(Actually, people aren't confused about this, but it makes for great clickbait articles and ad revenue because it fans the flames of both sides of the argument.)
There's a couple places within the article that's more superficial though.
Maybe because "positive discrimination" is pursued using equality of outcomes as a key indicator?
I would say that you're right, and this means that specific outcome equality is not indicative of equality of opportunity.
But, it's also true that equality before the law and other measures that have been tKe over the last 3 generations do not guarantee equality of opportunity. Opportunity is in itself a sticky concept.
If you grow up in a family/community where a lot of people are tradesmen then you'll be more likely to be a tradesman. Same for long term welfare dependance, academia, public services, fireman/police, housewives, medicine...
The dynamics are complex and not fully understood. People form an idea about what's normal, possible and such. This is a socially based phenomenon.
I would argue that it is important to the concept of opportunity. The boxes that you put yourself into by default, and others put you in.
There aren't necessarily physical barriers. Lots of people don't end up on the default paths, maybe even most. But, defaults matter.
Anyway, let's think about it in reverse. Let's say we are abolishing sexism, racism, caste systems or some other form of big societal discrimination. A generation later, no change in outcomes. Did we succeed?
OTOH, it's hard to know what to do with specifics like women in tech. It's not inherently bad for women to have different average preferences and that will cause unequal outcomes. But, we also can't ignore outcomes (imo) entirely either. There must be actual outcome changes.
There's also the issue of feedback loops. If certain spheres are male realms, with women in a tiny minority this makes it harder for women to enter in the future.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3030621/
And of course, far more studies that have been going on for decades.
The vast majority of women will not continue dating a man who earns less than they do. Two thirds of young women in Germany expect their future partners to take care of them financially.
So for men, earning money and status is absolutely essential.
See also: https://www.theguardian.com/lifeandstyle/2016/jul/02/a-lette...
(2) What makes you think it wouldn't? After all, it affects everyone in society. And motivation is a pretty powerful force, it is basically how our entire economy is structured.
See also:
https://economix.blogs.nytimes.com/2009/11/18/money-gender-a...
"...it typically takes a lot less money to get women to say they are satisfied with their work than it does to get men to say it."
"Women were more likely to tell PayScale that say they find their jobs “very meaningful” than men were, with 35 percent of women and 27 percent of men describing their jobs this way."
So women have more freedom to find meaningful jobs. Men less so.
For free market choice on which jobs pay more -- I think this one is pretty clear. Do you think differently?
As for biological preference, I don't think it takes into account the average pay at all. Career path is usually chosen at a younger and more idealistic age, and probably more importantly - poorer countries have women overriding their preferences and going into higher paid lines of work, while richer countries let women do what they want. As per this very study. In the case where money doesn't matter because your family is rich, biological preferences become more important and people stop choosing based on average pay. So the biological differences seem to explain it really well.
So if you assert that discrimination is the cause for the imbalance, the evidence shows that your assertion is wrong, because it makes the wrong prediction.
This is not just a matter of degrees it is a matter of sign (the arrow is pointing in the wrong direction).
Robert Sapolsky - we need you now more than ever, https://en.wikipedia.org/wiki/Robert_Sapolsky
Most people seem very capable of thinking statistically, but they don't communicate in distributions.
For two people communicating in good faith, the two statements (1) "Boys are taller than girls" and (2) "Anne is taller than Sean" don't prompt talk about how both can't be true at the same time, sexist. People communicate statistical differences as "absolutes", but when communicating in good faith, everybody knows what everybody means.
Look at stereotype research. Most stereotypes are statistically true. People communicate them "as if" absolute, but nobody's head explodes when individual examples don't "conform" to the stereotype. That's because people can think in distributions.
That's not surprising, it's how to survive in a stochastic environment.
This gap between thought and speech only becomes problematic ("makes it really hard to have any discussion") because we can rhetorically exploit the fact that most people can't explicitly communicate about distributions. Just tell them they're bigoted, and they don't know how to refute you. Yay, you win.
When you average all women and all men, you get a "gap". However, like the imbalance in fields, this is almost entirely due to different choices. When you control for factors such as field, hours worked, years worked the gap effectively disappears.
Choices made by women are influenced by the glass ceiling, society expectations that women will take care of children and the elderly, and the traditional fields where they're expected to work. The decision to measure the gender gap by comparing equal working hours and equal job positions is not innocent, is itself a politically motivated and comforting way to make rational people think "I told you, there's nothing to see here". You need to evaluate what causes women to have those less hours in less paying jobs, to see whether there's discrimination or not.
Yet the salary wages discrimination reappears when you factor in hours dedicated to domestic work (the "second shift"), differences in salary between fields with equal risks (exposition to chemicals, hardship) where workers are majority women vs those with more men, or companies' (irrational) tendency to default in promoting male workers over females in similar situations.
https://www.washingtonpost.com/graphics/2017/business/women-...
It is seen in human beings and animals that statistically females mate sidewards or upwards in terms of providability. This in turn applies a pressure on males to compete for more and accumulate more resources, whereas females are not measured by males on their providability hence have no pressure on them to fight for the same resources. So, men will always compete more to get a greater share of the pie, this combined with capitalism explains whatever income gap remains after attaining equality of opportunity. Shouldn't the underlying metrics of mating preferences also be taken into account before putting it all under male sexism and patriarch umbrella ?
You can't ignore that we come from a culture with a tradition of 60+ centuries of male-favouring sexism and patriarchy, which has not vanished merely because of a theoretical solution to it (legal equality) has been formally declared. There's a lot of social inertia introducing biases in the opportunities available to women.
You can take into account mating preferences all you want, but how do those affect choices of job positions other than very indirectly? On the other hand, irrational prejudices in recruitment and career advancement processes have a direct impact on the quality of job positions that women have access to, and therefore in their opportunities. Unless you're downright denying that such prejudices exist, this cultural factor would have a larger influence than biology.
The incentive models, which take data from both biology and culture, claims a prediction that as you measure equality of opportunity and compare nations it will correlate to inverse gender segregation in the work force. The more equality of opportunity the higher gender segregation will be, caused by incentives that is derived from gender roles, sexual strategies, and sexual dimorphism.
The prejudices model make the opposite prediction. It claims that when measuring the equality of opportunity the higher it is the lower the gender segregation should be.
In order to resolve the question one measure equality of opportunity and gender segregation, creating data. This create a scientific valid argument in favor of one model over the other.
That is one aspect of the theory. An other aspect would be to measure difference in prejudices between different professions. If the model is correct then it should be able to predict based on such measure which professions are gender segregated and which are not, and inversely predict the level of prejudices exist per profession given any country. The incentive model claim that incentives are a large influence and such you could not make a predictive model based on measuring prejudices.
The incentive model also predict that nation where the incentives are different (based on Maslow's hierarchy of needs), you should see a correlated change in gender segregation. Incentives such as reproduction is listed last among the physiological need, so nations where water and food is scare for women should result in higher incentive to seek professions where those are easier achievable. The prejudices model would argue that there is no such correlation.
This is of course just a small sample in order to scientifically evaluate the two different theories. There is also matching philosophy theories (their names escape me right now), arguing in different direction to explain the data.
I personally look at the outcome (you can reach a billion people just by having a cool idea and sitting down and doing it) then look at the hurdles to that outcome.
One of the hurdles is that engineers love wasting literally hundreds of each others' hours. If I were building something as a newbie, people have no trouble sending me off on a wild goose chase to spend a hundred hours with my head in documentation. When an alternative ecosystem lets me literally type a single line and it's done. (In this case the first ecosystem hates my time and the second one respects it.)
This isn't really related to gender, but it's another hurdle. So when you look at the incentive (reaching people, fame, fortune, money), if there is a highly unequal distribution you have to look at some of the hurdles.
By the way no one knows your gender online when you're learning stuff, you can use a non-gendered username and get treated exactly the same asking stuff online and so forth.
In this case the hurdle is artificial: being cooped up with a computer while the tools you're trying to use waste thousands of your hours. But it's still a hurdle that impacts genders in a different way, and that matters.