employees hired from the former. As an example, suppose 990 men and 10 women of equally high quality are applying for 100 jobs, and the company’s quality measurement tool has a near-zero failure rate. If the women are all accepted and the men are chosen by the roll of the dice, the ending 90-10 ratio will still look discriminatory against women, although it’s actually discriminatory in their favor. If the initial pool measured 800-200 and the ending pool, 90-10 again, it would be evidence of anti-female discrimination.

This said, the manifesto’s author wasn’t fired for misunderstanding statistics.

17

Steffen 08.11.17 at 8:07 pm

“…if the Google Manifesto was correct, then you would expect to see that Google was full of mediocre female employees, who had been hired by a process biased in their favour despite being inadequate to the task.”

No, it would only explain “why we don’t have 50% representation of women in tech and leadership” as Damore states. It’s mainly about preferences not about ability.

18

BruceJ 08.11.17 at 8:13 pm

I used to be depressed at the generally very poor level of statistical education, now I’m depressed at the extent to which people with an excellent education in statistics still don’t really understand anything about the subject.

I rather suspect it’s a variant on Upton Sinclair’s famous maxim:“It is difficult to get a man to understand something, when his salary depends upon his not understanding it!”

19

Crissie Brown 08.11.17 at 8:13 pm

To keep the numbers simple, let’s say that the average gender percentages nationwide in university computer science programs (roughly 80:20 m/f) follow through to Google applicants.

A study of GitHub “pull requests” found that code written by women is approved slightly more likely than code written by men, but only if the female author’s profile hides her gender. If she reveals her gender, her code is less likely to be accepted. So it’s plausible that women are, on average, slightly better coders than men. (Note: the difference in acceptance rates, when female authors’ gender was hidden, was very small.)

Google gets roughly 130 applications for each job opening. Again, to keep it simple, let’s say that 130:1 ratio applies across entry-level job openings, including coders. And let’s say that only the top 20% of applicants are good enough to meet Google’s standards.

Given those numbers, for any entry-level coding job Google would typically have 26 qualified applicants: 20 men and 6 women. That’s the 80:20 ratio referenced above, with the ‘extra’ qualified applicant a woman, based on the GitHub study results.

If Google hire a qualified man from that pool of 26, they will turn away 19 qualified men and 6 qualified women. If they hire a qualified woman from that pool, they will turn away 20 qualified men and 5 qualified women. Either way, they will turn away about two-dozen qualified applicants, including both men and women.

If Google decide their code shop should more closely resemble the demographics of their market — roughly 50:50 m/f — they have plenty of qualified women in the pool for a typical coding job.

What’s more, if they have 10 coding job openings and hire the best 5 men and the best 5 women (as best their HR procedures can determine that) then they’ll turn away about 190 qualified men and 55 qualified women.

That works out to a qualified-but-turned-away ratio of 77.75:22.25 m/f, which is still very close to the ratio in the qualified applicant pool (80:20). Indeed you could justify the ‘extra’ 2.25% for women by the small difference found in the GitHub study.

And here’s the key point: not a single qualified male will have been turned away to hire an unqualified female.

Yes, there’s an exception: HR mistakes. No HR hiring process is perfect, and invariably at least some new hires don’t work out. But if the HR assessment of qualification is truly gender-neutral, then the washout rates for men and women should be about equal … as David predicts in his post.

If the washout rate is higher for men, it’s possible (but not proven) that Google’s HR process overestimates men’s qualifications. If the washout rate is higher for women, the converse is true. But so long as the washout rates are about the same, it’s likely their HR are assessing and hiring applicants well …

… and that remains true even if Google hire 50:50 from an 80:20 pool, because they have so many applicants for every position.

As for Google’s (hypothetical) decision that their coding shop should more closely resemble the demographics of their market — because they think that will help produce more products that meet the wants/needs of more of their market — in employment law that’s known as a “legitimate business reason.”

20

Dave 08.11.17 at 8:21 pm

The true underlying distributions would be useful if Google’s hiring process was to select people at random from the population, put them through a standard test of the single “quality” variable of interest, then take the ones who passed the test and discard the ones who failed.

Dan, you know that hiring processes do not select people at random. Is this true in the City? In the Premier League?

I did a Psychology degree (I know, not a proper science: tell me about it) in the 90s. Then, the trend was “we can do away with interviews”. Me (always contrarian) “the whole bloody point of interviews, you self-serving imbeciles [I wasn’t popular], is to determine where the candidate WANTS to go. Her CV already laid out her abilities.

Google tech employees aren’t a random subset of US population; they’re the ones motivated to work on certain engineering problems. And that’s where you’re wrong. Women [as a population] have different motivations than men [as a… etc]. That’s your difference.

If we’re doing videos. How about?

This? Horrendously sexist, imagine thinking women are different from men. Oh noes!

21

MrArt 08.11.17 at 8:29 pm

Belle@7

It would be pretty annoying I agree. (Though I’ve never seen that attitude expressed by my colleagues. Not overtly anyway. I work in software, but in the UK)

Tech companies can and should contribute to closing the “inclination gap” by treating their women better – but it’s a variable that’s largely (IMHO) outside their control. You can see the gap well before software companies are involved, e.g. in the rates at which girls and women choose to study computers.

I only make the point because the industry seem to be routinely attacked for having an unequal gender ratio, and individual companies should be able to use the defence of having an unequal ratio of applicants. (This is certainly the case where I work: 90% of applicants I’ve seen in the last 6 months were male).

22

P O'Neill 08.11.17 at 8:38 pm

Statistics is a discipline with a core tool, regression, whose name reflects a misunderstanding that occurred at the same time it was developed, so this has been going on for a long time. And the Google business model involves very large economic rents. Very difficult to argue that the techbro quotient is disciplined by “the market.”

23

Quite Likely 08.11.17 at 8:49 pm

The best explanation I’ve heard for the gender gap in programming is that it’s not so much about ability as it is about interest. A lot more men get into computer related stuff from an early age, and thus a lot more come out the other end of the pipeline into the job market. So there’s no reason to expect a very different average level of skill between male and female programmers, but there’s also no reason to expect there to be proportionately more female programmers than there are female computer science majors, which are overwhelmingly male.

24

Omega Centauri 08.11.17 at 9:01 pm

I would agree that we don’t even know how to measure programming ability. I would also say, even is we had some sort of programming IQ test, it would miss the fact that there are ia a wide collection of skills/aptitudes/personalities and individuals with the same “IQ” score will do well/poorly is so small part because of a good match of skill/interest set to job demand. For some the ability to understand the hardware and how to interface and exploit it will be crucial. For others enough math chops to come up with innovative new algorithms -or know enough existing ones to quickly settle on a good one. For some jobs, leaps of logic will be a key asset -for others a great detriment. Who is fast and bold? Who is slow but meticulous? One size doesn’t fit all by any means.

On to Google and some of the valley firms nashing their teeth over the poor representation of women (or blacks and hispanics say) in their technical work places. Shouldn’t it be difficult, and frustrating, and potentially non-optimal to try to get an employee distribution that differs substantially from that of the recruitment pool? In the extreme case, it might mean a bidding wars for the few qualified members of minority X, as every firm considers they absolutely have to have one at any price, and there aren’t enough to go around.

In any case, where I work, its high end mechanical engineers plus software. Until recently women were like non-existent. But, we hire a lot from China and France, and apparently the recruiting pools there have more than a few excellent female candidates. Thats a data point that
supports the argument that the selection is cultural rather than due to intrinsic biologic differences. But, IMHO, by the time someone enters the recruitment pool (or not), the damage has already been done. I don’t think that in more than rare cases, many are going to be able to make up for a lost decade simply because of raw ability. So to a fair degree, these companies are going to be stuck with the recruitment pools, that come out of the education system.

25

bianca steele 08.11.17 at 9:04 pm

MrArt’s suggestion might equally well prove that Daniel’s advice was being followed, that they were restricting recruitment of stars to white men and were recruiting women and racial minorities deliberately into support positions.

(I haven’t read the thing, nor any piece about it longer than this one, nor do I plan to. I will say that the idea of the “supercoder” was well known in my youth and I have no desire to see how many otherwise sympathetic people who couldn’t place a semicolon correctly to save their lives simply assume they are all men.)

26

CJColucci 08.11.17 at 9:31 pm

I’ve spent decades litigating employment discrimination cases and have long thought that where discrimination has its bite is largely at the level of the mediocre. The really good people are, mostly, getting hired despite being [fill in the blanks] because they are needed or it would be hard to disguise a discriminatory reason for not hiring them; the bad people, whether [fill in the blanks] or not, aren’t seriously considered because they’re bad; and at the level of the meh prospects, there is wiggle room to discriminate. To analogize to a current situation: if Colin Kaepernick were a better quarterback, he’d be playing somewhere; if he were worse, no one would question why he isn’t playing; but he’s just mediocre enough that non-football reasons look like the best explanation.

27

Cugel the Unclever 08.11.17 at 9:34 pm

Most of the academic work that James Damore gestures at[1] is about differences in *personality* and *preferences* as these things are measured in social science, rather than ability.

Damore keeps making references to “ability” as well, which is unfortunate, as his substantive claim is as follows:

Differences in distributions of traits between men and women may in part explain why we don’t have 50% representation of women in tech and leadership. Discrimination to reach equal representation is unfair, divisive, and bad for business.

He then blunders into this statement:

I’m simply stating that the distribution of preferences and abilities of men and women differ in part due to biological causes

This is a facepalm-moment, because he mentions “abilities” then spends most of the rest of the piece talking about *personality* differences, interest in things vs. people, neuroticism etc. He claims there is a difference in skill traits, which is not what most of the evidence he gestures at actually indicates. So the memo is clearly sexist, and definitely worth a stiff talking-to from HR and an early warning.

But it does bring me back to the OP above. I’m genuinely curious if Dan thinks the entire gender disparity in “tech” or Google is attributable to sexist hiring policies. Dan says:

The male/female ratio at Google is not the outcome of a neutral process; it’s a variable under Google’s control. And when you think of the male/female ratio as an input rather than an output, you can start thinking about recruitment as a quality control process and everything becomes much simpler.

But is it though? That’s the key point in question. If women are on average more interested in people and men are on average more interested in things you might simply expect more men to apply to work on jobs that require dealing with things than women[2]. At this point you’re no longer hiring from a random sample of the population, but from a biased sample. In the US, in 2014, 18% of computer science majors were women. So Dan’s claim that the ratio is wholly under Google’s control is just wrong. Unless you want to start talking about exactly what sort of company Google wants to be, which doesn’t really illuminate matters one way or the other.

Dan states:

if the Google Manifesto was correct, then you would expect to see that Google was full of mediocre female employees, who had been hired by a process biased in their favour despite being inadequate to the task

At no point does James Damore make this claim. His argument is that the observed gender disparity is *mostly* a result of preferences which, on average, differ between men and women. He doesn’t discount sexism entirely, but his point is that relentless pursuit of 50%/50% representation is, as he puts it, divisive. If you pick the top 5% of applicants from category M and category W, and you get 82% of applicants from category M, and only 18% from category W, then it’s unlikely your output will be 50%/50% between M and W.

I’m more of a people-person myself, so I may have made some mistakes in the HTML tags in this comment. I’d be grateful if someone who was better at that sort of thing could correct any errors I’ve made.

[1]: Apparently links were redacted from the version of the memo posted at Gizmodo. I’m sure we all agree that lack of explicit citations might have contributed to the general bad temper and lack of clear understanding of the content of the memo.

[2]: The question of whether Google should focus on “things” or “people” is another question entirely. It’s been weird seeing people attack the memo on the basis that software engineering is as much people work as technical work, which seems to accept the memo’s basic claim that there *are* significant gender differences; and as a result hiring more women is the right thing to do because they’re on average more interested in people.

28

Belle Waring 08.11.17 at 9:53 pm

On reflection I prefer dsquared’s response.

29

DCA 08.11.17 at 9:57 pm

Of course this was Silicon Valley where everything was just invented, so something else missed is how moth-eaten this particular topic is. Back in the Early Paleolithic (that is, say the 60’s and 70’s) the fraction of women in computer science was high; then to quote from an article in 2000, “The proportions of Computer Science (CS) graduates who are women has been declining in the last fifteen years. ” Historically, most computers (the human kind) were female, and this carried over into the early computer era. Then it became prestigious and the men showed up.

And the article is “Underrepresentation of Girls and Women in Computer Science: Classification of 1990s Research”, with citations back into the 1980’s to papers about possible causes. The writer may have been a great programmer, but he sure didn’t do his homework.

30

Orange Watch 08.11.17 at 10:05 pm

Crissie Brown@16:
This absolutely nails the issue, at least IRT Google or other high-profile, high-prestige companies. As long as there is a sufficiently large hiring pool for there to be a significant excess of qualified candidates for every job opening, the underlying assumptions fielded against practices such as Google’s are problematic so long as we accept there isn’t some (essentially magical) means of concretely and deterministically grading job suitability in an extremely fine-grained and accurate manner. If there is not a clear advantage to one qualified candidate above another, then some arbitrary criteria will be applied to differentiate between them. Seeking to have demographics similar to that of the population at large is rather hard to construe as discriminatory so long as all other things are equal.

31

peterv 08.11.17 at 10:33 pm

@ orangeman #5:

“If, after 50 or 100 years of teaching statistics, most mathematically inclinded people don’t get it , is that because
a) statistics is actually a *really* difficult subject”

It is. Moreover it relies on probability theory, a mathematical theory whose syntax dates from the decade around 1666, whose standard axiomatization of that syntax dates from the late 1920s, and for which there is, as yet, still no agreed semantics. If people who study the formal representation of uncertainty professionally can’t agree what probability statements mean after 350 years, what hope is there for the rest of us.

In these circumstances, the lack of agreement on the semantics of and choice between alternative approaches to statistical hypothesis testing is, after a mere 100 or so years of disagreement by statisticians, not something we should be especially alarmed about.

32

Omega Centauri 08.11.17 at 10:38 pm

If we for analysis purposes divide the applicant pools into men and women. Surely assuming the average quality of each pool is the same is only a default assumption. But, we already know that there was significant cultural selection going on to form those pools. I doublt the average quality would be close to equal. The most credible arguments say the womens pool mightbe better. I.E. only women who are reaaly confident in their ability will choose to enter the pool, because of the perception of bad cultural vibes. And, two, if firms are really actively trying to close their gender gaps, they could have preferentially removed (by hiring them) proportionately more women than men from the recruitment pools. But, I’m sure you could come up with arguments pointing the other way too. Lacking an objective measurement tool, we just can’t know.

33

Matt 08.11.17 at 11:24 pm

The original author tries to explain the lack of women in tech with biological differences, which aren’t just socially constructed because:

• They’re universal across human cultures

• The underlying traits are highly heritable

• They’re exactly what we would predict from an evolutionary psychology perspective

So we should swiftly reject the “timeless biological differences” explanation for outcomes that do vary significantly by social environment, or that have varied within a single culture too swiftly to be the outcome of genetic changes, right?

In the USA, women’s share of bachelor’s degrees granted in computer science more than doubled from 1970 to 1980, peaked at 37% in 1984/1985, and then fell back into the high teens by 2010.

The change is lightning-quick compared to human generations. This is like looking at a graph of American polio cases and concluding that selection pressure abruptly produced polio-proof offspring starting in the mid-1950s. It’s mutation and selection as explained by X-Men comics.

Why does India have twice the percentage of women among HackerRank participants as Canada? I don’t know. Here’s what I do know: not because of biological factors that are innate to all women! There’s not much variation in sex ratios among working age adults across these countries. There is a large variation in how male-weighted HackerRank participation is by country.

34

Kiwanda 08.11.17 at 11:30 pm

Speaking of “quant” organizations that need more diversity, I see that the senior team at Frontline Analysts is 83% male, and the female senior staff in on the more non-technical end: coaching and communications. I wonder why Frontline is unable to hire, or perhaps to retain, more women on their senior staff.

35

RichardM 08.11.17 at 11:40 pm

Under very general assumptions, we can say that if the recruitment process is “fair” with respect to male versus female employees (or any other groups), the defect rate in the groups will be identical.

a: If you worked for google, would you feel comfortable posting under you own name the opinion that some of the women who worked for them were ‘defective’?

b: google’s recruitment process involves a step where minority candidates (only) who fail one interview can request a second. I presume you understand how this policy, which is new, will very likely lead to a higher proportion of ‘defective’ minority candidates, as there are now two chances to get the wrong result. The claim otherwise, in defiance of all logic in favor of making soothing noises, seems to have been what annoyed the sacked guy initially.

So your position here seems to be a criticism of google’s recruitment strategy. For which see a above.

36

Alex 08.12.17 at 12:06 am

mojrim:

That is, if the median skill levels of men and women in the industry are similar, but women enter the field at 1/3 the rate of men, doesn’t that imply self selection more than anything else?

Nope. First, because Google is not interested in the median – it wants those who are very good programmers – that’s why Daniel talked about the tail of the distribution. Second, you’ve kinda missed the whole point of the post (and what the Google memo writer was trying to argue against) – namely, that sexism in the tech industry (whether explicit or implicit) plus other social processes (e.g. the way we bring girls up versus boys) are the reason for that discrepancy between being 50/50 and only 1/3 or whatever women entering the profession instead. The memo writer thinks it’s (mostly) about inherent biology. He at least tried to argue that, it was a flawed argument but still, yet what you have done is simply point to the discrepancy and go “Ha! This discrepancy SHOWS it must be about self-selection! Logical error!” You haven’t actually eliminated any of the other possibilities anyone else is talking about. Finally, “interest” is (or can be) a socialized variable – simply asserting “women have less interest in tech stuff” might be true, but you need to show why that can’t be explained through a sociological model.

37

Alex 08.12.17 at 12:11 am

Alex K:

Then the reasoning gets a little opaque, I’m afraid. To check for possible discrimination, we should be looking at the male/female ratio both in the pool of applicants and in the pool of new employees hired from the former. As an example, suppose 990 men and 10 women of equally high quality are applying for 100 jobs

No, let’s not suppose that. Instead, let’s suppose the pool of applicants is a variable which Google has at least some control over. That is, it doesn’t just sit there waiting for people to apply by chance, but advertises all over the place, sponsors people, visits university open days and jobs fairs, sets up community coding classes for kids, you name it. Then if only 1% of applicants are women, we can say that Google really needs to work on those outreach programs, and focus more on getting the number of women applying up.

38

dsquared 08.12.17 at 12:16 am

Crissie: you’re right, but if you look at this a bit deeper you’ll see that the reason that you’re right is that the ratio of false negatives has to be the same at the optimum, because whatever the distribution is, the definition of an optimum ensures that if the distribution is at least reasonably well behaved, they’re all locally identical at the optimum.

39

Ian Maitland 08.12.17 at 12:24 am

I submitted a comment many hours ago. It is still “awaiting moderation.” Is there some problem?

40

Collin Street 08.12.17 at 12:33 am

To check for possible discrimination, we should be looking at the male/female ratio both in the pool of applicants and in the pool of new employees hired from the former.

Not so! That’s DD’s point: you can look at it on the other side, failure rates per thousand hires. It’s an unusual way to frame it but it reveals essentially the same information, and does so using only data that the firm keeps reliable track of. “Failure per thousand hires” is a rate; it’s already adjusted for pool sizes and doesn’t need to be adjusted again. It’s adjusted for the output pool rather than the input pool… but if there are no problems then the two sides will match, obviously, and if there is a problem then the mismatch will show up just as well seen from the inside as the outside.

[it doesn’t work for firms where the absolute failure rate is so low as to be too noisy to reveal the information you need; this is a real problem for small firms — but no statistical methods will work for small numbers — and a theoretical problem for firms with such outstanding hiring practices that they never mishire anybody, despite thousands a year, but this latter is a never-happens-in-the-real-world situation]

41

Jose Arcadio Buendia 08.12.17 at 12:41 am

I agree with most of the OP, but the problem is that you can’t really read this on the merits as such. You have to read it on the perceived merits of how the alt-right freakout machine will take it. It was made to sound very reasonable, using all of the so-called SJW language and references to science and yet elicit the very reaction that it brought on: termination, thus giving them the martyr that this person seemingly wanted to be.

The key thing that they will take away from this is that you can’t even acknowledge that differences between men and women—constructed or not—exist. Is that what it says? Is that what it means? No, but that’s the text for the alt-right reader. And of course that straw man text makes it absurd on its face to argue with.

You could post forever about how and why the memo is wrong, but you’re wasting your time.

42

nick s 08.12.17 at 12:44 am

“A lot more men get into computer related stuff from an early age, and thus a lot more come out the other end of the pipeline into the job market.”

Get thee to YouTube to see the “for fathers and sons” advertising for American home computers in the 1980s, around the same time that comp-sci undergraduate major numbers start to drop for women. The “pipeline” was deliberately narrowed, and while today’s brogrammers aren’t responsible for that, they are responsible for any attempts to retcon that history, especially if their first move after being shitcanned is to seek martyr status in drearily familiar places.

Or, as Anil Dash noted re “pipeline” arguments a couple of years ago: “We can’t keep saying it’s reasonable to have robot cars & face computers in 3 years, but empowering women/PoC is going to take 20 years.”

43

Sebastian H 08.12.17 at 12:46 am

I fully expect that women are treated poorly in programming, largely because women are treated poorly all over the place. I have no idea if it is worse in programming than in say doctoring or lawyering or investment banking–all places where I have heard some nasty anecdotes. It seems perfectly plausible that women may be discriminated against, and legally, any time we see big differences in hiring to population ratios, there is a huge potential for legal liability.

Your discussion has some holes though.

“Under very general assumptions, we can say that if the recruitment process is “fair” with respect to male versus female employees (or any other groups), the defect rate in the groups will be identical. “

Yes, if there is a nice neutral method for evaluating employees, this would be true. But in actual fact, you know what gets much more legal scrutiny for bias against protected classes than hiring decisions? Firing decisions. If you tried to fire a protected class in proportions larger than the current already hired population you would be begging for a lawsuit without absolutely ironclad evidence that the members of that population were ‘defects’ under your definition. You see it all the time in age discrimination lawsuits–an employer is taking a somewhat risky path if they don’t hire some people 40 and older. But if they disproportionately fire people 40+, you can tell them to start writing the settlement check now. You’re assuming an approximate symmetry in hiring and firing that absolutely does not exist.

Now vaguely looking at the numbers you can see that the hiring ratios are pretty much in line with interest numbers and graduation rates in the field. That would tend to suggest that women are either genuinely less interested in the field, or that they are pushed away from it much earlier than the hiring level. Or very likely both.

” If there were a genuine problem with a biased recruitment process, you would expect to see that the small minority of conservative professors were startlingly good and universally recognised as being so intelligent and productive of the best scholarship that they had got through the discriminatory process.”

Again this presumes that “productive of the best scholarship” was subject to a neutral evaluation, while in many of the less scientific fields the politics of the field certainly could contribute to the citation levels or lack thereof. And shouldn’t this argument be reflexive? If women were really being discriminated against shouldn’t the small minority of female programmers be universally recognised as being the very best programmers in the world?

I tend to think that surviving in a field with heavy discrimination doesn’t always suggest that you’re the best of the people who would have made it had there not been heavy discrimination. Surviving in a field with heavy discrimination requires ancillary skills and talents like stubbornness, hard headed refusal to be pushed down, or in some cases hard headed refusal to notice it.

44

Kiwanda 08.12.17 at 1:52 am

The philosophy department of the National University of Singapore has 85% male faculty; the economics department of the University of Queensland has 78% male faculty; the philosophy department at the University of Wisconsin-Madison is 75% male; philosophy at the Univ. Bristol, 72%; International Relations group within the Poli Sci Dept. at George Washington Univ, 80%.

Why are these institutions, like Frontline Analysts, unable to hire, or perhaps retain, more women? I think some investigation is needed regarding their likely biased hiring practices, their hostile work environments, and the splainy dudebros that must work at these places.

45

tib 08.12.17 at 2:22 am

The argument that women just aren’t interested in a highly compensated field like software development is hilarious.

46

baa 08.12.17 at 2:37 am

if the Google Manifesto was correct, then you would expect to see that Google was full of mediocre female employees, who had been hired by a process biased in their favour despite being inadequate to the task.

This has it exactly backwards. Google *does not have* 50/50 representation in tech positions. That is the fact to be explained. What the GM suggests is, and I quote, “differences in distributions of traits between men and women may in part explain why we don’t have 50% representation of women in tech and leadership.”

If Google discriminates against women in the positions where women are <50%, we would expect the average woman to be superior to the average man (the Jackie Robinson effect). If, instead, the Google Manifesto is correct, you would expect the performance of men and women at Google to be relatively equal in the positions where women are <50%.

47

Anonymous 08.12.17 at 3:21 am

When you speak of evaluating the hiring process according to the defection rate you’re assuming that the process which weeds out employees is unbiased.

In an environment where you can be summarily fired for even thoughtfully and respectfully questioning the value or method of diversity efforts, that would seem like a strong assumption to make.

I very much enjoyed reading your analysis minus the attack on conservatives at the end.

48

nastywoman 08.12.17 at 4:19 am

I also always wondered why so many man babies got dressed in blue from birth on and we women get dressed in pink?

49

Krzys 08.12.17 at 4:27 am

Your argument is that a functional of two input distributions is unaffected by their locations? Are you sure YOU understand statistics?

50

Chris Mealy 08.12.17 at 6:36 am

I’m just a shitty old hacker and could never get hired at Google, but maybe a non-super-genius like me can look up some numbers. I wonder if women have innate interests and talents in other competitive fields? Let’s see what the old Statistical Abstract says. Percentage of female graduates by year:

Medical degrees
1970 8%
1980 23%
1990 34%
2000 43%
2009 49%

Law degrees
1970 5%
1980 30%
1990 42%
2000 46%
2009 46%

I only took one stat class in college so I don’t know what you call it when you make a long, complicated argument without eyeballing real-world some numbers first.

51

Lobachevsky 08.12.17 at 7:28 am

MrArt @8: If you want to measure fairness, comparing the averages of the tails of two distributions isn’t very helpful. What counts is the cutoff point, which arguably should be the same in a fair process. After that, let the averages fall where they will.

Or more crudely: if Google recruit the next Turing, they shouldn’t have to recruit a dozen halfwits of the same gender to bring the averages back into line.

52

casmilus 08.12.17 at 9:07 am

When I got my first job in software development 20 years ago the firm I worked for had quite a few middle-aged workers who had moved into programming after other careers. These people were no worse than any of the recent graduates.

From this entirely unrepresentative sampling I concluded that there is no truth in the saying “can’t teach an old dog new tricks”, the real reason old people are resistant to change is just the well-founded ones that (1) it’ll be unnecessary/ doesn’t improve things, (2) be used as an excuse to get rid of old workers who could actually retrain and be useful, and (3) older people have lots of other responsibilities that cause them to be cautious and conservative in general. All bloody obvious really.

I say the real battle line soon will be over age discrimination, as we all have to stay in work till we’re 85.

53

casmilus 08.12.17 at 9:10 am

@21

The correct place for the semicolon is at the end of each statement in C++.

54

BigHank53 08.12.17 at 9:48 am

on why Google-bro is both (a) not very smart, and (b) not employed by Google any more. It’s doesn’t waste any time on gender; it’s about engineering and management.

I am sure that everyone is here is aware that most orchestras have switched to blind auditions, where the judges listen to potential musicians from behind a curtain. Hiring of female musicians rose significantly after this was introduced. It’s not just something that happens to professionals–some sociologists gave sample elementary school math worksheets to teachers for grading. Worksheets with female names got lower marks than ones with male names, or worksheets with the names redacted. It is facile and self-serving to claim a lack of female interest in a technical field when there’s been active social pressure steering them away from it for their entire lives.

55

SusanC 08.12.17 at 10:38 am

I think I basically agree with Daniel here.

Semi-seriously, I’m wondering if a better model might be something like the Fermi Paradox, with “why are there so few women working for Google” substituting for “Why haven’t we encountered extraterrestial life yet”, and something a bit like the Drake equation having terms for the various reasons people do not work for Google.

Just an anecdote, but a woman I know who does work for Google says the reason she doesn’t work for Microsoft is the way woman get treated at Microsoft. A proper social science study might take this a research hypothesis and see if the effect is more general than just that person. Hypothesis: (a) “Potential candidates don’t want to work with those sexist dillweeds” (hat tip to Belle here) is a significant term in the analogue of the Drake equation; (b) certain other employers in the same industry have a reputation as noticeably worse offenders.

[Of course, there are plenty of other reasons why you wouldn’t want to work for MS apart from the sexism factor … ]

I am slightly puzzled by the hiring practises of some of these tech companies…

a) The story they tell the immigration authorities when applying for visas (in both the US and EU) is that there are absolutely no qualified applicants for the post with the right to work in the relevant country. Zero qualified applicants. Not even one, of either gender.

Now of course you, or some suspicious minded person from the relevant government’s immigration department, might doubt the truth of this claim.

But, on the other hand, the salaries those companies are willing to pay suggests there may be some truth in it. (i.e. “offer more money to attract more applicants” has definitely been tried, and might well be into diminishing returns by now)

b) On the other hand, the stupid version of stack ranking, as practised by e,g, Microsoft, is not entirely consistent with (a) and is more consistent with an oversupply of candidates.

How not to do stack ranking: You have (say) 1000 applicants. You hire the top ten. You tell the immigration authorities the other 990 were not qualified for the post (so you can hire the ones who needed a visa). Each year, you sort your employees and fire the bottom 10%, even though you know they are actually competent — even really good — at the post they’re in, and are better than any of the 990 you didn’t hire last time. (because this is how dumb stack ranking is done). You then attempt to refill the vacancy you’ve just created … (this bit would work if candidates are in over supply, contra to previous claims about there being no remaining qualified applicants).

56

Dave 08.12.17 at 11:35 am

I don’t know how many of you actually read the “manifesto,” but it’s mush-brained nonsense that uses the veneer of scientific skepticism to justify entirely unscientific arguments. The actual core is built on nothing but entitlement, a sense of persecution, and a mountain of unfounded assumptions — biologism about how men and women are just different; the begged question that, if those differences exist, they must relate to tech skills/interest; the assertion that affirmative action is discriminatory against the majority; the denial of discrimination against the minority; casting doubt on the qualifications of women in tech in the abstract while demurring on the subject of any particular female coworker; the preemptive defense that people who criticize his misogyny have an “authoritarian” liberal bias; and on and on.

Never mind the statistics, it’s hard to believe that anyone who understands the fundamentals of logical argument could ever take this crap seriously. Once you get past the bramble of insinuations, defenses, and misused words, the basic argument is little more than a gross tautology. Women are underrepresented in tech because they’re biologically less suited for coding work as evidenced by the fact that they’re underrepresented in tech. There’s no there there at all. It really is the most mindless, selfish sort of conservatism. The status quo works fine (for me), so the people who say it should be changed must be wrong because those changes could make things worse (for me).

That’s not really what bothers me though. What bothers me is that the whole internet seems to be talking about it and a lot of people are taking it seriously. The “liberal fascism” crowd is having a field day with this and it’s maddening. It seems half the audience (the intended half) were only willing to read as far as his initial claims of being a persecuted conservative man and were off to the races. Much of what he says later on is completely indefensible, but as long as people can pose this as a free speech issue, they can just ignore what he actually said and pivot onto the topic of liberal echo chambers and the evils of affirmative action. Meanwhile, a certain folks in the center and left (perhaps myself included) are using this whole blowup as an excuse to slag Silicon Valley “tech bro” culture in a way that mostly just seems to validate the original author’s persecution complex.

Never mind that posting a rant on Google’s internal network is hardly mounting a soap box in the town square or the inconvenient fact that belittling and undermining your coworkers on the basis of sex is illegal, the real story here is clearly that liberal social justice warriors just can’t handle hearing conservative viewpoints in their safe spaces.

57

Dipper 08.12.17 at 12:44 pm

To take one point from this. If we knew what the actual distribution was, would it be appropriate to recruit in the ratio that exactly matched that distribution? I can think of two reasons why we would want to err on the side of diversity.

Firstly, if I have a group of workers from one clearly defined social group, it is unlikely I will get applications from talented people outside that group. If I recruit someone from outside that group, then I am more likely to be able to attract talented people from outside the core group in the future as they can see this is a team where they could prosper. That added optionality on future recruitment is worth something, so on a like-for-like basis it is advantageous to recruit outside the group.

Secondly, one consequence of having employees from the same social group is the possibility of them developing a group-think that becomes damaging to the organisation. For a specific example, I know some senior managers in banks who think it is no co-incidence that desks involved in Libor fixing were predominantly white male and all from similar backgrounds, and who think that recruiting a diverse group of traders is a way of increasing the chance of having a proper open culture. Hence recruiting a diverse range of people is more likely to produce an open culture that is less likely to tolerate destructive behaviour.

58

Jay L Gischer 08.12.17 at 1:31 pm

I am a programmer. I do not think there is any such thing as “programming ability”. I think there is such a thing as “programming skill”. Maybe I’m splitting that hair too finely?

Another point: Many of the diversity program Damore objects to do not pertain to recruiting but to in-house career development. There are several women-only programs – to train women how to negotiate for salary, for instance.

Damore is very veiled about this, it’s probably not intentional. But I think he wishes he could have access to some of those programs. Instead, he’s asking for them to be ended. This might strike one as odd, but it’s a common pattern – “I didn’t get that, why should you?”

59

Faustusnotes 08.12.17 at 2:55 pm

CJColucci , kaepernick may not be the best quarterback in the league but he is definitely above average and would have a job this year if it were not for his anthem protest. It’s an insult to his achievements to suggest otherwise, and undermines your argument to suggest highly qualified people are not rejected or underpaid on a purely political basis. See also: Marie curie.

60

Belle Waring 08.12.17 at 3:04 pm

Please everyone actually read this “”manifesto”” to see that it is cribbed poorly in equal parts from the quite thoroughly discredited Charles Murray and rightly never credited pick-up blogs. Please. Also it’s racist in passing in the most telling conceivable way.

61

bianca steele 08.12.17 at 3:22 pm

DCA @ 29 makes a good point but I was just thinking about this while brushing my teeth, and I think it’s more complicated. In 1980 I was 13, was taught programming by teachers who’d learned it as adults and found it very difficult, and it was considered amazing that any of us could do it well. It was considered normal that out of 150 college bound grads, only 2 were good programmers. Of course only maybe 35 were even exposed to programming at all.

There was a period in there where it was a hobbyist thing, it required money and willingness to appear geeky, or later able to have lots of friends who did it (so the worry about appearance wasn’t there). That seems to be around the time it became a guy thing on the order of metal rock and comics. I wouldn’t be surprised if the “genius” tag hung around them even as it got easier and easier to do it.

My husband says his company gets fewer and fewer resumes from women.

62

harry b 08.12.17 at 5:26 pm

Kiwanda:

” The philosophy department of the National University of Singapore has 85% male faculty; the economics department of the University of Queensland has 78% male faculty; the philosophy department at the University of Wisconsin-Madison is 75% male; philosophy at the Univ. Bristol, 72%; International Relations group within the Poli Sci Dept. at George Washington Univ, 80%. Why are these institutions, like Frontline Analysts, unable to hire, or perhaps retain, more women? I think some investigation is needed regarding their likely biased hiring practices, their hostile work environments, and the splainy dudebros that must work at these places.”

I don’t really understand how this is a response to Daniel.

I’m in the UW Madison Philosophy department. Our graduate students are disproprortionately men but, nevertheless, I learned one year that a group of women in the department (who had been admitted in a 50/50 year) all believed they were beneficiaries of affirmative action in an admission year during which my (now deceased) colleague who was a feminist philosopher was on the admissions committee (she was never on admissions during the 23 years we were colleagues). I promptly did an analysis that showed that on every publicly available metric — publications, conference presentations, time-to-degree, etc — the female graduate students considerably outperformed the male graduate students, which suggests to me that, indeed, we (no doubt inadvertantly) do favour men. Unless, of course, we educate them better. Which I doubt.

63

William Berry 08.12.17 at 5:37 pm

@Faustusnotes:

I mean no offense, but I think you need to give CJColucci’s last sentence a closer reading. I think he covers your objection:

” . . . if Colin Kaepernick were a better quarterback, he’d be playing somewhere; if he were worse, no one would question why he isn’t playing; but he’s just mediocre enough that non-football reasons look like the best explanation.”

64

William Berry 08.12.17 at 5:38 pm

Maybe “just good enough” would have been better wording on CJC’s part. But, whatever.

65

bianca steele 08.12.17 at 5:50 pm

Something else occurs to me: how many commenters here are not in academia or journalism and would think it ordinary to publish a long attack on their employer’s management practices, in a venue provided to them by their employer, not even to mention having it picked up by major media?

I’m not defending firing people for being critical of management (whatever the perspective, political or otherwise), but on the other hand, the only two s/w engineers I’ve known essentially fired for cause (one was sacrificed to a scheduled RIF) had become disgruntled to the point where they were spending significant time trying to convince people who worked for them or whom they’d been assigned to interview, and this seems rather similar.

66

Ian Maitland 08.12.17 at 6:51 pm

Dave @ 36

Reading your comment the question is unavoidable: “but have you ACTUALLY READ IT?”

67

Ian Maitland 08.12.17 at 7:24 pm

Dipper @ 57

You say that having employees from the same social group my result in group-think that becomes damaging to the organisation., and you cite the LIBOR scandal in which “desks involved in Libor fixing were predominantly white male and all from similar backgrounds.”

That puzzles me because, first, you’d expect foreign exchange desks to be more, not less diverse, than most other bank departments. So, by your logic, a scandal should have happened elsewhere.

Second, was LIBOR a purely a lily-white affair? I take the following from WIKIPEDIA:

“In court documents filed in Singapore, Royal Bank of Scotland (RBS) trader Tan Chi Min told colleagues that his bank could move global interest rates and that the Libor fixing process in London had become a cartel. Tan in his court affidavit stated that the Royal Bank of Scotland knew of the Libor rates manipulation and that it supported such actions. In instant messages, traders at RBS extensively discussed manipulating Libor rates. In a released transcript of a 21 August 2007 chat, Jezri Mohideen, who was the head of yen products in Singapore, asked to have the Libor fixed in a conversation with other traders:
Mohideen: “What’s the call on the Libor?”
Trader 2: “Where would you like it, Libor that is?”

68

Dipper 08.12.17 at 8:16 pm

@Ian Maitland. Most desks I had experience of in the UK were white male desks, usually people who had joined the bank straight after school and worked their way to the desk from the back office (and most were implicated in some way in fixing scandals). This is not to implicate people who didn’t have degrees, just there was a closeness about them. They understood each other.

Major G7 currencies don’t require knowledge of the economies to trade but they do require people to be mentally very quick and sure of themselves and have a feel for liquidity in the market. The best way of establishing whether people had this was trying people out from the back office, hence the recruitment of local staff.

I guess in Asia they would have been the Asian equivalent. Also note the recognition from Asia that London (the biggest LIBOR products market) was doing this.

69

smh 08.12.17 at 8:17 pm

Damore proves that white boys need a mentoring program to help them in their career.

70

Lupita 08.12.17 at 9:05 pm

Google grants different periods of parental leave to primary and secondary carers. What if parental leave were enforced equally for both parents, say, each had to take six months during the first year of the child? What if working a maximum of eight hours a day were also enforced? These practices would enable both men and women to compete on equal terms for leadership positions given that both would have put in equal time and would continue to do so in the future. As it is, men and women are not competing on equal terms for leadership positions until equal overtime and parental leave are enforced equally.

Are women not entering professions were employees work vast amounts of overtime because they find the conditions oppressive and exploitative and won’t be able to compete for leadership positions on equal terms? Is tech one of them? Google? How much overtime do employees work in countries were more women enter tech? What are the parental leave laws in those countries?

I think the solution to this problem will be found by also exploring the material conditions of tech employees rather than just social prejudices and preferences, particularly those of male engineers. But then, dwelling on the thoughts of individuals, and trying to change those thoughts in order to solve social problems is a very Western, as in liberal individualist, bias.

71

Ivo 08.12.17 at 9:06 pm

“female employees in tech are generally very good.”

I don’t know how I can phrase this question without it sounding like I’m challenging this assertion, but I’m just genuinely interested: do you have evidence for this claim?

72

Omega Centauri 08.12.17 at 11:41 pm

Dipper, good points, in favor of overhiring underrepresented groups. Its not always the direct effect of the employee’s work, but also contribution to corporate culture and outreach that may matter.

Also if one wants to take a poll, and there is reason to suspect that preferences of two identifiable groups might differ. Minimum variance (expected error) is not minimized by
polling proportionately more of the smaller group -then correcting by having different weightings.
Of course a simple math question like error minimization doesn’t necessarily apply to hiring, but
it does show that an optimal strategy may be counterintuitive.

Too many of the arguments assume the goal of hiring is to minimize the “failure” rate. However, in many situations it is better to accept a higher failure rate in order to find more superstars. I suspect many potential superstars look risky (for instance if they don’t like the opportunity they will probable leave).

73

Suzanne 08.12.17 at 11:47 pm

@60: Damore was canny when you think about it. Although race is mentioned along with gender when he speaks of diversity and its fell influence on the company, he is careful not to cite any of the related pseudo-science. Note he had no such inhibitions when it came to gender.

Had he actually quoted Murray or his fellows or wrote at length about racial differences and the difficulty blacks have with coding or handling stress, he would have been fired within 24 hours. As it is, his memo was allowed to bop around the company until it was made public, or such is my understanding. In other words, it’s still okay to write and talk this drivel about women — as this very thread demonstrates.

@56: I was amused to read Damore’s complaint that Google was not a welcoming environment for conservatives. Guess this special conservative snowflake couldn’t cope and needed his own safe space. (Which I gather several wingnut organizations have offered him.)

74

Collin Street 08.12.17 at 11:50 pm

Please everyone actually read this “”manifesto”” to see that it is cribbed poorly in equal parts from the quite thoroughly discredited Charles Murray and rightly never credited pick-up blogs. Please. Also it’s racist in passing in the most telling conceivable way.

Eh, you’ve done it for me: my time’s probably better spent looking at pictures of cute rescue wombats or what have you. My turn next time you’ve got some vile shitness you want somebody to have a look at.

75

Alex K. 08.13.17 at 12:28 am

@19:

… for any entry-level coding job Google would typically have 26 qualified applicants: 20 men and 6 women. That’s the 80:20 ratio referenced above…

If Google decide their code shop should more closely resemble the demographics of their market — roughly 50:50 m/f — they have plenty of qualified women in the pool for a typical coding job.

Agreed. It will result in anti-male discrimination, but you’re saying Google should be able to get away with it as a legitimate business decision. So be it for the sake of this thought experiment.

@40: If only the OP said clearly, “the failure rate is significantly lower for the women at Google.” Instead, it observes that women in tech are generally very good. But so is pretty much everyone at Google, so the failure rate seems equally, negligibly small for both sexes. If the failure rate is about the same, it doesn’t give us any new information unless we measure it against the incoming quality averages.

76

Mark Engleson 08.13.17 at 12:30 am

This jerk got fired for not having the sense and restraint not to discuss politics in the workplace and not to refrain from going hard on a major critique of management policy. You don’t pull that crap in the corporate world. It breaks every norm of workplace etiquette.

77

Belle Waring 08.13.17 at 12:56 am

72: Done and done. Enjoy the wombats or what have you.

78

bianca steele 08.13.17 at 1:15 am

I don’t think the goal of hiring is clear. A goal of hiring for some might be to ensure a supply of thick white boys (and maybe girls, in the new dispensation) making poor use of elite educations, who can be kicked up to management in a few years, without ever having dirtied themselves by thinking like a prole. Or to ensure a supply of middling engineers in their early twenties who can be trained on doing boring work for a few years, after which some will be promoted to tech stardom or management, some will move to marketing or more (or less) customer-facing, more or less demanding, tech roles, and some will continue doing middling work. Are either of the first two necessarily “failures”? Or, I don’t know, maybe there are possibilities I wouldn’t come up with but someone else would.

79

bianca steele 08.13.17 at 1:19 am

I’ve said it before, though maybe not here: I doubt there is any profession as reliant on group work, and more full of people who think the world will come to an end if their essentially aesthetic preferences aren’t followed by all around them, than programmers, unless it’s actors.

80

Collin Street 08.13.17 at 7:32 am

Are either of the first two necessarily “failures”?

The clever thing about DD’s model is that we can treat a person’s pass-or-failureness separately for each potential future.

A person’s selected for a… vector? of potential future roles; if they turn out suitable for one of those roles, that’s a “pass”, for that role, if they turn out unsuitable then that’s a “washout”, for that role. Presumably they’re hired if their assessment “is this person likely enough not to be a washout” comes back “yes” for enough future roles. For _each_ role they are separately considered; if they turn out to be unsuitable for any specific role they’re washed out of that role, and if they turn out to be unsuitable for any role they’re washed out of the company. Which means: we can separate out the washout rates for each role individually, and assess the recruitment process for each role individually, even though as a practical matter the recruitments and the washout rates are all mixed together. And all this without access to any information beyond what we generate automatically in the hiring process.

[of course, this won’t allow for internal management problems that affect washout rates, such as for example widespread tolerance of misogynistic dickweedery among the lower decks a concommitant morale problems among female staff. But the chances that “higher standards in [minority] hiring” and “worse environment for [minority] workers” exactly cancel out are — I hope without good reason — pretty low, so any discrepency in washout rates should reveal some sort of problem somewhere…]

81

Ivo 08.13.17 at 7:48 am

@76: I’ve seen people claiming that the memo was initially sent to an internal mailing list specifically meant to discuss diversity issues. I think it changes things quite a bit if that’s true, but I have not been able to verify that claim.

82

Ian Maitland 08.13.17 at 1:21 pm

Ivo @ 81

This may be what you are referring to: “Mr. Damore said he shared the missive, titled “Google’s Ideological Echo Chamber,” about a month ago with specific individuals and groups focused on diversity before posting it to a mailing list called “skeptics” on Aug. 2. Then, Mr. Damore created a companywide discussion group for the document. As more employees took notice, Mr. Damore’s words soon spilled out onto the internet”. From the NYT August 11, 2017 “A Crisis Forces Google to Uphold Its Values While Fostering Debate.”

83

anon 08.13.17 at 1:32 pm

I was an IT professional for over four decades.

The reason there aren’t more women in IT is cultural.

Young children are told that math and science is hard. When young girls have difficulties they are ‘encouraged’ to do something else. When young boys have the same issues they are told to ‘work harder’.

In my ( admittedly anecdotal ) experience most people regardless of gender, skin color, religion or any other such attribute are only ‘adequate’ programmers. To be a great programmer a person needs to be closer to an extreme end of the Asperger scale.

Most people, thankfully, are not.

84

RichardM 08.13.17 at 2:21 pm

I am sure that everyone is here is aware that most orchestras have switched to blind auditions, where the judges listen to potential musicians from behind a curtain. Hiring of female musicians rose significantly after this was introduced.

The results when someone tried something similar for tech interviews are perhaps interesting:

http://blog.interviewing.io/we-built-voice-modulation-to-mask-gender-in-technical-interviews-heres-what-happened/

They seem to generally support Damore’s point that there does appear to be a relevant difference in preferences that you would need to account for in order to come up with an optimal diversity strategy. And in a world where literal Nazi’s are parading the streets and Trump is the president, it would be pretty unexpected if a clumsy or problematic diversity strategy was able to prevail by sheer force of will.

Freedom of speech isn’t an abstract or first order goal, it’s a tactic. Like any tactic, if you abandon it, you lose the benefits of applying it.

85

James Wimberley 08.13.17 at 4:27 pm

As an extreme example of genetically influenced disparities in ability, take tennis. (Sailing would also do). It won’t I trust be controversial to posit that in the population as a whole, tennis-playing potential will be skewed toward men by virtue of being taller and having longer arms and a stronger upper body. The disparity holds in the upper tail of the distribution; the results of the various “battles of the sexes” between high-ranking players organised as stunts are pretty lopsided, and nobody has ever bothered to organise one between reigning champions. Now suppose there is an eccentric, prestigious and tony suburban tennis club that only holds both-gender competitions, and you are a gambler betting on the results. Is there a Bayesian prior in favour of male players here? I doubt it. Far too many other factors enter into the recruitment of the club, such as money, motivation, and peer pressure, and it is these the gambler should concentrate on. It’s too far from a random sample of the general population, or even the upper tail of the tennis-genetic-ability distribution. The same goes for Google. Its recruitment pool is graduates of highly ranked college computer science programmes in California. Stanford’s is 30% women, Harvey Mudd’s 50%. Google’s intake is a much lower percentage.

86

Omega Centauri 08.13.17 at 5:00 pm

Ivo. If it was intended to be some internal discussion group where all opinions are welcome, and the intent was that comments say within, unless specifically authorized by their authors, it would put the suit over the firing in a different light.

I did see something that tends to support that, that it wasn’t a blanket emailing out of the blue, but had originally been part of a discussion group, but over a period of time it began circulating throughout Google. Doesn’t affect the merits/demerits of the “manifesto”, but it does affect whether the firing was fair and proportionate.

87

Phil 08.13.17 at 5:03 pm

I worked in computing from 1983 to 1995. Admittedly, all my workplaces were majority-male, but none of them had a sex ratio of 4:1 or above; 5:2 or 5:3 was much more usual. Apparently the industry has got more male-dominated – unlike pretty much every other walk of life – in the last twenty years. But clearly this is just the way things are, and any attempt to fix it would be the rankest political correctness and also doomed to failure.

88

Ian Maitland 08.13.17 at 6:43 pm

Jason @ 85

I have wondered about chess. Any take on that?

Not so fast on Google’s recruiting pool. Nationally women account for less than 18% of college computer science grads. So after top schools than can pick and choose the numbers seem to tail off sharply.

Of course Damore did not raise the issue of Google’s recruiting pool. But he might have pointed to the low proportion of women computer science graduates to support his claim that biological differences may largely explain why women on average choose the tech field less often than men.

The same goes for Google. Its recruitment pool is graduates of highly ranked college computer science programmes in California. Stanford’s is 30% women, Harvey Mudd’s 50%. Google’s intake is a much lower percentage

89

Mark Engleson 08.13.17 at 9:09 pm

Someone brought up in this conversation the onerousness of a diversity training. Oh, the horror! When you consider what Google engineers get paid, if they have to suffer through a seminar they find uncomfortable, well, I just can’t get sympathetic to that. A lot of jobs, a lot of poorly paying jobs, have regular conditions that are dangerous, that are demeaning.

Grow up already. Feeling a little uncomfortable for a few minutes or a couple of hours at a highly desirable and extremely well-paying job is small potatoes. Get over it. It’s an entitled attitude to think your workplace should be so suited to you that never feel any stress there.

90

Kiwanda 08.13.17 at 10:06 pm

harry b:

I don’t really understand how this is a response to Daniel.

In the most direct way, perhaps not: it’s true that there’s been no memo from Daniel Davies, John Holbo, John Quiggin, yourself, or Henry Farrell, giving some broad gender-differences-based explanation for why each of your respective departments (or in Davies’s case, senior staff) is about 80% male. Nor are academic departments or consulting firms the same as large technical companies in their staff turnover etc. But still: I understand from the OP that staffing ratios are just a simple matter of appropriate recruiting. If recruiting to these organizations is not discriminatory against women, and their work environments are not hostile to women, what explains these lopsided ratios?

91

Suzanne 08.13.17 at 11:29 pm

@89: Those diversity training sessions, like sexual harassment seminars, are more necessary than many realize. As the priest observes in “Brideshead Revisited”: “But these young people have such an intelligent, knowledgeable surface, and then the crust suddenly breaks and you look down into the depths of confusion you didn’t know existed.”

(The foregoing applies to older people, as well.)

92

Ian Maitland 08.14.17 at 1:50 am

Mark Engleson @ 89

I have taken the liberty of re-writing your post.

“Someone brought up in this conversation the [workplace gripes of women at Google and the occasional microaggressions than they experience]. Oh, the horror! When you consider what Google engineers get paid, if these women have to suffer through some [oafish male pranks that make them] uncomfortable, well, I just can’t get sympathetic to that. A lot of jobs, a lot of poorly paying jobs, have regular conditions that are dangerous, that are demeaning.”

“Grow up already. Feeling a little uncomfortable for a few minutes or a couple of hours at a highly desirable and extremely well-paying job is small potatoes. Get over it. It’s an entitled attitude to think your workplace should be [redesigned around you so that] you that never feel any stress there. [For goodness sake, man up].”

93

Ian Maitland 08.14.17 at 2:10 am

94

J-D 08.14.17 at 10:11 am

If employees and/or applicants for employment receive differential treatment from Google and other large tech companies on the basis of gender, then it’s very likely to have an effect on people’s propensity to apply for jobs with them, and perhaps even on people’s propensity to study computer science. If the pool of job applicants and/or the pool of computer science graduates has a low percentage of women, the behaviour of Google and its peers is probably part of the explanation, and if they change their behaviour it could the effect of increasing that percentage.

And so —

I agree with most of the OP, but the problem is that you can’t really read this on the merits as such. You have to read it on the perceived merits of how the alt-right freakout machine will take it. It was made to sound very reasonable, using all of the so-called SJW language and references to science and yet elicit the very reaction that it brought on: termination, thus giving them the martyr that this person seemingly wanted to be.

The key thing that they will take away from this is that you can’t even acknowledge that differences between men and women—constructed or not—exist. Is that what it says? Is that what it means? No, but that’s the text for the alt-right reader. And of course that straw man text makes it absurd on its face to argue with.

You could post forever about how and why the memo is wrong, but you’re wasting your time.

— arguing against the memo isn’t going to stop the ‘alt-right freakout machine’, but that doesn’t make it a waste; if arguments are voiced, it’s not only the alt-right freakout machine that hears them.

95

sanbikinoraion 08.14.17 at 12:48 pm

@Crissie, Daniel

As someone who’s done hiring in software (not for Google) the idea that 20% of candidates are qualified for even a moderately-challenging job is laughable. Try closer to 2%. Of those 130, maybe 60 will be illegible CVs that are heading directly to the bin. Another 60 will be CV writeoffs for other reasons: not enough experience, wrong kind of experience, etc. You might phone screen the remaining ten and bring only 3-5 to interview. That’s approximately my experience. Maybe Google gets better CVs than I do on average but I seriously doubt that; I suspect anyone and their dog applies to every available Google vacancy.

Let’s do the numbers again assuming 2% “qualified” instead of 20%. You get 2 men and 0.6 women. Round that up to 2 qualified men and 1 qualified woman and hire the woman, and suddenly your thought experiment looks rather worse.

Note that I think Google Guy is basically wrong on all points — but I also think this example is way, way out.

96

CJColucci 08.14.17 at 3:13 pm

FaustusNotes: It seems we disagree somewhat about Colin Kaepernick’s abilities, but I don’t think it matters much to the main point. I believe as much as you do that he would have some job in the NFL but for his anthem protests. And I believe it precisely because Kaepernick is good enough to hold some job in the NFL — not a starter for any team with the possible exception of Cleveland or the Jets, but certainly as a back-up somewhere. Given the type of player he is and the type of QB he’d be backing up, Seattle would have made a lot of sense. But I think if he were a significantly better QB than he is he would have a job regardless of his anthem protests; and if he were a worse QB than he is, that would sufficiently explain his lack of a job. Not to say that the anthem protests wouldn’t have hurt, too, but you rarely bother with non-football reasons for why a bad player isn’t playing.

97

TM 08.14.17 at 3:14 pm

I don’t think this whole debate has much to do with statistics either being hard or being poorly taught. Or really, with statistics at all.

98

map maker 08.14.17 at 5:37 pm

Fun piling on to Google. Lord help us if someone ever leaking the minutes of a search committee at an R1. “Open searches” that are anything but, job descriptions that have everything but the preferred candidates name, etc., etc. But yes, easy to throw bricks at Google.

99

anon/portly 08.14.17 at 6:19 pm

Currently 60:

Please everyone actually read this “”manifesto”” to see that it is cribbed poorly in equal parts from the quite thoroughly discredited Charles Murray and rightly never credited pick-up blogs.

I can’t see where even one single sentence in the memo (let alone 50% of it) is “cribbed” from Charles Murray. The word “race” appears 5 times, by my count, twice in headers, but is there mainly (by my reading) because the diversity programs the author is objecting to include both gender and racial diversity. (“IQ” makes an appearance in a footnoted aside).

Meanwhile, although the author never brings up “race” to specifically object to or discuss practices concerning racial diversity, he brings up “woman” over and over and over. The memo is clearly focused on what the author doesn’t like about Google’s efforts to hire a greater proportion of women.

Perhaps this is all a Trojan horse, and what the author really hates is the effort to hire more minorities, but what’s the point of surmising that? Yes, it could be true, but it might not be true. Maybe all of the specific practices that he’s seen or been subject to and didn’t like have involved women and not minorities.

The “evo-psych” stuff about women and their interests and the whole discussion of why women are under-represented in some fields and over-represented in others may be complete twaddle, but it doesn’t seem like the same sort of stuff that Charles Murray does. I think I understand at least some of the objections to Murray’s ideas and interpretation of data, so maybe I am just unfamiliar with how those objections carry over to these other ideas and interpretations of other data.

Currently 89:

Someone brought up in this conversation the onerousness of a diversity training. Oh, the horror! When you consider what Google engineers get paid, if they have to suffer through a seminar they find uncomfortable, well, I just can’t get sympathetic to that. A lot of jobs, a lot of poorly paying jobs, have regular conditions that are dangerous, that are demeaning.

Grow up already. Feeling a little uncomfortable for a few minutes or a couple of hours at a highly desirable and extremely well-paying job is small potatoes. Get over it. It’s an entitled attitude to think your workplace should be so suited to you that never feel any stress there.

This gets at what the Memo is actually about, how Google’s efforts to hire to hire more women are, in the author’s view, counter-productive in that they have created less attractive working conditions for some other workers, and that on balance this is hurting the company. But, criticizing him this way obviously pre-supposes that he’s wrong. What if he’s right?

Much of the response by the Left to this memo seems to implicitly assume that in this case, Google, because they’re trying to hire more women, is getting it right. Well, how does anyone know? Not every bureaucratic outcome in a corporate context is a good one, not even ones with a “progressive” intent, surely. (Also, “slack?”).

I have myself never been subject to any onerous diversity training or practices of any sort. But I do know people, left-wing people in particular, who have expressed strong reservations about various “diversity bureaucracy” practices (in an academic environment mostly).

If a worker in a job that was unnecessarily “dangerous” or “demeaning” made an effort to improve working conditions, I think everyone would agree this is good. If the author of the memo is correct, then it seems the invective above (and also of course his firing) is hard to justify.

Of course the author might be wrong also, Google’s actions in this regards may be working to make the company better. My point is that I don’t see any evidence for this either way[1], and yet everything else is kind of besides the point. In a way harry b’s comment 62 or the discussion of Kaepernick are the most relevant comments, because at least there’s some recognition of the desirability of having evidence.

[1] I guess you could take the idea that the author’s evo-psych arguments about women are bad as evidence that his arguments about Google’s working environment are bad, but is that necessarily so? I also find myself less sympathetic towards the author and his concerns because of his focus on “conservatives” and ideology – if these programs are really a problem, I would think they would be alienating many non-conservatives as well, yet the author never seems to entertain this notion – yet I don’t know that this is really evidence against his view, either.

100

William Berry 08.14.17 at 6:24 pm

@ian mainland:

So very MRA.

Yes, of course being a little stressed or uncomfortable on occasion in the workplace is exactly the same as being sexually harassed and discriminated against.

Christ, what an asshole.

101

William Berry 08.14.17 at 6:27 pm

Well, OK, Maitland, then, not “mainland”.

102

Doug K 08.14.17 at 8:01 pm

DCA – I had that cite all ready to go, glad someone beat me to it – thank you ;-)

With Github it is in fact possible to have blind auditions for programmers. Guess what, turns out to be the same as blind auditions for musicians, as BigHank53 mentions – seem to be a lot more good women once the gender bias is disarmed. How odd.

https://news.ncsu.edu/2017/05/gender-bias-in-programming-2017/
“Programmers who could easily be identified as women based on their names or profile pictures had lower pull request acceptance rates (58 percent) than users who could be identified as men (61 percent). But woman programmers who had gender neutral profiles had higher acceptance rates (70 percent) than any other group, including men with gender neutral profiles (65 percent).
78.7 percent of women’s pull requests were accepted, compared to 74.6 percent for men.”

It’s certainly not a pipeline problem either. As James Wimberley mentioned, there are more CS women and minority graduates entering the workforce than are getting hired. This includes graduates from the highly competitive universities where the tech companies do all their recruiting – MIT, Stanford, Harvey Mudd, etcetera.
These statistics are quite simple.
From 2014,
https://www.usatoday.com/story/tech/2014/10/12/silicon-valley-diversity-tech-hiring-computer-science-graduates-african-american-hispanic/14684211/
From 2017,
http://www.slate.com/articles/technology/technology/2017/08/facebook_s_employee_diversity_numbers_are_not_a_pipeline_problem.html
Every time a study of this kind is done, the results are the same.

We are forced to conclude that the gender disparity in tech is due to rancid fuckwit techbros like Damore, and the hostility they engender towards women and minorities. I’ve worked in the industry since the 80s, and I saw it happen. The 10x programmer is a myth, see link from my name. On the other hand, women in tech tend to be 2-3x better than the men in my experience – they have to be in order to survive.

103

Ian Maitland 08.14.17 at 9:37 pm

William Berry @ 99

I hope you mean “Maitland” rather than “asshole.” Regardless, I’ve been called worse. (And I get a kick out of provoking assholes).

But, empirically speaking, where do you get the idea that women are sexually harassed and discriminated against at Google? True, there may have been a few isolated egregious cases, but what workplace is completely free of those? The difference is that the liberal media machine seizes on any misbehavior in Silicon and uses it to smear every male employee. The political motivation is obvious.

I think that Doug K @ 100 backs up my claim that the issue is about being a ” little stressed or uncomfortable on occasion in the workplace” (as you put it) rather than something as serious as being sexually harassed. Doug concludes that “gender disparity in tech is due to rancid fuckwit techbros like Damore, and the hostility they engender towards women and minorities.” (Please disregard the stereotyping).

Really, if the earnest gentle well-intended Damore is the major barrier to women in tech, then my pastiche of Mark Engleson @ 89 is closer to the mark than I imagined. We are back in the make-believe feminist world where a micro-aggression can cause a nervous breakdown. (Damore has described his mandatory diversity training: “You’re not allowed to say that — that’s sexist” ).

Surely, the most anti-feminist meme around is the (contemporary) feminist meme of women as passive and helpless and devoid of any agency. This is why women are underrepresented on Google’s workforce?!

104

engels 08.14.17 at 10:16 pm

easy to throw bricks at Google

🎻🎶🎶🎶

105

William Berry 08.14.17 at 11:48 pm

@Ian Maitland:

Well, OK, I took it that the Internet debate had broadened to the general issue of discrimination against women in hiring and promotion in tech, and not just at Google.

Your silly “re-writing” of Engleson’s post includes the priceless “workplace gripes of women at Google and the occasional microaggressions that they experience” and “oafish male pranks” [that make women “uncomfortable”]; smarmy cracks that are so typical of the MRA’s technique of trying to minimize the actual harms experienced by others so that your own whining doesn’t look so contemptible in comparison.

What you are really doing here is just confirming what I have always known about you MRA types (and right-wingers in general): the one thing you are really good at is crying.

From what I’ve seen about you online, you’ve made a good career out of it.

Crooked Timber really needs a better class of troll.

106

Ian Maitland 08.15.17 at 2:29 am

William Berry:

Ouch! I am sorry that I don’t live up to your standards.

OK, my rewriting of Engleson’s post was silly and tongue-in-cheek, but what is truly dismaying is that it is another case of nature imitating art: It faithfully reflects a lot of the everyday silliness that goes down in today’s workplaces. Try reading some of the media accounts of supposed workplace sexism. Take the case of the “Anita Hill” of Silicon Valley, Ellen Pao. Pao sued Kleiner Perkins, demanding an eight-figure payout, complaining that she had not been seated at the center table during key meetings. She said a Kleiner partner behaved inappropriately when he gave her a book of Leonard Cohen’s poems and drawings (entitled Book of Longing) on Valentine’s Day and asked her out to dinner on a night when his wife was out of town. Kleiner’s attorney elicited that Pao had engaged in a consensual affair with a married co-worker (and later intervened to prevent him being fired, suggesting a female colleague be fired instead). It is truly cringe-inducing stuff. “Crying” is not a bad description. Even though the jury found against Pao, the liberal media canonized her and credited her with having broken down barriers facing women. What is our culture teaching young women about how to advance professionally?

Btw, William, I had to go online to find out what MRA was. The truth is that I am an old-fashioned feminist. I didn’t leave the feminist movement, it left me. Real feminism would never have insisted that the whole workplace be redesigned around the particular needs and wants and phobias of women employees. (Men are humans too). I am no men’s rights advocate, but I despise people who think they are entitled to special rights.

How are real women or men helped if men perceive every woman as a lawsuit waiting to happen? How is it helpful if male colleagues walk on eggshells around women? Or clam up out of fear of saying the wrong thing? What older man (Like John Doerr at Kleiner who mentored Pao) is going to take that risk next time? Will we know that women’s liberation has finally arrived when it is taboo for male engineers to give female engineers books of poetry by Leonard Cohen?

107

wierdnoise 08.15.17 at 5:26 am

BigHank53 already posted this link but I think it well worth reading so I reposted it here.

I’ve worked as a software engineer for 40 years, and observed both the rise and fall of women’s participation in the field and the development of toxic brogrammer culture. And I’ve worked (and continue to work) at some of the big-name Silicon Valley internet companies and can say that Yonatan’s experience matches mine. I spend about 15% of my time actually writing code (on average — a bit more when I was more junior, a bit less today). The rest is spent in communication with managers, customers, and other programmers. I have noticed some differences between women and men software engineers. Like Yonatan observes, the women seem on average better at the whole mutual support, teamwork, and yes, political aspects of software engineering. Thing is, we don’t interview for those qualities. We interview on “objective” criteria like high-stress coding challenges. Yet my experience is that successful teams are the ones where those other things are most present.

Software engineering isn’t like most other engineering disciplines, and in fact many would refuse to call it engineering at all. There is a lot more informational communication, and a lot less math and detailed specifications than, say, electrical or mechanical engineering (something that may be entirely contrary to what folks outside of the profession might assume). The social aspect is highly important; frequent discussion of design and progress and, yes, empathy in the sense of understanding a coworker’s thinking. And although I cast a skeptical eye toward arguments from biology, the fact is that women are expected by our culture to do better in these domains, and thus often do so. Make of that what you will.

It amuses me, in a sad way, that there has been a lot of effort to engineer the social and communication environment of software development, techniques such as “agile programming” — and yet the brogrammer “hold my beer while I write some spectacular lines of code” ethos still predominates.

One final comment: that screed circulated within Google for weeks before its existence was leaked to the public. The result was a textbook case of a hostile working environment for the women there. Shame on Google for letting it go on so long.

108

J-D 08.15.17 at 5:36 am

anon/portly

This gets at what the Memo is actually about, how Google’s efforts to hire to hire more women are, in the author’s view, counter-productive in that they have created less attractive working conditions for some other workers, and that on balance this is hurting the company. But, criticizing him this way obviously pre-supposes that he’s wrong. What if he’s right?

I don’t know what kind of efforts Google is making to hire more women, but it seems to me that any kind of systematic effort to hire more women is almost certain to make any workplace less attractive to some people, namely, sexists, chauvinists, and bigots. I think that’s (as they say) ‘a feature, not a bug’.

It should be obvious that a premise like ‘You should never do anything that makes the workplace less attractive to some workers’ is bound to lead to bad conclusions and should be rejected.

Ian Maitland

The difference is that the liberal media machine seizes on any misbehavior in Silicon and uses it to smear every male employee.

To me, that statement looks like an indefensible lie, but if there is some defence of it that I can’t see, by all means share it.

109

Collin Street 08.15.17 at 6:24 am

To me, that statement looks like an indefensible lie

Every hard-right activist without exception displays obvious signs of theory-of-mind/empathy impairment.

110

Pavel A 08.15.17 at 7:07 am

Ian Maitland@106:

Are you a walking, talking bingo card of sexism?

– “Real feminism would never have insisted that the whole workplace be redesigned around the particular needs and wants and phobias of women employees.” (No True Scotsman fallacy and cluelessness about 3rd wave feminism)
– “I am no men’s rights advocate, but I despise people who think they are entitled to special rights.” (Women just want “Special Rights” and not “real egaliatrianism” which is actually more of the same old shit)
– “What is our culture teaching young women about how to advance professionally?” (Women aren’t qualified, they only complain to get ahead)
– “Will we know that women’s liberation has finally arrived when it is taboo for male engineers to give female engineers books of poetry by Leonard Cohen?” (Man, I was just being a nice guy and was totally not stalking her, why don’t women just like nice guys, etc)
– “The truth is that I am an old-fashioned feminist” (Insert hyperchicken joke here)
– “How is it helpful if male colleagues walk on eggshells around women?” (My ability to slap women on the ass and talk about my junk all day is being taken away. Wah)
– “Or clam up out of fear of saying the wrong thing?” (By asserting their right to a safe workspace, women are silencing men instead of… you know… asking them to be decent human beings and shit)

I think I got at least a row in there. Go read something about feminism newer than the 1970s.

111

Hidari 08.15.17 at 7:46 am

If James Damore is so furious about being fired, why doesn’t he just go to his trade union?

Oh but wait.

112

Jeff Guinn 08.15.17 at 9:53 am

Perhaps it makes sense to look just a little further afield than just Google or programming.

When it comes to mechanical aptitude, there is a significant difference between the genders, very nearly one standard deviation. It doesn’t require too great a leap of faith to suggest that there is a strong correlation between aptitude and inclination: people generally prefer to do what they are good at.

So, given that type of self-selection, assume that the population of male mechanics has a mean mechanical aptitude one standard deviation to the right of the male mean. There will be very, very few women whose mechanical aptitude is at/above this mean, because it is 2 SDs to the right of the female mean.

Therefore, one shouldn’t be surprised that any occupation with the word “mechanic” associated with it is over 97% male. There isn’t any physical reason why women couldn’t be small engine mechanics, but 99.5% are male.

Further, it is also true that men are significantly more adept at 3D visualization and systematizing.

Up until the mid-1970s women were effectively barred from being pilots (my profession). By 1980 or so, all formal barriers were banished. By 2000 women went from 0% of airline pilots to just below 6%; that ratio has remain unchanged ever since.

By this point, considering that any woman who has gained the required qualifications will get an interview, and, absent an extensive DUI history will get hired, the only remaining explanation is that women overwhelmingly choose not to be airline pilots. Because of that, there is no way in the world that an airline could get anywhere close to 50-50 F-M.

As it happens, there is a strong correlation between mechanical aptitude, 3D visualization, and systematizing to success in pilot training. And, since it is also true that women are more inclined to work with people/animals than things, and prefer social environments to asocial — and being a pilot is as much about things and as little about people as it gets — then there should be no surprise that women overwhelmingly don’t choose a high status/high pay occupation.

Perhaps the same reasoning applies with programming — that mechanical aptitude, systematizing and a preference for things instead of people are correlated with an inclination towards IT in general and programming in particular. After all, it is clear that women overwhelmingly choose not to do other things, why not this?

BTW, my son just graduated from WSU with a degree in electrical engineering. Despite all the outreach one could imagine, and free tutoring to women, only 4 out of 105 in his class were women.

113

James Wimberley 08.15.17 at 10:34 am

Slightly OT but while we’re on diversity, a significant transgender anecdote. Ever wondered why Silicon Valley’s hardware flagship Intel is being slowly eaten alive by upstart chip design shop ARM of sleepy Cambridge (England)? In recent years. it’s mainly because the business model (“decent and regularly improved low-power processors licensed openly to all comers, including Chinese” has proved superior to “powerful energy guzzling processors made only by us and sold at huge monopoly markups”. But the low-power bit goes back to Sophie Wilson, who as a young man designed the first processor in 1985 for what was then Acorn to meet a BBC specification. Advert for old blog post with the story.

You wonder why the obviously super-talented young Roger Wilson, as she then was, was not headhunted away by Silicon Valley for much more money. My speculation (somebody should ask her) is that it was the working environment. Discrimination, harassment and simple incomprehension must be much, much worse for transgender people than for women who are usually present in numbers for solidarity. If Roger/Sophie had found a safe working environment at Hauser’s little startup, she would have been very reluctant to leave it for the macho culture of Silicon Valley. She eventually did move, but to another Cambridge company. The city is notoriously tolerant of discreet sexual diversity, in a generally more tolerant country. Actions have consequences, as they say.

114

tomsk 08.15.17 at 12:33 pm

This has to be parody/trolling, right? Nobody could write this stuff, look back over it, think hmmm, yes, that’ll do nicely. That’ll win others over to the old cause, what. The comical moaning about Google’s failure to “defend the honor of its male employees” seemed like a red flag but this is clinching. Isn’t it?

115

tomsk 08.15.17 at 12:35 pm

Sorry, seem to have screwed up the formatting there. My comment was directed at Ian Maitland’s astonishing statement that “The truth is that I am an old-fashioned feminist. I didn’t leave the feminist movement, it left me.”

116

orangeman 08.15.17 at 2:13 pm

why is it ok to discriminate against people due to moderate differences in skill level ?

well, we all think it is ok for Surgeons or airline pilots
but programmers ?
why do private corporations have this right ?

117

engels 08.15.17 at 3:09 pm

Every hard-right activist without exception displays obvious signs of theory-of-mind/empathy impairment.

Not this again

118

Ian Maitland 08.15.17 at 4:29 pm

J-D @ 108:

A few seconds of Googling brought me this:

https://qz.com/1020790/sexual-harassment-is-finally-starting-to-topple-silicon-valleys-powerful-male-elite/

SEXUAL HARASSMENT IS FINALLY STARTING TO TOPPLE SILICON VALLEY’S POWERFUL MALE ELITE
By Alison Griswold

“Dave McClure. Justin Caldbeck. Matt Mazzeo. Jonathan Teo. Amit Singhal. Ed Baker. Emil Michael. Travis Kalanick.
This is a short list of men who have lost or left their jobs in technology after alleged sexual misconduct surfaced at their companies….
Something is changing in Silicon Valley, where sexism has long been a known problem, but has rarely been dealt with…
Sexism runs deep in Silicon Valley, where “brilliant jerks” have long been tolerated and even prized, and the poor treatment of women, as reporter Jeff Bercovici once said, “is a feature, not a bug”…
Sexism has been ignored for in tech for a very long time. 2017 appears to be the year when Silicon Valley can no longer remain willfully ignorant”.

Notice two things:

1. Griswold’s blurring of sexual misconduct and sexism.

The unsupported implication is that there is a link between raping or sexual harassing a female colleague, on the one hand, and telling a joke that offends her, on the other. All men — or all men who have made a female colleague uncomfortable — are therefore (in some degree) guilty by association for the outrageous behavior of a few. I call that a smear.

2. The stereotyping of male software engineers as sexists.

Griswold does not say that sexism is universal among male engineers in Silicon Valley, but she comes pretty close. Consider: The poor treatment of women in Silicon is a “feature, not a bug”; and in Silicon Valley “sexism has been a long known problem.” (Remember that Damore was fired by Google because he allegedly stereotyped women; the charge was false). Or look no further than the comment from wierdnoise @ 107: [s]he says that in Silicon Valley a “‘brogrammer hold my beer while I write some spectacular lines of code’ ethos still predominates.” I believe these unsupported attributions smear a lot of people.

119

Ogden Wernstrom 08.15.17 at 6:01 pm

Will we know that women’s liberation has finally arrived when it is taboo for male engineers to give female engineers books of poetry by Leonard Cohen?

So, it was the book, not the asking-her-to-dinner-on-a-night-when-his-wife-is-out-of-town?

120

anon/portly 08.15.17 at 6:31 pm

I don’t know what kind of efforts Google is making to hire more women, but it seems to me that any kind of systematic effort to hire more women is almost certain to make any workplace less attractive to some people, namely, sexists, chauvinists, and bigots. I think that’s (as they say) ‘a feature, not a bug’.

I think one could agree with your point entirely and also agree with my point entirely. Right? My point is that, just because you or I agree with the purpose of a bureaucratic process in a company or governmental institution, doesn’t mean that bureaucratic process in a specific instance can’t be poorly or unfairly implemented.

As I said, I think the (mem0) author’s focus on harm to “conservatives” kind of off-putting or odd, as I would think that the type of harm he’s describing, if significant, would apply more generally. Also whereas I objected to someone saying it was half cribbed from Charles Murray, I didn’t object to the idea that it was half-cribbed from “pick-up blogs” (a la Roissy, perhaps). But then as far as something like “discernment concerning academic research” goes, maybe this is just what the 50th percentile for software engineers looks like.

All the stuff about “ideological echo chambers” and “[s]peaking up without fear of being harshly judged,” I find hard to relate to, as I’ve experienced nothing like it. Sure, there are times where you start discussing something with a fellow employee (or group of fellow employees at a meeting), you realize it’s pointless to say certain things, as the person or people you are talking to don’t have nuanced enough views to be able to suitably process what you’re saying. At times I’ve held back from making a point because I realize that to them, it’s just a “conservative” point, and it would just take too long to get them to a better understanding. At the same time the people I can talk to about anything have generally been left-wing people.

So I think the idea that this guy just wants to be able to make points that are “sexist” or “chauvinist” or maybe just plain garish, is hard to reject. But maybe also at Google it’s just difficult to have a reasonable conversation.

121

Niels 08.15.17 at 9:24 pm

PatinIowa (12) I agree completely, except, I would still fire him even if he were one of those so-called star programmers.

You don’t need to protect the welfare of a white male over-privileged (Harvard eh?) youth.

122

Moz of Yarramulla 08.15.17 at 10:31 pm

anon/portly@120 the idea that this guy just wants to be able to make points that are “sexist” or “chauvinist” or maybe just plain garish, is hard to reject

That is my take on it. He was fired for making it impossible for anyone to work with him, because he felt the need to make points that are unacceptable (and possibly illegal) to his coworkers. But he was also discriminated against, in a self-fulfilling way, because he said things that his management “couldn’t bear to hear” about the inherent abilities of certain types of employees.

My experience of hiring programmers into small companies is that we bin between 90% and 99% of applications at first glance. I can reliably go through a set of 500 emailed applications in less than 10 minutes to end up with between one and fifty that are worth more than a glance. Fifty would be exceptional and rare, zero would be less surprising than fifty. Most of the rejects are “I can’t write readable English”, “I need visa sponsorship”, and then “I have no relevant qualifications or experience” being the rest of the instant rejections. After a while I start questioning myself, but then I think “if someone can’t put their qualifications front and centre in the cover letter they don’t understand or perhaps care enough about what their manager wants to be useful”. Viz, a big part of being a good employee is thinking “what does the other person want from me?” … and acting on that. The question of what to do if a woman ever makes it through that initial filter has never arisen.

Which feeds back into Google-boy’s stuff-up. When he sent that epistle to his coworkers, what did he think would happen? If he’s sufficiently naive to have expected them to say “OMG, you’re right, we’ve been *such* idiots about this” could he recover? How much effort would it take to turn him into a useful part of their organisation?

123

J-D 08.15.17 at 10:40 pm

Ian Maitland
‘Widespread’ and ‘universal’ are not synonyms; a report that says a problem is widespread in Silicon Valley (or any other place or context that happens to be under discussion) is not fairly interpreted as saying that the problem is universal in Silicon Valley and is not, therefore, a smear of everybody in Silicon Valley.

124

Ian Maitland 08.16.17 at 1:39 am

The absolute best thing I saw last week arising out of the Damore/Google affair is Megan McCardle’s story of how she decided that a job building servers and workstations was not for her.

It wasn’t because the work was heavily male. It wasn’t about client outings involving strippers, to which she wasn’t invited. It wasn’t the sexual harassment — entirely from clients, not colleagues. It wasn’t because she thought being a woman would impede her career. She liked working with the bros just fine. None of this bothered her or her work; she didn’t lose any sleep over any of these things.

“No, the reason I left is that I came into work one Monday morning and joined the guys at our work table, and one of them said, “What did you do this weekend?”
I was in the throes of a brief, doomed romance. I had attended a concert that Saturday night. I answered the question with an account of both. The guys stared blankly. Then silence. Then one of them said: “I built a fiber-channel network in my basement,” and our co-workers fell all over themselves asking him to describe every step in loving detail.
At that moment I realized that fundamentally, these are not my people. I liked the work. But I was never going to like it enough to blow a weekend doing more of it for free. Which meant that I was never going to be as good at that job as the guys around me.
So I went to business school….”

There is more. For you lefties, today she is the best blogger/columnist around. They don’t make women like her anymore. I’d call her an old-fashioned feminist.

https://www.bloomberg.com/view/articles/2017-08-09/as-a-woman-in-tech-i-realized-these-are-not-my-people

125

William Berry 08.16.17 at 4:20 am

@Pavel A: What you said @110

I admire your patience. I am getting old. I just can’t do it any more.

But, if you haven’t done it already, I suggest you take a look at Professor Maitland’s history on the Google. You will learn that he has been a “feminist”, by his idiosyncratic definition (well, not really idiosyncratic; it’s actually a fairly standard pose within the MRA crowd) for a long time.

Well, actually, what you will learn is that he is the poster-boy for White Male Privilege in Academia; the guy who cried all the way to the bank by socking it to those nasty women who dared to want to be paid as much as him.

126

Orange Watch 08.16.17 at 5:19 am

Doug K@102

This includes graduates from the highly competitive universities where the tech companies do all their recruiting – MIT, Stanford, Harvey Mudd, etcetera.

*sigh* In these conversations everyone always forgets my alma mater, even though it’s generally neck-and-neck with Stanford and MIT for top CS program in the country. Alas, dear Carnegie Mellon, I fear it’s because you’re in Pittsburgh, and even if Pittsburgh is an up-and-coming tech hub, it’s still Pittsburgh.

Anecdotally, my Language Technology cohort w/in CMU’s School of CS a little over a decade ago had ~25-35% female students (and a similar percentage of profs). The CS program proper had noticeably fewer, though.

127

Jeff Guinn 08.16.17 at 5:26 am

[Moz:] He was fired for making it impossible for anyone to work with him, because he felt the need to make points that are unacceptable (and possibly illegal) to his coworkers.

Could you quote just one?

128

J-D 08.16.17 at 7:30 am

Ian Maitland
If giving your weekend to your employer for free is a requirement for career success, that is a sign of something seriously wrong.

129

J-D 08.16.17 at 7:40 am

anon/portly

I think one could agree with your point entirely and also agree with my point entirely. Right? My point is that, just because you or I agree with the purpose of a bureaucratic process in a company or governmental institution, doesn’t mean that bureaucratic process in a specific instance can’t be poorly or unfairly implemented.

What is not clear is how you could possibly think that point is worth making in the specific context of this discussion. It is of course true that sometimes people bungle whatever it is that they are trying to do (whether that is making a workplace more welcoming to female employees or anything else), but this is so generally true and so easily perceived that it is hard to think of any context where it would be a point specifically worth making. It is this that makes it tempting to suppose, in a spirit of interpretive charity, that you had some other point you were trying to make, and if that is not in fact the case it is hard to blame people for misunderstanding you.

‘Is that all you meant?’ the reader naturally wonders. ‘That sometimes people blunder? We all know that. Did you really post a comment just for that? No, there must be something else there.’

130

J-D 08.16.17 at 8:00 am

Ian Maitland

How are real women or men helped if men perceive every woman as a lawsuit waiting to happen? How is it helpful if male colleagues walk on eggshells around women? Or clam up out of fear of saying the wrong thing? What older man (Like John Doerr at Kleiner who mentored Pao) is going to take that risk next time? Will we know that women’s liberation has finally arrived when it is taboo for male engineers to give female engineers books of poetry by Leonard Cohen?

People get sued sometimes. Sometimes the lawsuits are meritless (although sometimes they aren’t). For example, sometimes people get sued by their neighbours, and maybe sometimes those lawsuits are meritless. This shouldn’t result in people regarding their neighbours as lawsuits waiting to happen, and I don’t believe that it does. I, for example, don’t regard my neighbours as lawsuits waiting to happen. Sometimes people get sued by employees, and sometimes by employers. This shouldn’t result in people regarding their employers or their employees as lawsuits waiting to happen, and I don’t believe that it does. If a person reads a report of a woman bringing suit over the conduct (or alleged conduct) of her employer and/or her co-workers, possibly a suit without merit, and it leads that person to regard every woman in the workplace as a lawsuit waiting to happen, it doesn’t mean that there is something wrong with the system, or with the law, or with women in the workplace; it means there is something wrong with that person.

Not a good rule of conduct is ‘Do not do anything that might possibly offend anybody else’, because anything anybody does or says might possibly offend somebody else. But a good rule of conduct is ‘Give some consideration to how your conduct might affect other people’. Not offending other people is not a goal that should automatically override every other consideration; but it is a factor that deserves some consideration. If people feel that being asked to give some degree of consideration to how their behaviour might offend other people is equivalent to being asked to walk on eggshells, then once again it is a sign that something is wrong with those people.

Your account of Ellen Pao’s lawsuit against Kleiner Perkins is grossly distorted, and I’ll go into detail about that if anybody’s interested, but it is of limited relevance at most; even if her lawsuit was as completely without merit as you represent it as being, that fact would not support any conclusion relevant to this discussion. There is, however, one aspect of the way you present the case which is an indicator that there is something wrong with you; I will explain that, also, if anybody’s interested.

131

Phil 08.16.17 at 9:17 am

Ian @124: generalisations about hardware engineers don’t tell us anything about software developers. Perhaps box-shifting and cable-laying do appeal to under-socialised males; people in those roles can get away with not having people-oriented skill-sets (just about). Software engineers absolutely can’t.

But say, for the sake of argument, that male software engineers are (typically) autistic-spectrum gearheads with no social interests or skills. If that is the case, there’s all the more reason to make an effort to recruit more software engineers who can actually talk to people, who will (ex hypothesi) mostly be women. Also, all the more reason to try and create a hospitable environment for those vitally important members of staff, by socialising the males already on the payroll, or at a minimum schooling them in the basics of social interaction – which, as parents of toddlers know, starts by laying down the law about what they can’t say.

In short, where you – and Damore – see the natural order of things, the rest of us see a problem that needs solving, and that companies like Google are (creditably) making an effort to solve.

132

sanbikinoraion 08.16.17 at 11:13 am

It should also be said that in most roles (and perhaps Google is atypical in this respect), that actually writing the program code is the easy part of software development. Figuring out what the right thing to program – by eliciting requirements from the business and juggling multiple competing demands – that’s the hard part. So whether males are better than females at math(s) and software is less relevant than you might think.

133

J-D 08.16.17 at 12:20 pm

Jeff Guinn
It is not immediately obvious how valid or meaningful it is to treat mechanical aptitude as a single trait susceptible of one-dimensional measurement; but supposing it means something, if men on average show more of it than women, one might wonder why that is so. On the face of it, it would appear to me that mechanical aptitude could easily be the sort of thing that is improved by practice; in which case, one might wonder whether men have more of it because they practise it more. But then, why might that be so? There would probably be a combination of reasons; but boys being encouraged to use their mechanical aptitude more than girls is one obvious possibility. Shouldn’t we therefore explore the possibility that if employers of mechanics gave more encouragement to women as employees and potential employees, they would develop greater mechanical aptitude and take up more jobs as mechanics? I can’t see how it makes sense for any kind of employer to discourage (intentionally or inadvertently) a section of the population from developing the abilities the employer needs.

134

Ian Maitland 08.16.17 at 3:52 pm

Phil @ 131.

Re hardware and software engineers, I admit that my ignorance is showing. I am biologically completely unsuited for tech. I remember my father’s despair when I failed to share his enthusiasm for the beauty of the motor of his Standard Vanguard (don’t ask). But my argument does not turn on such differences and associated skill sets.

Rather my point is that, say, for the sake of argument, you are right about the optimal mix of (ex hypothesi) gendered skills, then we can safely rely on market forces to weed out tech companies that are slow to see the truth. We don’t need to impose a solution (50-50?) on them.

Relying on market forces to produce the optimal gender ratios has an additional advantage. Consider the possibility that you are wrong. After all, tech companies have been stunningly successful despite the fact that their workforces have been predominantly composed of “autistic-spectrum gearheads with no social interests or skills.” (Brilliant!). If you are wrong, then “enlightened” tech firms will be weeded out, and benighted ones will probably move to China or India.

So — unless you believe that tech companies are insufficiently motivated by greed — over time their actual sex ratios (or mix of “gendered” skills) will change, with lags, to reflect the evolving challenges they face. To believe otherwise is to believe that we know their business better than they do. And, it goes without saying, it is always in the companies’ own interests to work to create a hospitable environment for all their employees.

Damore would be the last person to oppose policies to create a hospitable environment for women at Google. In fact, his memo offers a number of suggestions about how to do that. But that wasn’t what Google was doing. You give credit to Google where it isn’t due. Under pressure from left-wing media and feminist activists, Google caved in and — to quote Damore — created “a climate of shaming and misrepresentation [that] is disrespectful and unaccepting of anyone outside its echo chamber.” The company instituted mandatory “diversity training” whose premise is that “all disparities in representation are due to oppression” and that discrimination is necessary “to reach equal representation.” The diversity program indoctrinates employees in this ideology. And, as Damore’s firing shows, it is a career-limiting move for an employee to question it. (Ironically, as I write, Google is making essentially Damore’s argument before the courts in trying to fend off harassment by the Department of Labor).

I think you have misidentified the problem Google is trying to solve. Diversity programming has nothing to do with creating a welcoming workplace. Rather it is about throwing Google’s male employees under the bus to bolster the company’s legal position in ongoing litigation. (“Judge, it’s absurd to say we discriminate against women. Why, look at all the petty and gross humiliations we subject men to”). In the process Google has created a hostile working environment for both sexes and has inflamed mutual resentments and suspicions. It is no coincidence that Damore decided to write the memo after attending a Google diversity session where he heard things he “definitely disagreed with.” “There was a lot of just shaming and ‘no, you can’t say that, that’s sexist’ . . .”

135

Jeff Guinn 08.16.17 at 7:00 pm

[J-d:] On the face of it, it would appear to me that mechanical aptitude could easily be the sort of thing that is improved by practice; in which case, one might wonder whether men have more of it because they practice it more.

That’s certainly a valid supposition — assume it is true. In that case, as Western societies have become vastly better in terms of equal opportunity (40 years ago, women were actively excluded from my profession to being actively sought after, and that is just one example of many), then the mechanical aptitude difference between men and women
should converge.

It hasn’t budged.

Further, there are easily available indicators out there to suggest what women’s preferences are right now. There is a series on youtube called “Project Binky”. It is about putting a turbo charged Toyota Celica drivetrain in an original Mini. Viewership: something more than 100% male. Same for AvE — he does tool reviews as well as shop tips and tricks.

Women practically never, even now, choose to spend time on that stuff. Guys in their hundreds of thousands (nearly 800,000 for Binky Ep15) do.

And I’ll bet the numbers are pretty much the same for, say, Arduino (popular micro controller) videos.

This means that young women are almost exclusively not choosing to look at these things, which could very well mean that they are not at all inclined to do so. Further, despite growing up in a culture that imposes no explicit limits, they are no more inclined to do so than their mothers.

So while your supposition might be true, there is no available evidence suggesting it is true. Rather, it seems that, as with many other things, their are distinct and ineradicable population-wide differences between men and women.

Which is Damore’s point: across many characteristics, the aggregate differences between men and women are such that at any point on the right side of the male distribution, there will be far fewer women. Within the easily and thoroughly quantified realm of mechanical aptitude, there is no doubt of this, and no indication it is amenable to social determinism.

There is no reason to believe that once artificial constraints are lifted — as in my profession — that female representation will change from its constricted amount to 50-50. It could be anything. In some professions, veterinary medicine, for instance, it has gone from nearly zero to more than 60%. In mine — zero to five percent, hasn’t budged since, and doesn’t look like moving ever.

This is a long-ish read, but looks in detail at the issue.

This American Life — which will never be confused with the right — had a show called Testosterone. Turns out, it makes a difference.

Damore was on solid ground, he used well established data and a coherent logical argument to make the case that Google’s goal for women software engineers will be defeated by the combination of women’s aptitudes and inclinations. Absent grotesque abuse of EEO law, there is no way to get there.

136

Ogden Wernstrom 08.16.17 at 7:43 pm

Mainhole:

Notice two things:

1. Griswold’s blurring of sexual misconduct and sexism.

The unsupported implication is that there is a link between raping or sexual harassing a female colleague, on the one hand, and telling a joke that offends her, on the other.

The link that comes to mind is fear. Once I can picture how this looks to the female colleague, especially if it comes from one with power in the workplace, I think the common factor is fear. Fear that the joke reveals a misogynistic point-of-view…which will lead to the joke-teller trying to undermine the woman’s work, derail her career, maybe even thrust himself into her personal life. Fear that a sexual joke about women is a hint that she will be considered a Scarlet woman, while the Libertine men get a pass.

All men — or all men who have made a female colleague uncomfortable — are therefore (in some degree) guilty by association for the outrageous behavior of a few. I call that a smear.

…and that rhymes with “Fear”, not mere discomfort.

2. The stereotyping of male software engineers as sexists.

Griswold does not say that sexism is universal among male engineers in Silicon Valley, but she comes pretty close. Consider: The poor treatment of women in Silicon is a “feature, not a bug”;

That Jeff Bercovici sure does stereotype men, doesn’t he? He wasn’t always that way, but reporting on the tech industry appears to have turned him into a…what would we call it?

…and in Silicon Valley “sexism has been a long known problem.”

Is the argument here that male software engineers are named-in-proxy by “Silicon Valley”? Because that shrill bitch Jeff Bercovici has shown that the problem exists much higher up the food chain, and he’s not alone. In the hundreds of small companies I have consulted for over the years, I find I can predict a lot about the attitude/personality of the owner by spending a couple of hours with the workers.

(Remember that Damore was fired by Google because he allegedly stereotyped women; the charge was false).

Is this stating that Damore was stereotyped as someone who stereotypes?

The assertion “false” sounds like the standard attorney’s assertion, if a court were to hold a hearing on that issue.

I believe these unsupported attributions smear a lot of people.

Imagine unsupported attributions that smear half the population. C’mon, show Collin Street he’s wrong – I dare you to try.

137

Moz of Yarramulla 08.16.17 at 11:44 pm

Jeff Guinn@135: There is a series on youtube called “Project Binky”. It is about putting a turbo charged Toyota Celica drivetrain in an original Mini. Viewership: something more than 100% male. Same for AvE — he does tool reviews as well as shop tips and tricks.

I watch most of AvE’s videos, and he’s unashamedly sexist. His latest video is called “BOLTR: BEAVER FEVER RELIEVER” for crying out loud. Sure, he possibly would argue that he does it in a self-consciously postmodern ironic way. But that doesn’t fly, because he does it consistently and there’s no counterpoint. He does it because it’s funny and it gets clicks. Because his audience is male, and comfortable with sexual, sexist humor.

Saying “women aren’t interested in sexist humor aimed at them” is almost tautological.

138

Orange Watch 08.16.17 at 11:54 pm

Ian Maitland@134

So — unless you believe that tech companies are insufficiently motivated by greed — over time their actual sex ratios (or mix of “gendered” skills) will change, with lags, to reflect the evolving challenges they face. To believe otherwise is to believe that we know their business better than they do. And, it goes without saying, it is always in the companies’ own interests to work to create a hospitable environment for all their employees.

So basically, your subsequent “but not like THIS” equivocatingnot withstanding, what Google is doing. The way you treat the idea in this paragraph is reminiscent of a free-market zealot describing how market forces – unsullied by human desires or action – will optimize markets. The companies will not change without their management actively seeking to change them. Individual low-level managers, hiring or otherwise, will not spontaneously try to optimize the business environment, and entrenched cultural forces will not dissipate without a fight. The way things are now is beneficial to people within the existing hierarchy and its social networks; they will not universally or unbegrudgingly sacrifice influence, promotion potential, and/or authority for the good of the company.

Companies are not actual persons with minds of their own; they are comprised of executives, managers, and workers which ARE actual people rather than mindlessly loyal and altruistic automata. Change will not organically occur because “the company decided to”; the leadership will have to subvert and alter existing corporate culture, either by threatening, persuading, or cajoling the constituents of that culture to change. And there will be pushback.

If that sounded pedantic… you might want to reconsider what it’s responding to.

139

J-D 08.17.17 at 1:33 am

Ian Maitland

… we can safely rely on market forces …

I deny you the right to treat that statement as axiomatic. If you want to rely on it, you need to indicate why it should be accepted.

So — unless you believe that tech companies are insufficiently motivated by greed — over time their actual sex ratios (or mix of “gendered” skills) will change, with lags, to reflect the evolving challenges they face. To believe otherwise is to believe that we know their business better than they do.

If that logic is applied consistently, it must follow that for you (or James Damore) to criticise policies/programs Google is adopting/implementing (such as diversity training) is to believe that you know their business better than they do. If you think that Google is a better judge of what it should do than anybody else can be, on what basis can you possibly support James Damore’s criticism?

Damore would be the last person to oppose policies to create a hospitable environment for women at Google.

That seems like a large claim to me. I am not inclined to accept it solely on the basis of your assertion.

Under pressure from left-wing media and feminist activists, Google caved in and — to quote Damore — created “a climate of shaming and misrepresentation [that] is disrespectful and unaccepting of anyone outside its echo chamber.”

If it were true that Google had created a ‘climate of misrepresentation’, it would be possible to present examples of misrepresentations. So far you don’t seem to have done so.

The company instituted mandatory “diversity training” whose premise is that “all disparities in representation are due to oppression” and that discrimination is necessary “to reach equal representation.”

You’ve mentioned several times that the training is mandatory, but so far the reports I’ve found don’t support that. I’d like to know where you’re getting that from. I’d also like to know whether the statements ‘all disparities in representation are due to oppression’ and ‘discrimination is necessary to reach equal representation’ are direct quotes from the training, as you appear to be representing them as being.

Diversity programming has nothing to do with creating a welcoming workplace. Rather it is about throwing Google’s male employees under the bus to bolster the company’s legal position in ongoing litigation.

What litigation are you referring to?

It is no coincidence that Damore decided to write the memo after attending a Google diversity session where he heard things he “definitely disagreed with.” “There was a lot of just shaming and ‘no, you can’t say that, that’s sexist’ . . .”

If somebody tells me that something I have said is sexist, and it actually is sexist, then the appropriate response is not for me to complaint about having been shamed. If I actually have been sexist, then I should be ashamed of it.

On the other hand, it’s possible that somebody might tell me that something I have said is sexist when in fact it wasn’t. The appropriate response then depends on more specific facts. For one thing, it’s possible that what I said might have been sexist even though I didn’t intend it to be sexist and didn’t realise it was sexist; so one part of an appropriate response might be to reflect on whether what I said was sexist, despite my first impulse to deny that. Even if what I said was not actually sexist, one part of an appropriate response might be to reflect on what led people to perceive it as being sexist, and whether I can learn anything about how to express myself more clearly in future. I can imagine that there are some circumstances where part of an appropriate response would be to make a public defence of my statement, explaining how it isn’t sexist and how perception of it as sexist is mistaken. What I can’t imagine is any circumstance where it would be an appropriate response to complain publicly about having been shamed. What is it, specifically, that James Damore said, and how was it, specifically, that he was shamed for it?

140

Moz of Yarramulla 08.17.17 at 3:33 am

Jeff Guinn @ 127: Could you quote just one?

Nope. Haven’t read it, don’t intend to. I’ve seen quotes that struck me as enough as single excerpts to make me not want to read it. Or find those quotes again. You want them, you look for them.

Enough people have read it that I don’t feel the need. I’ve read The Bible, I’ve read the first bit of “Atlas Shrugged”, I’ve done my bit. When a whole bunch of people I respect say “this is offensive garbage” I take their word for it.

I realise it’s an attractive nuisance but that doesn’t mean I have to respect the people who revel in it, any more than I respect people who suffer from other forms of bigotry. Sexism in tech is one of those things where denial just marks the denier as weird and obsessive. At best “It is difficult to get a man to understand something, when his salary depends on his not understanding it”. But a lot of people are out there insisting that the best case doesn’t apply to them.

141

Jeff Guinn 08.17.17 at 6:16 am

[Moz:] I watch most of AvE’s videos, and he’s unashamedly sexist.

Fair enough.

But the point still stands. There are untold numbers of videos on youtube dealing with tools, engine rebuilding, woodworking, electronics, programming tutorials, etc. that aren’t the least bit sexist. Near as makes no difference to 100% of the people making and watching them are men.

That counts as powerful evidence that women’s interests do not lie in those activities nearly to the extent men’s do, thereby substantiating Damore’s point.

[Ogden:] Imagine unsupported attributions that smear half the population.

Could you please provide one, using direct quotes, and showing how it was unsupported, and constitutes a smear?

142

Jeff Guinn 08.17.17 at 6:46 am

[Moz:] Nope. Haven’t read it, don’t intend to.

That’s on you then, not Damore. I can’t help but notice that no one who has leveled accusations at Damore has ever quoted him directly, nor impeached his sources. Not at the NYT, Slate, Gizmodo, nor here. Google management didn’t, either.

Having read it, I think you would not be able to find anything in there that constitutes sexism, or smears half the population, or any of the other charges hurled at him.

143

J-D 08.17.17 at 7:06 am

Jeff Guinn
If we have a situation where many occupations used to be heavily gender-segregated, but in the last forty years there have been huge changes in some of those occupations, so that the gender-segregation has been sharply reduced or eliminated, while in other occupations there has been little or no change, how do we explain that situation?

It seems that your preferred explanation is that in some occupations the gender segregation was largely or wholly a consequence of unfair discrimination, while in others it was largely or wholly a consequence of innate differences between men and women, and that over the last forty years unfair gender discrimination in our society has been sharply reduced, with the result that gender-segregation has disappeared, or at least been sharply reduced, in those occupations where it was a consequence of unfair discrimination, but remains in those occupations where it is a consequence of innate differences between men and women.

But I don’t know of any good reason to prefer this explanation to the alternative explanation that what has happened in our society over the last forty years is that some forms of unfair gender discrimination have been sharply reduced (affecting gender segregation in some occupations) while other forms have been little affected (leaving gender segregation largely unaffected in some occupations).

I think there are few if any people who would deny that there are differences between the interests and preferences of men and women; what is heavily disputed is the extent (if any) to which those differences are innate. It’s possible that there are innate differences which will remain no matter what else changes in society; but it’s also possible that there are still huge differences which could be sharply reduced, perhaps to nothing, if society changed.

Notice that you write that young women grow up ‘in a culture that imposes no explicit limits’. Even if this is true, your inclusion of the qualifying word ‘explicit’ represents an acknowledgement that there could be such things as implicit limits. Notice also that you refer to ‘distinct and ineradicable population-wide differences between men and women’. The fact that you conjoin the two qualifying adjectives ‘distinct’ and ‘ineradicable’ represents an acknowledgement that they are not synonymous, that it is possible there are distinct population-wide differences that are not ineradicable.

144

wierdnoise 08.17.17 at 8:55 am

Figures that Ian would cherry-pick one thing out of my comment and say absolutely nothing about the rest of it. Of course my experience is valueless, while his circular theorizing is somehow authoritative.

I was a lurker here for several years, and came back as a result of a link to this post, which I must thank Daniel for writing. But I now recall why I generally stay away these days. You folks are much too accommodating of folks like Ian, who keeps repeating his points over and over and over. It’s fine to have a variety of views — that’s what I valued CT for. But the haughty petulance of some of the most frequent commenters leaves a bad smell.

145

J-D 08.17.17 at 12:39 pm

Jeff Guinn

I can’t help but notice that no one who has leveled accusations at Damore has ever quoted him directly, nor impeached his sources. Not at the NYT, Slate, Gizmodo, nor here.

In my comments in this discussion so far, I have been responding to the contributions of other commenters, not to anything James Damore wrote. But I have now looked at what he wrote, and I notice several dubious points. For example (and I quote directly):

As soon as we start to moralize an issue, we stop thinking about it in terms of costs and benefits

If we think about an issue in terms of costs and benefits, we are implicitly moralising; what we count as a cost, and what we count as a benefit, is a moral choice.

Viewpoint diversity is arguably the most important type of diversity and political orientation is one of the most fundamental and significant ways in which people view things differently.

That’s not a claim that deserves to be accepted without argument.

We should focus on psychological safety, which has shown positive effects and should (hopefully) not lead to unfair discrimination.

That’s another claim that does not deserve to be accepted without argument. For one thing, it’s not obvious what is meant in this context by ‘psychological safety’.

146

Phil 08.17.17 at 1:24 pm

Sing it, wierdnoise. There’s just nothing there to reply to – Damore’s wrong about the science, wrong about the nature of the profession and wrong about the problem. By rights the OP should have been followed up with nothing but links and M3 T00s – not an endless stream of ah, but if you ignore this and redefine that, and hey, what about this other thing…. Hard to comment-police for blowhards, though.

147

hellblazer 08.17.17 at 2:04 pm

From the previous comment(er):

I was a lurker here for several years… But I now recall why I generally stay away these days… the haughty petulance of some of the most frequent commenters leaves a bad smell.

Seconded.

148

Jeff Guinn 08.17.17 at 2:25 pm

[JD:] It seems that your preferred explanation is that in some occupations the gender segregation was largely or wholly a consequence of unfair discrimination, while in others it was largely or wholly a consequence of innate differences between men and women, and that over the last forty years unfair gender discrimination in our society has been sharply reduced, with the result that gender-segregation has disappeared, or at least been sharply reduced, in those occupations where it was a consequence of unfair discrimination, but remains in those occupations where it is a consequence of innate differences between men and women.

Perhaps you should have quoted me directly.

Let me reiterate: until 40 or so years ago, women were effectively barred from all occupations that weren’t nurse, teacher, secretary. Then, in very short order, essentially all formal barriers to entry vanished. Subsequently, women’s participation in some occupations skyrocketed, while others barely budged.

Those occupations with skyrocketing female participation — veterinarian, family practice doctor, lawyer — all involve characteristics in which women, in aggregate, are far better than men.

But occupations with different characteristics — asocial, mechanical, systematizing — saw hardly any increase, and what little there was happened very quickly, and hasn’t changed since.

My profession is high in both status and pay. There are no physical barriers to female participation. Entry is down solely to desire, and success in training his highly correlated with mechanical aptitude, 3D visualization, and inclination towards systematizing. The work environment is extremely asocial, requires frequent moves and nearly non-stop travel.

That is why, in the aggregate far fewer women will be inclined to become airline pilots than men. Once the artificial constraints were removed, only the natural ones remained. There is an entire generation of women who have grown up absent those constraints, and with a great deal of explicit encouragement towards taking on non-traditional occupations.

My company has hired four hundred pilots in the last year. Fifteen were women. That ratio hasn’t changed in more than twenty years.

Why? Because contrary to your assertion, it isn’t highly disputed these differences are innate. In fact, quite the opposite.

Therefore, absent simply gainsaying reality, there exists some occupations that women simply choose not to do. Equality of opportunity is not synonymous with equality of outcome.

The fact that you conjoin the two qualifying adjectives ‘distinct’ and ‘ineradicable’ represents an acknowledgement that they are not synonymous …

That is not true. Distinct means that there is a significant difference in distribution. In aggregate, female mechanical aptitude is very nearly one standard deviation below male: this is a very distinct difference. It is almost certainly ineradicable because it hasn’t converged over time, is independent of culture, and gives every appearance of being innate.

Two entirely different concepts.

149

Ian Maitland 08.17.17 at 3:45 pm

Orange @ 138

That is wrong. It is beside the point what upper management or lower management wants.

Recall that I was responding to Phil’s scenario: “But say, for the sake of argument, that male software engineers are (typically) autistic-spectrum gearheads with no social interests or skills. If that is the case, there’s all the more reason to make an effort to recruit more software engineers who can actually talk to people, who will (ex hypothesi) mostly be women. “

In that environment (i.e., where social skills are needed), it follows that companies that don’t recruit more “software engineers who can actually talk to people” (ex hypothesi mostly women) will find themselves at a competitive disadvantage as compared with companies that do recruit such software engineers. They will lose market share and eventually go out of business. That will leave only companies that have something close to an optimal balance of gearheads (men) and software engineers with social skills (women).

Might individual low-level managers try to sabotage a change in the recruiting mix? Of course it’s conceivable, but then they had better be on the verge of retirement, because they will be putting themselves out of a job.

In that case, we don’t need your leftist-feminist conspiracy theory about pushback from “entrenched cultural forces.” The entrenched cultural forces can fight all they want, but they are history.

150

SusanC 08.17.17 at 4:46 pm

@Orange Watch:

I was going to mention Carnegie Mellon. As I understand it, they’ve recently had a push on recruiting more female CS undergraduates, which was reasonably successful. (There’s a question to be investigated about whether they’ve increased the total number of female CS undergraduates, or just persuaded the women who already wanted to study CS that Carnegie Mellon is the best place to do it). Which makes them an interesting experimental case for this kind of discussion.

151

Ogden Wernstrom 08.17.17 at 5:07 pm

@134 appears to be an argument based upon the magical thinking that markets can will cure all ills. I can’t expect an expert on Gender Differences to also have read Economics and understand market failures, or the tendency of high-rent, near-monopoly businesses to control a market (versus being-controlled-by-it).

Nor can I expect the sort of ideologue who was the first to be named a Senior Fellow at the Center of the American Experiment to concede that Google’s firing of Damore is a market response.

Damore, slapped down by The Invisible Hand.

152

Ogden Wernstrom 08.17.17 at 5:17 pm

@141 Jeff Guinn 08.17.17 at 6:16 am:

[Ogden:] Imagine unsupported attributions that smear half the population.

Could you please provide one, using direct quotes, and showing how it was unsupported, and constitutes a smear?

No, I can not provide you with an imagination. [Cue Collin Street….]

If you can’t start imagining a tactic beyond, “Prove it, using direct quotes”, I figure you will be asked to do the same for your assertions – so you can be kept busy with the same delaying tactic you are using. And your work can then be given the regard it merits.

Keep in mind that data which appears to show something inherent about women could be proving something about society. Also keep in mind that America-centric data that appears to show something inherent about women could easily be proving something about America.

Having read [Damore’s manifesto], I think you would not be able to find anything in there that constitutes sexism, or smears half the population, or any of the other charges hurled at him.

For example, you might be asked, “Please provide a direct quote which charges Damore with smearing half the population”.

153

Ian Maitland 08.17.17 at 5:48 pm

J-D @ 130, 139

I apologize if I appear to have neglected your comments. I had to think hard about them.

You’ll hate me saying this, but I find a lot to agree with in your post. I am not proposing anything like abolishing lawsuits, but I think we need something the English rule where the loser pays, among other overdue reforms. I would also be interested in your take on the Pao case. If you don’t want to share that here, you are welcome to email me. I’ve already been outed on this blog. I teach business at the Univ. of Minnesota. You shouldn’t have too much trouble finding my email address.

Having said that, I believe that the court system has been abused, particularly in the case of gender class actions over alleged pay discrimination at colleges. The problem is less with the court system itself than with college administrations. I don’t want to generalize, but very few of them are willing to let lawsuits brought against their colleges go to trial. They are paralyzed by the prospect of negative publicity and terrorized by fear of being labeled misogynists, and anyway they are PC and even if they are convinced that the particular lawsuit against their college has no merit, who could oppose hiking women’s pay? Consider the pay hikes to be some sort of righting of a historic injustice. So the administrations move quickly to settle the lawsuits, usually giving women generous “equity” pay increases, without any serious regard to the merits of the claim.

Of course, this process feeds on itself. Once the plaintiff’s bar figures out that colleges won’t fight women faculty’s pay equity claims, but will always offer settlements, then no matter how unfounded the claims are, a lawsuit is a one-way bet. They can’t lose. These lawsuits have greatly enriched the plaintiff’s bar. (I am speaking from first-hand experience).

Re your # 139

1. My original comments first: “we can safely rely on market forces …”
2. You: I deny you the right to treat that statement as axiomatic. If you want to rely on it, you need to indicate why it should be accepted].
3. Reply: I spelled out the economic logic. I think the burden is on you to give some indication of why the laws of economics may be suspended in this case.

1. So — unless you believe that tech companies are insufficiently motivated by greed — over time their actual sex ratios (or mix of “gendered” skills) will change, with lags, to reflect the evolving challenges they face. To believe otherwise is to believe that we know their business better than they do.
2. If that logic is applied consistently, it must follow that for you (or James Damore) to criticise policies/programs Google is adopting/implementing (such as diversity training) is to believe that you know their business better than they do. If you think that Google is a better judge of what it should do than anybody else can be, on what basis can you possibly support James Damore’s criticism?
3. My contention is that Google’s policies have been political decisions, not business decisions. Companies should not be placed in the impossible position of having to adopt bad policies in order to placate political and media bullies with their own political agendas.

1. Damore would be the last person to oppose policies to create a hospitable environment for women at Google.
2. That seems like a large claim to me. I am not inclined to accept it solely on the basis of your assertion.
3. As Jeff Guinn says to Moz, if you haven’t read Damore’s memo, “that’s on you then.” Don’t accept my assertion if you want, but (and I address this to almost all of you) please read Damore’s memo. That is where you will find the supporting evidence.1.

1. Under pressure from left-wing media and feminist activists, Google caved in and — to quote Damore — created “a climate of shaming and misrepresentation [that] is disrespectful and unaccepting of anyone outside its echo chamber.”
2. If it were true that Google had created a ‘climate of misrepresentation’, it would be possible to present examples of misrepresentations. So far you don’t seem to have done so.
3. That is not fair. For one thing, that is a quotation from Damore. But, if you want specifics, take the biggest lie of all, one that has girdled the globe billions of times, and appeared in thousands of media outlets (including the NYT, the Economist, the Guardian). It is Google CEO Sundar Pichai’s lie about Damore’s memo: “To suggest a group of our colleagues have traits that make them less biologically suited to that work is offensive and not OK.” For the quote see: https://www.recode.net/2017/8/7/16110696/firing-google-ceo-employee-penned-controversial-memo-on-women-has-violated-its-code-of-conduct
That is not what Damore said, and it is not what he meant. But it is what the media reported. But Damore’s detractors seem to fear that if they actually read the memo they will somehow become ideologically contaminated. Read the memo, please. (For one case, see Claire Cain Miller’s column in the NYT. The lie lives on in the archives because she refused to react despite numerous protests).

1. The company instituted mandatory “diversity training” whose premise is that “all disparities in representation are due to oppression” and that discrimination is necessary “to reach equal representation.”
2. You’ve mentioned several times that the training is mandatory, but so far the reports I’ve found don’t support that. I’d like to know where you’re getting that from. I’d also like to know whether the statements ‘all disparities in representation are due to oppression’ and ‘discrimination is necessary to reach equal representation’ are direct quotes from the training, as you appear to be representing them as being.
3. No, the statements are directly from Damore’s memo — the one that ignited this controversy. (As for mandatory, are you suggesting that they can be skipped with no consequences? I will try to verify the claim).

1. Diversity programming has nothing to do with creating a welcoming workplace. Rather it is about throwing Google’s male employees under the bus to bolster the company’s legal position in ongoing litigation.
2. What litigation are you referring to?
3. There are pending sex discrimination lawsuits in the courts. I believe they (or it) were filed recently. The Department of Justice is also in the courts demanding a huge data jump. (You see how the iron triangle of activists, media and government work?).

1. It is no coincidence that Damore decided to write the memo after attending a Google diversity session where he heard things he “definitely disagreed with.” “There was a lot of just shaming and ‘no, you can’t say that, that’s sexist’ . . .”
2. If somebody tells me that something I have said is sexist, and it actually is sexist, then the appropriate response is not for me to complaint about having been shamed. If I actually have been sexist, then I should be ashamed of it.
3. That is a big “if.” I am quoting Damore. He obviously didn’t believe it was sexist. If you have read up on campus speech codes, you will know how perfectly innocuous words have been proscribed.

1. It is no coincidence that Damore decided to write the memo after attending a Google diversity session where he heard things he “definitely disagreed with.” “There was a lot of just shaming and ‘no, you can’t say that, that’s sexist’ . . .” BIS
2. On the other hand, it’s possible that somebody might tell me that something I have said is sexist when in fact it wasn’t. The appropriate response then depends on more specific facts. For one thing, it’s possible that what I said might have been sexist even though I didn’t intend it to be sexist and didn’t realise it was sexist; so one part of an appropriate response might be to reflect on whether what I said was sexist, despite my first impulse to deny that. Even if what I said was not actually sexist, one part of an appropriate response might be to reflect on what led people to perceive it as being sexist, and whether I can learn anything about how to express myself more clearly in future. I can imagine that there are some circumstances where part of an appropriate response would be to make a public defence of my statement, explaining how it isn’t sexist and how perception of it as sexist is mistaken. What I can’t imagine is any circumstance where it would be an appropriate response to complain publicly about having been shamed. What is it, specifically, that James Damore said, and how was it, specifically, that he was shamed for it?
3. Yes, and when will they find time to write code? If you read Damore’s memo, you will see what he said, and you already know the consequences. As for your claim that you “can’t imagine is any circumstance where it would be an appropriate response to complain publicly about having been shamed,” I am horrified. I think you and your conscience need to have a serious conversation. (Can you imagine “any” circumstance where it is appropriate for a woman to complain publicly about having been discriminated against?).

154

Jeff Guinn 08.17.17 at 8:07 pm

[Phil:] Damore’s wrong about the science …

That was worth a spit take. Science is a process, not a conclusion. Science is a one word way of saying “rational inquiry.” If someone was to say to you “the rational inquiry is settled”, you’d look at them as if they had three heads.

But wait, there’s more. The rest of the article is a peon to how unsettled some conclusions are from rational inquiry. Damore isn’t wrong about the conclusions, because in some regards, the scales do not conclusively tip one way or the other. From the article:

The problem is, the science in Damore’s memo is still very much in play, and his analysis of its implications is at best politically naive and at worst dangerous.

Yep, nothing but science there.

But wait, there’s more more:

In general, he notes, women prefer to work with people and men prefer to work with things—the implication being that Google is a more thing-oriented workplace, so it just makes sense that fewer women would want to work there. Again, the central assertion here is fairly uncontroversial. “On average—and I emphasize that, on average—men are more interested in thing-oriented occupations and fields, and that difference is actually quite large,” says Richard Lippa, a psychologist at Cal State Fullerton and another of the researchers who Damore cites.

But trying to use that data to explain gender disparities in the workplace is irrelevant at best. “I would assume that women in technical positions at Google are more thing-oriented than the average woman,” Lippa says. “But then an interesting question is, are they more thing-oriented than the average male Google employee? I don’t know the answer to that.”

Emphasis added. This article cited a person who utterly clueless. Damore’s arguments were explicitly aggregate; he emphasized, although apparently not loudly enough, that what is true in the aggregate tells you nothing about the individual. Lippa, despite it being obvious enough, can’t fathom the point: within a representative population, there will be substantially more “thing oriented” men than women; therefore, to the extent “thing orientation” is useful, there will be far more men reaching some degree of “thing orientation” then women.

Anyone care to dispute that?

Remember, Damore’s contention was that Google had effectively imposed a determination of what reality must be, despite determination being, at best, on very infirm ground.

Damore provides a firmly grounded, well reasoned, argument why that determination is very likely wrong.

Then Google proved Damore’s point by firing him.

155

Collin Street 08.17.17 at 9:04 pm

The thing, of coutse, is
+ nobofy realises that they are in error (“I belueve X” and “I believe I am correct to believe X” are the same damned statement, not independent verification of each other)
+ a fortiori, a person’s conceptual errors are believed by that person to be correct
+ which means line-by-line refutation is largely pointless: if a person is acting in good faith, then every thing they write is genuinely believed.

I mean, how’s it supposed to work? What’s the mechanism of correction here, how is having your mistakes pointed out supposed to help you see that they are mistakes?

156

Kaleberg 08.18.17 at 12:26 am

One of the reasons fewer women go into computer software is because they know they’ll have to deal with assholes like Damore, armies of them. They’d even most likely have to work for one of them. Even for women with natural programming ability, this can be demoralizing. Who wants to work in a field where everyone thinks you are a moron and makes sure that you know it? It’s not as if there is a real metric for how much work you’ve gotten done.

Google has a problem with this. Those stupid Google hiring interview questions were a dumb ass idea. Anyone who has recruited programmers know that a lot of good programmers like stupid games – the ones you can look answers for on stackexchange or quora – while an awful lot of even better ones think that stupid games are stupid games. It’s the math team versus the real mathematicians. It’s about two different skills, solving math problems quickly and on the fly and applying or making major advances in mathematics. Google hired a lot of the math team types. They should have looked for classical music freaks instead. They’d have gotten better all around programmers.

157

Ian Maitland 08.18.17 at 3:45 am

wierdnoise @ 107

W-“I have noticed some differences between women and men software engineers. Like Yonatan observes, the women seem on average better at the whole mutual support, teamwork, and yes, political aspects of software engineering…. Software engineering isn’t like most other engineering disciplines, and in fact many would refuse to call it engineering at all. There is a lot more informational communication, and a lot less math and detailed specifications than, say, electrical or mechanical engineering (something that may be entirely contrary to what folks outside of the profession might assume). The social aspect is highly important; frequent discussion of design and progress and, yes, empathy in the sense of understanding a coworker’s thinking. And although I cast a skeptical eye toward arguments from biology, the fact is that women are expected by our culture to do better in these domains, and thus often do so. Make of that what you will.”

I- I don’t find anything to quarrel with here. In fact, I am ecstatic that, unlike other critics of Damore, you are not a biology-denier. I’ll settle for a biology-skeptic because that is a major advance over the dogmatic attitude of so many other commenters.

W-It amuses me, in a sad way, that there has been a lot of effort to engineer the social and communication environment of software development, techniques such as “agile programming” — and yet the brogrammer “hold my beer while I write some spectacular lines of code” ethos still predominates.

I-When I saw this, I thought it was a cheap shot, if a fairly harmless one. I still think so. Two things bothered me. First, I thought you were giving aid and comfort to some of the media attack pieces that suggest that tech workplaces are overrun with beer-swilling brogrammers who create a hostile work environment for women. Second, there is the double standard. Is it possible for there to be a predominant ethos without a majority or at least a plurality of the class sharing it? You are characterizing real people here — the “ethos” is not just an abstraction that hovers over Google or is pumped in through the air ducts. What is sauce for the gander should be sauce for the goose. I tried to imagine the furor you would have provoked if you had used an equally derogatory stereotype of women programmers.

W-One final comment: that screed circulated within Google for weeks before its existence was leaked to the public. The result was a textbook case of a hostile working environment for the women there. Shame on Google for letting it go on so long.”

I-I am probably going to repeat myself, but here goes. Of course I think your view is vile. To show why, let me oversimplify. The underlying issue is why women are “underrepresented” (only a lawyer could have coined that barbaric neologism) in Google’s workforce. There are broadly two theories. 1. The underrepresentation is the result of sexism. 2. The underrepresentation is the result of the career choices voluntarily made by women and men (which presumably reflect the different distributions of their interests).

Now, it is possible that the answer lies somewhere between the two theories. That is Damore’s view. He begins his memo by admitting that sexism may play a role. (He was torched because he dared to mention the B word that is anathema to feminist thought police).

Think for a moment about what accepting the first theory means. Again, sexism isn’t just something in the ether of the workplace. There can’t be any sexism without sexists. And that is where Damore and other male employees come in. They are the designated fall guys. (Who else?). If the low number of women at Google is a national scandal, it is THEIR fault. That is exactly the point of diversity sessions — that is where male employees are “re-educated” and brainwashed to purge them of all those nasty male toxins that make them discriminate against women.

It is no accident that that is precisely the stereotype that you will find endlessly propagated in the attack media whose goal it is, as I said in an earlier post, to force gender equality on tech firms regardless of what real men and women want. It is quite usual for the writers to heap scorn on the beer-swilling frat boys who are supposed to populate tech firms (SEE http://theweek.com/articles/476014/brogramming-disturbing-rise-frat-culture-silicon-valley). That may be why your caricature hit a raw nerve. (“Oh, Wierdnoise is one of THEM”).

Let’s imagine Damore’s reaction. He is stung by these unwarranted attacks. They don’t fit anything he has seen at Google. He knows that less than 18% of computer science graduates are women. So how can he be blamed for the small number of women at the company? To make things worse, there are the degrading and humiliating diversity sessions at which it is forbidden to question the premise that the underrepresentation of women at Google is due to the sexism of its male employees.

And you say that his speaking up in his own defense, in defense of his male co-workers, and for the truth is “a textbook case of a hostile working environment for the women there.” That is worse than vile. It is evil. Repent!

158

TM 08.18.17 at 8:44 am

Just finished reading Siri Hustvedt’s Blazing World . Fantastic read, highly relevant to this thread.

159

Mario 08.18.17 at 8:44 am

It’s so good that people focus on equality issues at Google. As far as I know, at Amazon’s warehouses everybody is treated equally.

160

Orange Watch 08.19.17 at 3:14 am

IM@149:

In that environment (i.e., where social skills are needed), it follows that companies that don’t recruit more “software engineers who can actually talk to people” (ex hypothesi mostly women) will find themselves at a competitive disadvantage as compared with companies that do recruit such software engineers. They will lose market share and eventually go out of business. That will leave only companies that have something close to an optimal balance of gearheads (men) and software engineers with social skills (women).

Again. How would this look different than what Google is doing? What measures would executives in a company undertake IOT effect these changes? You have repeatedly asserted that if there could be a business advantage in increasing the number of female SEs, companies would do so. What would that look like??? You say the desires of employees that support the status quo ante would be run over roughshod by the desires of “the company”. What would that look like???

You are indeed repeating yourself, and what you’re repeating is magical thinking – invisible hand nonsense where business practices change without any human actions being undertaken to cause them. Alternately, I suppose you might be asserting that if change was advantageous, it would have already occurred. Which obviously is no less nonsensical. But I don’t see a third option from what you’ve said. Either responses to market forces occur without human intervention, or they always and forever happen immediately if not sooner. Any other explanation of why the invisible hand hasn’t already fixed this would require you to actually demonstrate that it’s a non-problem rather than airily citing the infallible wisdom of inhuman capitalist divinities reigning over the marketplace.

161

mclaren 08.19.17 at 4:53 am

Clever argument which misses the larger point that available evidence shows quality of programs does not depend on star programmers but on the quality of the system+team that produces those programs. Producing good teams is the goal, or should be.

Turns out that hiring superstar programmers isn’t the key to producing good teams.
What is? Social intelligence and emotional sensitivity:

“…On the good teams, members spoke in roughly the same proportion, a phenomenon the researchers referred to as `equality in distribution of conversational turn-taking.’ On some teams, everyone spoke during each task; on others, leadership shifted among teammates from assignment to assignment. But in each case, by the end of the day, everyone had spoken roughly the same amount. ‘‘As long as everyone got a chance to talk, the team did well,’’ Woolley said. ‘‘But if only one person or a small group spoke all the time, the collective intelligence declined.’’

“Second, the good teams all had high ‘‘average social sensitivity’’ — a fancy way of saying they were skilled at intuiting how others felt based on their tone of voice, their expressions and other nonverbal cues. One of the easiest ways to gauge social sensitivity is to show someone photos of people’s eyes and ask him or her to describe what the people are thinking or feeling — an exam known as the Reading the Mind in the Eyes test. People on the more successful teams in Woolley’s experiment scored above average on the Reading the Mind in the Eyes test. They seemed to know when someone was feeling upset or left out. People on the ineffective teams, in contrast, scored below average. They seemed, as a group, to have less sensitivity toward their colleagues.”
https://www.nytimes.com/2016/02/28/magazine/what-google-learned-from-its-quest-to-build-the-perfect-team.html

So, as usual, the available research shows that emotional rather than intellectual qualities prove far more important to building successful teams and actually getting the job of producing high-quality programs done. Standard, typical, usual and quotidian. All the emphasis on intelligence and superstar programmers is wasted, all the effort to identify the very best people and gather them together proves not only irrelevant but counterproductive.

But of course neuroscientist Antonio Damasio had already summarized the overwhelming evidence that intellectual capacity is much less important in assessing the ability to accomplish difficult tasks than evaluating the person’s emotional qualities. See Damasio’s book Descarte’s Error.

As you’d expect, NASA — which has to produce mission-critical software where failure can mean death — also rejects the “star programmer” delusion:

“There’s a reasonable degree of trust when entering the workforce due to the federal background investigations you’re subjected to. It is also assumed, however, that everyone has something valuable to offer, and we trust the potential of the individual. Show that you’re good and the title becomes almost meaningless. It’s not uncommon for an intern to contribute to high-profile NASA software projects. “
https://mystudentvoices.com/a-look-into-nasas-coding-philosophy-b747957c7f8a

We now return you to the regularly scheduled rating and ranking of people in organizations, which Campbell’s Law AKA Goodhart’s Law has shown to be destructive, futile, and toxic.

162

Jeff Guinn 08.19.17 at 6:17 am

More evidence in Damore’s favor:

During the past week, tech and education writers have been talking about data showing that the gender gap in the tech world is evident even in high school.

They cite Barbara Ericson, director of computing outreach at Georgia Tech, who recently broke down the 2013 Advanced Placement exams and found that in three states (Mississippi, Montana, and Wyoming), zero girls took the AP computer science test. Even in the state where girls are best represented among computer science test-takers, Tennessee, girls only took 29 percent of the exams.

However, even those statistics don’t quite show the striking degree to which girls are underrepresented among test-takers not only in computer science but in nearly all STEM fields. U.S. News broke down the numbers and found not only that boys outnumber girls by more than 4 to 1 among computer science test-takers, but by more than 2.5 to 1 on Physics C tests, which test specialized fields of physics. Boys also outnumber girls by nearly 2 to 1 among test-takers in the more general Physics B, and by nearly 1.5 to 1 on the Calculus BC exam.

In that same year, 2.1 million girls took AP tests, compared to 1.7M boys.

The rates have remained essentially unchanged over the last 20 years:

At stake in the experience of women as undergraduates is the makeup of the technology-creating population. While girls and women may be using the Internet for communication and the web for information retrieval, it is predominantly men (the majority in the U.S. being white and Asian-American) who are programming the computers, designing and fixing the systems, and inventing the technology that will affect all aspects of our lives (National Science Board 1998). Only 15-20% of undergraduate computer science majors at leading U.S. departments and only 17% of the high school computer science Advanced Placement test takers are female (College Board 1998; Kozen & Zweben 1998)

Despite being 20 years old, that link is well worth reading. Among other things, it describes a very different interest in computers between boys and girls. Boys were far more focused on computers, and spent much more of their time with them; girls had other interests. At Carnegie Mellon, 40% of male CS freshmen tested out of the intro class; no women did.

Yes, I know, an anecdote is not data, but: my son easily passed the CS AP exam, despite his school offered no CS classes.

Comments on this entry are closed.