[StavrosK]

Certainly, systemic bias may still be a problem. The issue is that, unless you believe that men and women are exactly the same, and physical and psychological differences somehow stop at what interests they'll pick, the expected distribution of preferred professions wouldn't be 50-50.

If there are innate differences in what men and women find fulfilling, you wouldn't expect to see equal participation in everything. Shooting for 50% quotas is then misguided, because all it means is that you're doing more work to find candidates that will pass your filter, as you're looking in a smaller pool on one of the two sides.

[msluyter]

It's unclear to me that anyone is aiming for a 50-50 distribution, nor am I implying that that should be the end goal. Megan McArdle has a new article in which she generally agrees that there's probably a biologically rooted difference in interest levels in tech, but that:

"So even if the disparities don’t start off as discrimination, you can still end up with an environment in which women who could be great engineers decide they’d rather do something else. A “natural” split of, say, 65-35 could evolve into a much more lopsided environment that feels downright unfriendly to a lot of women. And the women who have stuck around anyway are apt to get very mad indeed when they hear something that seems to suggest they’re not experiencing what they quite obviously are."

Worth reading in full: https://www.bloomberg.com/view/articles/2017-08-09/as-a-woma...

[StavrosK]

Ah, yes, in that case I agree with you 100%.

[foldr]

Men and women don’t have to be “exactly the same” to be equally good at programming. Germans and Spaniards are not “exactly the same” on average, but that doesn’t mean we automatically expect there to be more German programmers than Spanish programmers or vice versa. The question isn’t whether men and women are exactly the same, but whether they are, overall, on average, equally good at doing the relevant tasks. Even if men and women have different cognitive strengths and weaknesses (which I doubt that they do, to any significant extent), these might very well balance out. So for example, Federer and Nadal are very different, but about equally good.

[StavrosK]

That is very much not the question. The question is whether there exists an equal proportion of men and women in the population. If you want to hire a senior developer, and women make up 60% of that population, it would be counter-productive to try and mandate 50% quotas, because then you'd be looking for people in the minority population.

Sexist hiring practices isn't having fewer women than men. Sexist hiring practices are having a men:women ratio that is different from the ratio in the applicant pool. If the general population of developers is 90% men and 10% women, hiring 9 men for each woman is exactly what you'd expect to see in an unbiased hiring process.

I don't think "equally good" even makes sense. Of course there's nothing preventing women to be as good as men at programming, or vice-versa. You have to look at populations per level of competence, and that still doesn't mean that someone "can't" be as good as someone else. Maybe they're just not interested in that thing as much.

[foldr]

The relative lack of women in the applicant pool is obviously one of the problems that Google and other companies are trying to fix with their various outreach initiatives.

If you take the proportions in the applicant pool as given and unchangeable, then that may not be sexist per se, but it is perpetuating an imbalance that has its roots in sexism on a societal level.

Again, the mere existence of differences between men and women absolutely does not mean that you should expect there to be a difference in the size of the respective applicant pools. That is just bad logic. The argument is the same regardless of whether we are talking about interest or ability.

[StavrosK]

> The relative lack of women in the applicant pool is obviously one of the problems that Google and other companies are trying to fix with their various outreach initiatives.

Certainly. We're just debating whether the proposed solutions are doing more good than harm.

> Again, the mere existence of differences between men and women absolutely does not mean that you should expect there to be a difference in the size of the respective applicant pools.

Why is that? Are we sure the distributions of things men and women like are exactly the same? Men and women are different both physically and psychologically, so what is the force that counterbalances the differences and makes the distributions of things each gender likes identical?

[foldr]

>Why is that? Are we sure the distributions of things men and women like are exactly the same? Men and women are different both physically and psychologically, so what is the force that counterbalances the differences and makes the distributions of things each gender likes identical?

Because it is possible for two people to be very different and yet equally interested in and equally good at any particular task. It is obvious that men and women are different in some respects. What is not obvious that men and women differ in such a way that men are going to be better at programming, or more interested in it, than women are.

[StavrosK]

> What is not obvious that men and women differ in such a way that men are going to be better at programming, or more interested in it, than women are.

What is also not obvious is that they don't. Since men and women differ in pretty substantial ways, I think the burden of proof is on people who claim they're both interested in the same things equally.

It's fine if some demographics just aren't interested in some things, if they're free to choose them. It's not fine to say "no! You must be interested in whatever everyone else is interested in, otherwise the notion of equality I have doesn't make sense!"

To summarize, my entire point is that we should make it so each person can freely choose what they want, rather than trying to make the posterior distributions fit the model that we imagine must be the right one. That means no "boys can't play with dolls", no "engineering isn't for women", but also no gender/demographic quotas.

[deecewan]

> if men and women have different cognitive strengths and weaknesses (which I doubt that they do, to any significant extent)

I don't agree. And I think this is a core reason that the outreach is needed. Because of different points of view, and a broader candidate pool with different psychological make-ups, we can expand our viewpoint and do better, more interesting things.

I think that Google, Facebook, etc are not only reaching out to women in tech as a community gesture. I think that they see the value in the expanded viewpoint and the new things that can be developed because of it.

[foldr]

Men and women certainly have different experiences and viewpoints, but that's not what I meant by a cognitive strength or weakness. I meant e.g. men being significantly better at spatial reasoning, or something like that. My impression is that the evidence for significant differences of this kind is rather scanty.