[spangry]

I don't think it's a non-factual claim that there is a significant degree of horizontal occupational gender segregation in pretty much every advanced economy. A random search using terms similar to those above yields the following: http://imgur.com/a/lcHPU

The segregation seems to follow a pretty stereotypical pattern: men cluster is occupations concerned with abstract, impersonal systems or that revolve around physical labour and women cluster in professions that involve a relatively greater degree of social interaction. I am not making any claim as to why this is the case, merely that it is.

A good paper (IMHO) that does a thorough analysis of the data on vertical and horizontal segregation can be found here: http://sci-hub.cc/10.1177/0038038511435063

One finding I thought particularly interesting was that the countries with the highest levels of horizontal segregation are Finland, Denmark and Sweden. This seems counter-intuitive to me, though I cannot explain why...

[tptacek]

Three responses.

1. I'm not sure it's surprising, let alone a rebuttal, to observe that there's significant gender disparity in most industrialized economies. In fact, it can be the case that the US leads the world in gender parity (in fact: I think this might be likely) and still simultaneously be the case that there are structural biases against women. Why would we expect to see wildly diverging outcomes in different countries? All these countries (note the exclusion of China!) employ essentially the same pipeline of children to the same set of Professions.

2. As has been repeatedly stated here and in the post we're responding to (which is premised on this observation): pretty much all the other high-status Professions have markedly better gender parity than technology. It may be the case that gender disparity in compensation is broadly the same in societies and that men and women generally cluster into similar compensation bands and occupations, but that could also primarily be a phenomenon of the (majority) of people who don't work in high-status Professions --- because in medicine, law, accounting, actuary, even sales, we have 40-55% female participation.

3. The notion that a study like this rebuts the argument I just made is a sort of flight to abstraction. We can look at women in science (where there is broadly equal representation between men and women in awarded degrees, and markedly better parity among postgraduates and practitioners) and see that CS (and EE) are outliers. I don't need to look at GINI coefficients and treat pay equity as a vector in R^2 to observe the problem, nor does it make sense for me to concede that the debate should shift into this mathematical abstraction. We have the actual facts here; no need for the models.

[spangry]

Addressing your responses:

1. Unless I've misunderstood, the original assertion was that the gender gap is present in many engineering professions. I'm not clear on what facts you are relying on, but the data I've seen (and linked to) seems to bear this observation out. Such an observation is not at odds with your assertion that there are 'structural biases' against women.

And if I has to guess, to put it bluntly, I suspect China was not included as their official statistics on any remotely economic matter are highly suspect. Additionally, there are quite disparate outcomes among countries. On no measure of gender workforce equality is the US in the lead, at least according to the data in the linked study. The US is better than most on vertical inequality (i.e. wage gaps within the same occupational groups) but roughly average on horizontal inequality (i.e. occupational gender segregation). The US is closest to South Africa in these respects.

2. I don't have much to say here really. I guess part of this might come down to what one subjectively defines as a 'high status' profession. I don't really have an opinion here, so I defer to yours.

3. I don't recall saying that that study I linked to rebuts your point of view. I linked to the study because it's fairly comprehensive, informative, and seems to have a sound methodology. The data analysed is fairly credible too: large cross country social surveys and Census data. I'm not sure where R^2 comes into this, as the authors are not constructing a predictive model. Rather, they're simply calculating observational statistics.

And I'm not asking you to concede that the debate should shift into some kind of weird, abstract mathematical exercise. But if you refuse to even glance at some of the data put forward (or any data really), preferring some nebulous facts that you haven't actually presented, then I'm afraid you're likely to have a hard time convincing people who don't already agree with you.