Is this argument any more rigorous?

1. There is a non-50% representation between men and women in tech.

2. Unjust discrimination and societal pressures are the only possible explanation for this imbalance. Therefore we must intervene until it is 50%.

There is no doubt that unjust discrimination and societal pressures exist, and that we should fight these so that anyone with talent and interest can succeed in this field.

However, if 100% of these pressures were removed, would we have exactly 50% representation of men and women in tech? Many people appear to believe very strongly that this is true. What is the basis for this belief? Are we willing to entertain the possibility that, even without these pressures, we would have less than 50% representation of women in tech?

Instead here are some of the issues Google seeks to address which seem plenty reasonable to me.

"Our goal is to create an environment in which every Googler can thrive. We check and recheck our people processes (including promotions, compensation, and more) to ensure fairness and equity in all things that impact Googlers. For example, we took action when we saw that women in tech were less likely to self-nominate for promotions. We’ve also long had gender pay equity in our workforce and we recently shared our approach to compensation with the hope that other organizations will adopt similar fair pay practices." https://www.google.com/diversity/at-google.html

https://www.theatlantic.com/magazine/archive/2017/04/why-is-...

> Well, not quotas. You can’t say quotas. At least not in the United States. In some European countries, like Norway, real, actual quotas—for example, a rule saying that 40 percent of a public company’s board members must be female—have worked well; qualified women have been found and the Earth has continued turning. However, in the U.S., hiring quotas are illegal. “We never use the word quota at Intel,” says Danielle Brown, the company’s chief diversity and inclusion officer. Rather, Intel set extremely firm hiring goals. For 2015, it wanted 40 percent of hires to be female or underrepresented minorities.

> But since it began linking bonuses to diversity hiring, Intel has met or exceeded its goals. In 2015, 43 percent of new hires were women and underrepresented minorities, three percentage points above its target. Last year, it upped its goal to 45 percent of new hires, and met it.

https://www.intel.com/content/www/us/en/diversity/diversity-...

> US female new hires 35%

For example, from your quote: "For example, we took action when we saw that women in tech were less likely to self-nominate for promotions."

This seems to presume that, in a just world, women would self-nominate at exactly the same rate as men, and that we should intervene until we achieve that world.

I agree 100% with this: "I just hope that women know it's open to them and feel comfortable joining the field. I don't care about convincing uninterested people." I agree with that both for entering the field itself, as well as for other aspects like getting promoted.

Almost every diversity activist: https://twitter.com/sehurlburt/status/894409695057764352

Women being underrepresented in tech is usually cited as an example of sexism in the system. The implication is that there needs to be equal representation for there to be no bias.

We could have an even higher percentage of women, like we do in many top-notch classical outfits today. There is some history here I'm qualified to speak about as a woman who took many orchestral auditions in classical music. Most of decision makers in classical back in the day asked exactly what you are asking about tech right now. The results were stark. (see study below) My example is not intended to imply that tech is "the same" as classical music or anything--just pointing out a strong similarity in today's tech interviews to the gender/race-biased audition practices of orchestras in the pre-1970s, when it was believed that women/brown candidates were not as strong/accomplished/technically skilled/artistic/confident as their white European male colleagues. Here is a 1974 NYT article about "the change to screened auditions"; the reporter still felt comfortable calling Asian players "Oriental" at that time.

http://www.nytimes.com/1974/10/15/archives/orchestra-auditio...

The Met Orchestra is considered the overall "best orchestra" in the United States. It is a dream job in terms of pay and quality of music making. By best, I mean that winning a job at the met is like being a star player in the majors, more so than in any other orchestra. The Met's auditions are considered by classical players to be the most impartial, although they are still partial. They are the blindest, anyway, in terms of using a screen, if players can make it past the tape round. (that is another issue altogether for another time) Today, more than half of the NY Phil and 50% of the Met orchestra players are women, due to blinder audition practices. Still, most orchestras remove the screen after the first or second round today. (?) I believe the MET waits until the finals to remove the screen. (I didn't make it to the finals, but it is what I hear from players who have)

I believe that tech could learn a great deal from classical music's hiring practice innovations of the 1970s and 80s. Today, most US Orchestras are comprised of at least 50% women, up from none or just a couple in the 1970s. Today, St. Louis is majority female. If anyone in tech is interested, they could dig into this study published in The American Economic Review

http://www.academicroom.com/article/orchestrating-impartiali...

Blinding the auditions is brilliant, because it isn't a "bias for bias" approach. It doesn't attempt to combat one bias by intentionally introducing another bias. It just removes the jury's ability to be biased, by concealing the musician's sex.

Because of this, the results of that study are nearly unimpeachable. It's basically a perfect approach. If tech could do anything comparable to that, I would be 100% in favor.

Why aren't there totally blind coding interviews? I just don't get it! The science (I've read but can't remember the source now- if you know it, please share) says that in-person interviews are not helpful in helping to identify successful candidates, and are often harmful.

http://blog.interviewing.io/we-built-voice-modulation-to-mas...

"After running the experiment, we ended up with some rather surprising results. Contrary to what we expected (and probably contrary to what you expected as well!), masking gender had no effect on interview performance with respect to any of the scoring criteria (would advance to next round, technical ability, problem solving ability). If anything, we started to notice some trends in the opposite direction of what we expected: for technical ability, it appeared that men who were modulated to sound like women did a bit better than unmodulated men and that women who were modulated to sound like men did a bit worse than unmodulated women. Though these trends weren’t statistically significant, I am mentioning them because they were unexpected..."

I get your point that there still could be biases if you can recognize a musician's sound. I should have been more clear about what I meant by "perfect". What I meant is just that this intervention (adding screens to auditions) almost certainly didn't create any bias, it only removed it. It may not have removed 100% of it, but it didn't combat the bias by introducing a counter-bias.

Can you recognize a coder's work by their style? Definitely to some extent. One aspect of working on a shared code-base with a whole bunch of engineers is that uniformity of style is a significant plus (makes code easier to read when it's consistent), so many of the differences that might show a "personal flare" are removed. But still, there is plenty that could belie a person's coding style.

I think there are way too many programmers out there, however, for this to be an issue in interviews. Estimates say there are something like 2-4 million software engineers in the US. Compare this with orchestral musicians: it looks like about 60 orchestras in the US work 40+ weeks per year; if we estimate 100 players per orchestra this is 6,000 full-time orchestra spots in the USA (you might have better numbers on this). If you specialize this by instrument, it might be more like 240 full-time orchestral oboists in the whole country. It seems much more likely that you could recognize a specific person with those odds.

(By the way, I'm a musician too. :) Though not in the orchestral world, and not full-time. I'm a singer (small ensembles mainly), though I did a lot of collaborations with orchestras as a boy growing up in Boychoir).

I understood/agree with you about how you spoke about our audition removing bias without introducing a counter-bias. There is no way to be perfect here, but the orchestras who implement this screened protocol are some of the top 10 in the country, and are respected throughout classical music for making the best attempts to be fair. (most orchestras do a sloppier version of the screen, and small orchestras don't bother) After I got wise to it, I only took "top 10" auditions for this reason, because the process is an expensive burden on the player.

You are right-ish about the numbers of players in circulation. This is an interesting point about why we or our colleagues know who is playing behind the screen: 1. we know who's at the audition. 2. At any given point, there are around 20-30 of us in the group of around 200-250 per instrument (who might all show up to a top 10-er) who are "in it to win it" at a top-ten audition. We all play gigs together often or went to school with one another. So the group we are selecting from is very small. Still, I believe we each could do it out of 100 players. It would be instructive to test this. I only mentioned it because it is interesting and most people don't realize this. (people are surprised when a musician can identify who is playing on a Mozart Concerto recording within a few bars at parties, but it isn't a trick-- your ear gets trained to recognize people in that way.)

I only asked about recognizing someone by their code because there is a scene in Silicon Valley where Dinesh "falls in love" with the code of Guilfoyle, without knowing it's his. I thought that was hilarious, but unlikely, since the characters code together every day. Your comment taught me something about how teams work together on code. I'd say it is like the "house style" of a newspaper.

I agree that it would be unlikely for anyone to recognize someone's code in a blind coding interview due to the numbers of engineers out there. Very interesting comment. Thank you! Cool that you are a singer and can keep it up.

There is a difference between succeeding in difficult tasks and succeeding at tasks that are seeing very high growth in overall accumulation of knowledge.

In high growth areas its not easy to make it big without working insane hours. Or return to regular practice after long breaks.

To me tech will will see near equal representation over time. As our trade becomes normal like any other. Of course that will mean the high intensity action will move onto else where.

>In high growth areas its not easy to make it big without working insane hours.

In classical music (and jazz etc.), you can't succeed at even a prosaic task such as winning an orchestra audition without working life-destroying hours. From age 10(or earlier, really) to age 40+, most classical musicians who "succeed" practice alone at least 4hrs a day, and rehearse, perform (and teach, when they are older) another another 4-6 hrs per day. There really isn't room for anything but this in their lives. That is the info I can offer to help w understanding this half of what you are saying. I think I didn't realize that this was not widely known. Coming from this life, it is easy to lose perspective on what is generally understood.

If every few years you have to throw away a good part of what your learned and redo it, taking big breaks becomes very hard.

You can do this in music because you start exactly where you stopped. In programming we haven't yet reached a state where knowledge of tools, frameworks and ecosystem can remain the same for years. Things change pretty rapidly here.

But you claim we have no possible theory for how DNA could make men and women's minds different, or that society could create differences in how much genuine interest women have in tech?

We don't really know why some people are straight and some people are gay. Does this mean we deny that gay people exist? My father is gay and his identical twin is straight. How did this happen? According to Wikipedia, studies have shown this trend often enough to conclude that genetics alone can't explain sexuality. So do we just argue that there really is no difference between gay and straight people (ie. it's a choice), because we don't understand the precise reason for the difference?

Even if your description of the situation is accurate, just because there's a theory on one side and none on the other does not remotely imply that theory is automatically correct and should be acted upon.

For the people living in those historic eras. We don't know the effects it has on people living in the modern era.

And then you have to make the leap on why tech companies should be enacting social justice in the first place. They're not the guardians of society. Social justice is inflammatory and hiring is basically a zero sum game because "diversity" doesn't affect head count (other than for the liberal arts majors and recruiters who have made careers out of the diversity issue)

3) Differences in representation are explainable by gender differences

Really? I read it as

1) Gender differences exist

2) Differences in representation may be at least partly due to gender differences.

3) Differences in representation should not be taken as automatic proof of bias.

Probably because nobody actually knows right now what actually causes the differences and how. The case that is caused exclusively by entrenched sexism - somehow driven by people graduating from the same institutions who have been fighting sexism for decades now, and by the same people who repudiate and denounce sexism on every corner - also does not sound very convincing. Sexists, of course, do exist, so does sexism, but presenting it as sole driver and concluding any deviation from 50% or whatever the population demographics would be is the ironclad indicator for sexism is as "unrigorous handwaving" as the opposite extreme.

It's not hard to see why people would get angry at a special pleading intended to excuse gender disparities.

1) Sexism exists

2) ???

3) Differences in representation are explainable by sexism

I think the original article was making a more nuanced argument allowing for an interconnected set of factors explaining the gender imbalance.

In other words: you'll need to demonstrate either that it's somehow harder to write the 900th instance of the same SQL query than it is to understand supersingular curve isogenies, or that somehow we've acculturated people to believe it's OK for women to nerd out about supersingular curve isogenies but not about SQL queries.

Neither argument makes any sense. Virtually all of what software developers do, even the hyperqualified developers working at Google, is less intellectually challenging than real hard science work, and if anything it's become far more socially acceptable to work in computer fields than to be a hard scientist.

We're in Occam's Razor country here, so the charge that concern over sexism is itself "special pleading" isn't going to stick.

Here are male/female distributions of bachelor degrees across many majors. Few of these sit right at 50%:

http://www.randalolson.com/2014/06/14/percentage-of-bachelor...

Are you arguing that every field that doesn't sit right at 50% is sexist against the underrepresented gender?

(I don't dispute that sexism in tech exists. I just dispute the expectation that representation will always be equal absent sexism.)

As I've pointed out before, that shortcut doesn't work.

If we just look at the higher level computer science your jobs, then again we will see that in the most comparable field, mathematics, women underparticiate something like 30-70.

If we ignore all that and pretend working with computers is uniquely bad for women judged against comparable careers, can we start calling out the enablers yet?

Well, bad news, there's another hurdle to jump, which is that female participation in computing is inversely related to a country's ranking on gender equality tables. I've yet to see that stubborn fact integrated successfully into the argument. It is mostly just ignored.

And that's off the top of my head. There are probably more. It's a complicated question to answer. So you must demonstrate at least familiarity with the obvious hurdles before questioning the moral worth of the people who disagree with you.

Well the obvious difference is that CS is not a science at all. It's really more akin to engineering, which AFAIK has a similar gender imbalance.

Is is possible that two things are true? That (1) there is sexism in tech that must be addressed, and (2) even if women felt welcome in tech, the composition of tech could be <50% women?

I can believe that a switch to NP's as the go-to for routine care might be a good idea. But even if that is the case, you're making a very different argument here. I wonder if you notice the difference. You're arguing that more men in nursing might be beneficial for customers, not that men are being kept out of nursing by sexism.

I believe the same is probably true of tech (that more women in tech is better for customers), if for no other reason than something like half of customers are women.

But these two arguments are very different in their moral implications. When you argue that sexism is keeping women out of tech, this is fuel for the outrage machine of moral indignation, like the Washington Post article above. It is evidence that he industry is wronging its female employees and potential female employees. It is cause to criticize the industry as sick and oppressive (which it certainly is in some cases, most notably Uber).

So I think that this criticism is well warranted in many cases. The question is: where are the goalposts? Is the industry sick and oppressive towards women until it hits 50% representation? I don't think that's a fair expectation.

It's hard to find a benign reason why the percentage of women CS majors would halve in the last 30 years while the profession as such has become more attractive.

http://www.npr.org/sections/money/2014/10/21/357629765/when-...

Now, look, this explanation may be wrong, but, no, it's not hard to imagine alternate explanations. Nice and neat conclusions in the social sciences are hard to come by. Anybody in this thread claiming that they know exactly which explanations are the right ones is full of shit.

I don't think it is hard to find a good reason why someone wouldn't go into nursing (even if this wouldn't disprove sexism in nursing). It has barriers to entry by requiring formal education, which can be competitive. The working conditions are somewhat difficult having to work irregular hours and be exposed to emotional stress as in seeing people dying. At the same time it still relatively low status, not necessarily well paid and has limited career prospects.

Of course you can argue that this shouldn't be the case. That the profession should help people dealing with stress and have good working conditions etc. Still it much harder to imagine these sort of "hard" reasons, negative things that are inherent to the profession, when it comes to programming.

As you justly said, we're in Occam's razor country. So, do you think the 2 observations above are more easily explained by socialization, or biology ?

One side points out that gender parity is overwhelmingly better in other STEM fields, that in modern science those fields are all themselves largely defined by software and technology, that there's a clear history of women being excluded from other professions, and that those professions later and enduringly corrected those problems, and that gender parity in CS has sharply decreased over time.

The other side has something to say about monkeys.

I'm sorry you feel the need to trivialize my point about monkeys. Unfortunately, evidence is going to keep rubbing you the wrong way, you're going to love this new thread :

https://news.ycombinator.com/item?id=14942022

“At the time, it seemed clear to me that any between-sex differences in thinking abilities were due to socialization practices, artifacts and mistakes in the research, and bias and prejudice. ... After reviewing a pile of journal articles that stood several feet high and numerous books and book chapters that dwarfed the stack of journal articles … I changed my mind.”

Why? There was too much data pointing to the biological basis of sex-based cognitive differences to ignore, Halpern says. For one thing, the animal-research findings resonated with sex-based differences ascribed to people. These findings continue to accrue. In a study of 34 rhesus monkeys, for example, males strongly preferred toys with wheels over plush toys, whereas females found plush toys likable. It would be tough to argue that the monkeys’ parents bought them sex-typed toys or that simian society encourages its male offspring to play more with trucks. A much more recent study established that boys and girls 9 to 17 months old — an age when children show few if any signs of recognizing either their own or other children’s sex — nonetheless show marked differences in their preference for stereotypically male versus stereotypically female toys.

Halpern and others have cataloged plenty of human behavioral differences. “These findings have all been replicated,” she says. "

You understand, right, that my side of this argument doesn't have to predict any part of human society from the behavior of monkeys? That all my argument requires me to be able to do is count?

The studies showing that female infants are more interested in looking at pictures of faces, while male infants are more interested in looking at pictures of mechanical toys wasn't enough, so researchers are one upping and studying monkeys. It does not help though, I heard from someone saying this in good faith that the sex differences in toy preferences in monkeys are due to societal norms in monkey groups.

At this point I'm seriously starting to consider the socialization hypothesis (which is obviously true to some degree, i.e. socialization plays a huge role in the observed sex differences) to be unscientific, that is, unfalsifiable.

Actually, there must be a similar female superhero stealing all of the boys' medical school applications as well (and overall college applications, which are majority female)

Or... could this all be the result of decades of encouraging young people to follow their heart when thinking careers ?

People, Occam's razor please.

https://news.ycombinator.com/newsguidelines.html

1. There is problem X

2. Possible cause of problem X is the reason Y

3. In fields where we are sure reason Y is rampant, we have successfully solved problem X by attacking instances of Y

4. In the field F, we have had no comparable success doing the same, despite all efforts.

5. Therefore, it is obvious that reason Y is much worse in F than everywhere and we should double and triple and quintuple our efforts there, until there is success. Since we reached success in other fields, it is stupid to assume we can't do the same here.

Do you see a problem with this chain of thought? If not, let me apply this logic to another model:

1. Cough is frequently caused by bacterial infections.

2. Antibiotics treat bacterial infections with much success in many cases

3. In patient P, antibiotics did not lead to any improvement of his cough.

4. Obviously, we need to double, triple and quintuple antibiotics doses until P is finally healed.

I hope you realize here that maybe P has a viral infection instead, or maybe he needs to stop smoking, or maybe he has another one of the hundred of causes of cough there could be.

>you'll need to demonstrate either that it's somehow harder to write the 900th instance of the same SQL query than it is to understand supersingular curve isogenies,

What you mean by "harder"? Is it harder to lift 600lbs than to give birth to a baby? Why do you assume that the question is about "harder" and not, for example, "more pleasurable" or "more intellectually satisfying"? Maybe SQL queries just bore prospective biologists to tears?

BTW, among composers the gender balance is even worse. Is it harder to write a symphony than to decode human genome? What metrics would you use to compare them?

Maybe instead treating everything as a nail we should try and consider some alternative hypotheses than "CS and musicians hate women"? It might be more efficient than force-feeding antibiotics to somebody who does't have bacterial infection.

What's worrying is I know a couple PhDs - in STEM - that were granted when they probably shouldn't have been. One lacked any originality, and the other was basically incomprehensible.

Women were literally not allowed to be doctors or lawyers, either by law or by certification boards. Programming has never been a field requiring certification.

http://memory.loc.gov/ammem/awhhtml/awlaw3/women_lawyers.htm...

You can not remove the biological factors. So they're a dead end.

Too me, it sounds more like people claim we have nowadays removed sexism, that Google has none, and others disagree, that they are still there, while not as strongly maybe, and until they are fully addressed, we can't tell if the system is fair to biology or not.

"In mathematics, just 15 percent of tenure-track positions are held by women, one of the lowest percentages among the sciences, along with computer science (18 percent), and engineering (14 percent)."

https://www.theatlantic.com/science/archive/2016/11/math-wom...

NPR talks a bit about the subject, including how personal computers were initially marketed at young men. It's a good read.

http://www.npr.org/sections/money/2014/10/21/357629765/when-...

IMO the biggest counter-example of the "women are better at working with people / soft skill" theory is politics.

Politics is all about who you know, who has your back, how you can get people to do what you want (the voters, other party members), how you can get people to like you, can you predict when someone's going to stab you in the back? can you predict when to stab someone else in the back? Should you support person X? What will person/group X think if you support person Y/policy Y? It's all about soft skills, it's all about reading people, it's all about working people.

You don't need maths or logic, or numbers,or rationality, or science for politics. I'm sure we all know how bad many polticians are at science/STEM.

If these theory were true, you'd expect women dominate politics in the same way men dominate tech. But that's not what you see at all. Advocates for the "women are better at soft skills, and worse at numbers" need to explain this glaring counter-example. The other theory, of institutionalised sexism against women, fits observable reality better.

I'm not sure how you can merge the two theories that "Women are just better at people skills" and "Women aren't better than the best". The whole thesis is claiming that women are as naturally superior at people skills as men are naturally good at tech/STEM. Google is supposed to hire the best, and it's >50% men, so why isn't something that "hires the best" (in politics) still <20% women (e.g. US Congress).

I remember my time in the Harry Potter fandom, which is ~99% female. Boy, was there politics. I can remember some truly epic [1][2] incidents of Internet drama all driven by people seeking social status through any means of manipulation possible.

[1] https://fanlore.org/wiki/The_Ms.Scribe_Story:_An_Unauthorize...

[2] https://fanlore.org/wiki/Victoria_Bitter_Wank#Further_Readin...

How much counter-examples can you ignore before you say well maybe this "women are good at people skills, men are good at STEM". It seems like people really think "well men are better at basically everything"

In industries that actually are 50/50 or majority female, you do see a number of women in top leadership positions. For example, the president of the national teacher's union is a woman [1], as are the three co-presidents of the national nurse's union [2], as is my alma mater's (which only went co-ed in 1974) college president [3].

[1] https://en.wikipedia.org/wiki/Randi_Weingarten

[2] https://en.wikipedia.org/wiki/National_Nurses_United

[3] https://www.amherst.edu/amherst-story/president

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3149680/

So where are the places which are 90% women? Why isn't politics 90% women. Instead it's about ~20% and that's after years of campaigning & activism.

[1] https://www.dol.gov/wb/stats/occ_gender_share_em_1020_txt.ht...

Here's a UK survey of chefs: http://www.bighospitality.co.uk/People/Survey-finds-imbalanc...

It's almost like there's some sort of social factors which will ensure men get to the top!

The problem with the "???" is it's not in its proper place in the chain of logic. The syllogism instead goes:

1) Do we observe certain traits cause individuals to tend toward science, while others cause them to trend toward the humanities?

Yes. Whether or not you will enter the humanities or the sciences is correlated with whether you display a higher empathizing quotient ("preferring to work with people") vs. systematizing quotient ("preferring to work with things/numbers"), regardless of your gender.

2) Do we observe these traits vary by gender?

Also yes. Men generally have higher systematizing quotients, while women have higher empathizing quotients.

[source for both points: http://www.somersetcanyons.com/ourpages/auto/2016/3/11/48638...]

3) Therefore, some amount of difference in representation logically comes from differences in these gendered traits - whatever might have caused them to become gendered in the first place.

>Okay, so that's why the majority of accountants are women? That's why STEM fields that don't involve people at all (math, biology) actually have quite a lot of women, while programming doesn't have much women despite being on the social/people-oriented end of the spectrum?

Women represent over 60% of all undergrad university students, so we should expect fields that require an undergrad university degree to be a majority women barring other factors. Yet math, physics, chemistry, earth science, statistics, engineering - every one of them has fewer women than it should proportionately. Even biology, which for a brief time had a higher proportion of women than 60%, recently dipped back below that point. [graph: https://www.aps.org/programs/education/statistics/upload/Deg...]

Conversely, here is a list of majors that have consistently had above 80% female proportions:

Fashion Design, Interior Design, Education, Social Work, Nursing, Occupational Therapy, Art History

The specifics of why one given STEM major is more under-represented than another is an interesting question, but it shouldn't obscure the overarching trend - whether by a little or a lot, all the STEM majors have fewer women than they should. The data does indeed support the idea women prefer people-oriented professions over thing oriented professions on average.

At the core of diversity hiring (which corporations have spent billions of dollars on), the benefits of racial and gender diversity in a workforce have never been proven.