[dahart]

Yes, "leaving computer science" means that participation rates are going down, that fewer women are choosing it now versus in the recent past. The phrase always meant women as a group are leaving, it never meant that lots of individual women were starting CS careers and then quitting.

> If you look at the most equal and rich countries where women are doing as well or in some areas better than men you don't have many who chooses software development as a career.

That is not true. Participation rates have been very high in some of the countries that are now low. Participation rates in some of those countries are currently high.

You're trying to suggest that women as a group don't want computer science, by their nature as women, but to make that claim you have to deny history and ignore facts.

> It could be just that fewer women want to be software developers and that is all that there is.

The numbers changed a lot, and there is a reason. What you're saying is you don't know the reason, and I agree with you.

> it is also very possible that it [discrimination] does not affect the number of women in tech much at all.

I think it's extremely, extremely unlikely. It's demonstrated that the choices are cultural and not intrinsic to gender. So, your job then is to show that cultural attitudes are not affected by cultural discrimination. You're setting yourself up for an impossible task.

[cc81]

>That is not true. Participation rates have been very high in some of the countries that are now low. Participation rates in some of those countries are currently high.

Yes, and my point is that when those women got more power, more choice, more money and more freedom they started to become less likely to choose software development and more likely to select other professions. Other professions that had historically been male dominated.

So it could be that if you are free to choose there are certain professions that will have a larger part male or female and that will amplify in a culture of freedom. You will choose what your friends choose.

For example I don't think that you would ever have 50% men/women that are interested in working in kindergarten. I think there is some biology in that and then it is amplified by most men not wanting it making it less likely for those that might want it to actually choose.

[dahart]

> when those women got more power, more choice, more money and more freedom they started to become less likely to choose software development

You're cherry-picking there, and incorrectly assigning causation to correlation. Before 1984 in the US, women's choice and power was increasing, and so what their participation in CS. In some other countries, women's participation rates in STEM have continued to increase as they integrate with the workforce. So you could use the very same argument you used to come to the opposite conclusion.

> You will choose what your friends choose.

Totally agree there! That may or may not have to do with discrimination, but you're right about that.

> I think there is some biology

This is the root of it. You've decided that women intrinsically don't want to do math or engineering.

There's data that shows otherwise, but it is indeed difficult to prove anything concrete. Nonetheless, we have strong indicators that today's ratios are not natural. And why would they have settled yet? It's barely been 3-4 generations since women were allowed to even vote. It's clearly still in flux, so suggesting that biology explains today's difference is obviously stretching.

Bottom line is I don't doubt preferences are at play, and what I'm saying is that biology can't explain the current preferences in the US or in Sweden either. Biology is not fluctuating on the 100 year time scale, and cultural norms are. Assuming that the known, documented sexism during the last century is fixed now and that cultural attitudes have settles and job choices reflect intrinsic gender differences... well it's my opinion, but looking at all the facts we do have, that seems so unlikely I believe it's impossible.