Can you name any of these male dominated companies that hire men disproportionately to the gender demographics of their applicants/interviewees? The issue isn't the distribution of employees, it's whether there is sexism present or not and the first thing you'd probably want for evidence of sexism is disproportionate hiring of a demographic compared to their applicant/interviewee demographic.
Just because something is more represented by a demographic does not make it sexist. e.g. Nordic countries seem to be generally celebrated as being socially progressive and ridding themselves of sexism, yet female representation in STEM fields is incredibly low.
This really isn't a problem anyone wants to solve. STEM especially has unique issues where if you take extended absences things such as your skills or research will be invalided over the short periods you may be gone. Tackle that with a lack of a childcare and you have a system set to ensure the failure of women. Retention is close to impossible as the cost of childcare is impossibly high.
I'd say this is more industry-specific than STEM. In software I can jump back onto a project I left behind years ago and get productive immediately. This is something I've done with multiple projects. But there is a difference between some of those and other kinds of projects where the deadlines are tied into hardware cycles, like, automotive engineering. And generally in engineering, childcare is affordable because your salaries are high.
I was willing to listen to your point (that perhaps some companies favor women and so women disproportionately percolate to those companies… at least in theory seems like a possible argument, even if I’m skeptical) until you got here:
>> Also, because the intelligence distribution skews more male at the higher end, a company will be disproportionately male simply by being more technical and having a more difficult hiring bar.
Okay, buddy. Women used to be dominant in computing. Literally were the computers. And when artificial computers came to the fore, many transitioned to writing programs for them. Until it became a more prestigious field and men came to dominate.
And once the personal computer became widespread with computer game culture arising and a culture of harassing women until they leave became common enough to be an issue, why wouldn’t one expect that those who developed an interest in computer programming (ie from contact with computer games, etc) to be predominantly men?
(And I love computer games, but there is an insane amount of harassment. Everyone who plays knows this.)
>Women used to be dominant in computing. Literally were the computers
This is pathetic cope and completely misleading. The "programming" of that time had almost nothing to do with the modern practice. These "computer" roles were for number crunching, not software development.
You're bending over backwards to deny the fact that men and women are biologically predisposed to certain interests. Why? Is it some weird sort of insecurity?
>It has nothing to do with intelligence curves.
It has literally everything to do with intelligences curves. The male distribution of intelligence has a wider variance - that's a statistical descriptor of the normal distribution. This means there are more men at the extremes of both ends of the curve. And an inescapable consequence is that men will be overrepresented among knowledge workers in meritocratic system. Don't worry, you can take solace in the fact that this also suggests a larger population of male criminals because it also means there are more dumb men at the other end of the spectrum.
The lengths to which people will go to deny the effects of meaningful, proven differences among demographics is infuriating. It's gaslighting and it helps absolutely no one in the long run, assuming you want a functioning society.
Hey, whatever helps you sleep at night, but I agree with the other poster that EVEN IF you think the male/female curve distribution thing is completely correct without caveats, you just don’t have to be massively intelligent to do software development. You’re building a corporate SAP system, not proving some theorem or unifying the fundamental forces or something.
This isn't about my personal feelings, it's about the insistence that the demographics are not a reflection of the distribution of competence and personal choice, but discrimination. The entire foundation of discrimination against white males is dishonest because proponents refuse to allow for any explanation for overrepresentation beyond unfair hiring practices - which are being used to implement the systemic discrimination that they claim to be fighting.
>You’re building a corporate SAP system, not proving some theorem or unifying the fundamental forces or something.
And regardless of what you're building, you want the most competent employees that you can get for the lowest price if you are operating within a meritocracy. In that case your demographics will be strongly influenced by the distribution of competence, before even considering that the pipeline is skewed because women are simply not choosing to pursue software development - and that's not a problem that needs correcting unless you're playing tribalist team sports.
The anti-tribalists are using mere accusations of tribalism to justify their own vicious, exclusionary tribalism, and simply not allowing for discussion of alternative explanations which could appear to be motivated by tribalism. That's what keeps me up at night.
Look, it’s not helpful when making the case that white males (which I am, BTW) are being discriminated against and that we should feel bad about that and fix it when folks start saying “but actually women aren’t as smart at the high end as men and so…”
My target demographic is people who can distinguish between the phrase you gave, "women aren't as smart at the high end as men" (which is false or even nonsensical) and the premise of my prior comment and the matter under dispute, that most smart people are men (which is true) and that the m:f ratio increases as you raise the bar (which is also true). These are logically distinct.
Then how are we supposed to argue against the false claim that discrimination is the primary driver of demographic underrepresentation if this is the true cause?
Agreed. A vast majority of supposedly "high tech" jobs are solidly average intelligence, including coding, so if anything women should be rewarded in such positions because males' wider variance in I.Q. cuts both ways. It's only pure delusional thinking that keeps up the "muh I.Q. is one of the highest" attitude among "tech" bros.
Programmers are not solidly average intelligence at all. Everybody writing software is at the upper end, even if defined very loosely to mean something like top 10% of Americans. Note that among SAT scorers, the top 11% of males score >700 on the math section and top 7% of females score >700. This isn't some super-elite group, and you're already looking at 1.6x as many men.
But you're arguing somehow that we should specifically discriminate in an attempt to hire dumber people, so... well, that seems like a bad idea to me.
Average intelligence (around 100 points) does not get you hired in any high tech job. It is called "high tech" for a reason, it's not "average work" by any measure.
>And once the personal computer became widespread with computer game culture arising and a culture of harassing women until they leave became common enough to be an issue, why wouldn’t one expect that those who developed an interest in computer programming (ie from contact with computer games, etc) to be predominantly men? (And I love computer games, but there is an insane amount of harassment. Everyone who plays knows this.)
I'm sure this is a part of it, but I think you're overselling how impactful the harassment part is. The skewed ratio in CS has been something going on for decades. Online gaming, especially with voice, is the primary offender of online sexism in games. These types of games weren't that popular until early/mid 2010s. If you were interested in other types of games and largely avoided online games, sexism wasn't nearly as bad. That's not to mention a sizeable portion of CS graduates have no interest in games, let alone online games, as a whole.
Meanwhile, this doesn't answer the gender-equality paradox. At the same time, these "less gender-equal" countries definitely don't have a shortage of misogynist messages on their online boards, gaming communities, forums, etc. Still as misogynist, but more women entering CS: what gives?
Not at all. It still holds. Let's suppose the field were overall 90% women, 10% men. Even then, companies looking to hire the smartest candidates would be disproportionately male -- say, 15% male instead of the average 10% -- as the population male:female ratio increases as you raise the intelligence bar.
For many species, males have a higher standard deviation across a variety factors. There are biological reasons for this, since males are more disposable than females for reproduction. Think of one rooster for a dozen hens - this is still a viable flock, even if the other roosters get eaten. Nature can experiment more with males simply because the stakes are lower, and sometimes those experiments are worthwhile. [edit: Turns out this is not true for birds, so I learned something!]
Now, if you have two normal distributions with the same average and area but a tiny difference in the standard deviation... well, let's not go there because the math is too politically incorrect. [edit: And besides, many other factors affect outcomes besides genetics, so I do believe we should keep policies gender-unbiased as a matter of principle].
> There are biological reasons for this, since males are more disposable than females for reproduction. Think of one rooster for a dozen hens - this is still a viable flock, even if the other roosters get eaten.
Note that birds do not have mammalian sex chromosomes - males have ZZ and females have ZW - and females have been measured to have greater variability in birds.
So there is not really a biological reason as you say -- it just happened to go one way or the other in the past, and now different trees of species are stuck that way.
If we can repeatedly find differences in variance despite similarity of mean for different sexes within a species or other groupings, then that would be a kind of thing that would be reported in biology research.
I've not come across this before ("males have ZZ and females have ZW - and females have been measured to have greater variability in birds"). Do you have some references/books you can recommend? I'd like to find out more.
The standard deviation for the population of men is higher than women, so you get more outliers for men, smarter and dumber, but the average is the same. If that's true, then the math says the upper percent will consist of more men than women, assuming a split population. The dumb end will also have more men.
I realize this is taboo, but I don't see how any other conclusion can be made. I don't think it matters though, in the grand scheme of things, assuming the hiring is based on merit rather than bias. And, if the above is true, there's a better chance that you hire a dumb man than a dumb woman, so merit is important.
> Really? You think there are inherently more intelligent men than women?
Uh, yes. This is a well-established fact.
Women have two X chromosomes, both of which get used [1], and men only have one. So this should be your baseline expectation. In women, that genetic information used to construct the brain, gets, loosely speaking, averaged together.
It is of course also consistent with every empirical observation. E.g. the on the math SAT, 1.6x as many men score 700-800, 2x as many men score 800, despite the fact that the school system is biased against boys. And on grade school and high school math contests (such as the AMC, the AIME), the male/female ratio increases more and more the higher the scores go, up to 89% male before sample size was too small for me to get a read on the proportion. Of the top 10 Jeopardy champions, 9 out of 10 are male.
Sorry but that's pseudo-scientific bollocks. Far from being "well-established fact" it is actually contentious and poorly supported by the evidence.
Intelligence testing is a human created concept. There are all manner of cultural biases encoded in it. Intelligence is more than just being good at sums which is what your math SAT example seems to suggest.
Jeopardy champions are hardly a representative group. They're a self-selecting group by definition : those who chose to go on a quiz show. Maybe men show a bias towards showing off on TV or are better at remembering random facts. Your example doesn't control for any of those.
These are good points ("Intelligence testing is a human created concept").
However it wouldn't be surprising to me if higher end SAT scores would be highly correlated with likelihood of being hired at FAANGs of this world later in life.
If the above is true, and given that the SAT score distribution is shallower and wider for males, then it'd probably follow that there would be more males employed at those places than females, even if the hiring would be totally controlled for biases?
Or, is it your point that there's (probably) correlation here, but not causation? As in, there's bias against females (there most likely is), so more males are hired, and it's an independent fact that males have broader SAT scores distribution. Ergo, if we controlled for biases, then faang hiring results would be 50/50 male/female, even if SAT score distribution would stay the same?
Did you edit your post? Maybe I overlooked a paragraph.
The math SAT questions aren't a "bunch of sums." And I agree with you about cultural bias. As I stated, the school system is biased against boys. So the SAT math scores should understate the male/female ratio at that level.
Interesting thing about the chromosomes - in other species where the chromosome situation is reversed, like birds I think, the female is the one that displays more variability.
That's not what the parent comment said. You could see this picture even if average make intelligence was lower than female — all you need is a wider distribution.
Don't our genes dictate how we interact with the environment? I'm not sure how it could be anything but.
That being said, yes, we can change our environment and see changes in measured intelligence. A person with a head is smarter than one without.
To tie it back to your original point, you're saying the male/female differences in intelligence are purely environmental? As in, if two people with different genetic make ups were brought up in the same environment they'd have the same outcomes?
If we're seeing equality it's likely the result of changing our environment to force this.
Can you explain to me how our genetics don't dictate our responses? Short of God or some other unknown unknown (which is certainly possible), I don't see how it could be anything else.
Identical twins aren't in identical environments. It's not possible.
Sorry for the edits, I was trying to shortcut the argument based on what I thought you were saying.
Identical twins aren't exposed to identical environments; if the programmed response was particularly sensitive to small input variations, genes could completely determine interaction with the environment and identical twins could still be very different.
No. Genetics shape how the individual responds to the environment. Just because you can change the environment and get different results doesn't mean it's not due to genetics.
"Broadly the same" and averages don't work well when you're talking about a very specific sub population that would, necessarily, be picked from the few at the top end of the intelligence distribution.
No, please see the context/comment chain that this comment exists in. We're talking about hiring in tech. The average person doesn't apply to, get hired for, or succeed in, tech positions. Most tech positions require a higher degree because most tech (programming, computer science, hardware engineering, etc) requires or directly benefits from that higher degree. Those that hold those higher degrees have an above average intelligence. Those with the highest intelligence usually having the most success in the technical part of the tech workforce. Tech prefers the higher side of the intelligence distribution, because it's hard.
Your comment is so ambiguous that I'm not sure how to interpret it, but I encourage you to read the article completely, as the number you quoted is not part of the conclusion.
That is obviously garbage statistics, isn't it? The conclusion they've drawn about the female distribution is based entirely on the performance of one female player. And they're obfuscating rank ordering with rating gap.
Also, for an unexplained reason, they only consider Indian players. That's weird, isn't it? Why Indian? Let's investigate.
Using the Oct 20th 2020 data, since the Oct 6th 2020 data isn't available on the FIDE website, we could look Chinese players (birthday <2000 to follow their criteria):
I picked China because they're a large country and because the top active female player is Chinese.
Of the top Chinese players, the ordering is 8 males, 1 female (Hou, Yifan), 12 males, 1 female (Xie, Jun), then 3 M, 1 F, 5 M, 1 F, 1 M, 1 F, 2 M, 1 F, and so on.
The overall Chinese female percentage is 30.47%. Here's a table of percentage over rating threshold, with cumulative sums shown.
There is a big pile of low-rated Indian men with FIDE ratings, while mediocre women aren't interested. Note that the article fallaciously compares the male average to the female average as if that means something.
If we were to compute the article's bogus statistical argument about the top woman, a more appropriate female percentage to use would be more like 12.69%, the maximum on that list, before the hoard of mediocre males, instead of 6.1%.
It would make more sense to base our statistics on the set of all the world's active players, instead of one woman from a particular country against an irrelevant percentage. Here is that data:
M F M sum F sum Fsum/(Msum+Fsum)
2800 2 0 0 0 -
2700 32 0 32 0 0.00%
2600 186 1 218 1 0.46%
2500 437 11 655 12 1.80%
2400 1049 39 1704 51 2.91%
2300 1928 87 3632 138 3.66%
2200 3385 156 7017 294 4.02%
2100 5699 231 12716 525 3.96%
Oh, and one other thing, at a meta level: The entire construction of the author's argument is to pick a statistical measure with a wide standard deviation, instead of a better one that would use more data to get a low standard deviation. Then he finds that one real-life outcome is within that wide standard deviation.
Not only that, he defines it via the exponentially increasing ELO rating distribution, instead of rank order, and that would have the effect of piling up a bunch of lower ranked players together within one standard deviation of the mean. This means it would be virtually impossible to get a result 1 standard deviation out, let alone 2.
Yep. In fact, something even narrower. That it would do so without any sexual dimorphism between men and women.
(And I'm not even here to have this argument -- it's just a basic fact behind one of the reasons why some tech companies would skew more male than others -- but then some people decided to argue.)
> The skew only means the smartest 1% of men are smarter than the smartest 1% of women.
I don't think you can put it like that. Even if the IQ (or similar "cognitive capacity" score) distribution is broader&shallower for males, it doesn't mean what you had said.
Judith Polgar was #7th in chess rankings, even if overall number of players scored with ELO is counted in millions?
Dumb people need to be handheld into understanding your precise meaning because you can't assume they'll attach correct qualifiers so that the sensible meaning is inferred. (And even then it probably won't work.) The purpose of your post was to communicate with dumb people.
Most tech jobs don't actually require significantly greater than average intelligence. If you crunch the numbers it takes a pretty high bar to skew the gender of the posterior distribution much. It's not inconceivable and shouldn't be discarded out of hand, but there are plenty of other factors (including non-discrimination ones) worth keeping in mind.
As an aside-- when you make the variance argument it's useful to point out that men are vastly more over-represented in convictions or some other negative characteristic... otherwise it's just too easy for people to read as a suggestion that men are better rather than simply being a population which as a whole tends to have more diversity in its performance.
Communication is a two way street, and there are enough assholes out there who do think that men are inherently superior that it's in everyones interest to put in some more effort to make it clear that it isn't what you're saying.
I don't think it is useful here (on HN) to point out the low end. That just gets tedious and I think comes across as tedious. The pertinent fact is that m:f ratio increases at the top, and variance is just the most irrefutable factor that breaks apart any presumption that the distributions should be identical.
Also, I'm not trying to pretend that's the only difference between men and women. Men sleep 20 minutes less than women. Is that "inherent superiority"? No, but it's an advantage. Or maybe men undersleep and it's a disadvantage.
But -- we write because we hope to communicate, and if you're mistaken as a flaming misogynist then the communication will fail and you won't reach the people who could benefit the most from your perspective.
I think it can be worth it to put on a bit of a dance to make it clear that you're not just looking for reasons to be dismissive towards women, even if in some cosmic fairness sense you shouldn't have to prove you're not a witch.
In my mind a reasonable way to do that is to point out that variance differences often make men worse. Particularly so since someone one extra sd more intelligent on the top end probably has little benefit to their life, while someone being one extra sd more violent on the bottom end makes them much more likely to live a life thats miserable for them and those around them.
Just because something is more represented by a demographic does not make it sexist. e.g. Nordic countries seem to be generally celebrated as being socially progressive and ridding themselves of sexism, yet female representation in STEM fields is incredibly low.