It would be interesting to a breakdown by race done well. How do you categorise someone who has parents from different racial backgrounds? What about adopted children? How do you control for other relevant variables?

Typically it's the accusers of racism who make these conflagrations. These are usually well-intentioned progressives who, in their zeal to protect minorities, pounce on anything that could possibly be warped into an attack on minorities. For example, that time James Damore was crucified for suggesting that population level gender differences might better explain Google's hiring pipeline than discrimination.

Please don't take HN into full blown ideological flamewar. The more generic such tangents get, the more unmoored they become from anything specific the thread is discussing, and the more of a black hole they turn into.

Your comment here is emblematic of exactly what HN discussions need not to do. That requires the strength to resist temptation.

https://news.ycombinator.com/newsguidelines.html

I'm taken aback that my comment was "emblematic" of bad behavior. I made my critique charitably ("progressives are well-intentioned..."), and perhaps it was more generic than it could have been, it's certainly not in the bottom half of HN's distribution. I'll try to make my points more substantial going forward.

OP said "Google", not the "US". What's "silly" about thinking males are the majority at Google?

OP said "population level gender differences might better explain Google's hiring pipeline". Men may be the majority of Google employees, but they clearly aren't the majority of the population. That suggests there is discrimination somewhere, either in Google's hiring approach or in wider society (it's unlikely that there is some innate difference between men and woman that would account for such a disparity).

> Men may be the majority of Google employees, but they clearly aren't the majority of the population. That suggests there is discrimination somewhere, either in Google's hiring approach or in wider society (it's unlikely that there is some innate difference between men and woman that would account for such a disparity).

OP was citing Damore's claim about population level differences. Damore's claim wasn't that the US is 80% male so of course Google is 80% male, it's that men and women have different distributions for a given trait (hence, "population level differences") and these differences in distribution might account for Google's demographics better than discrimination. One such population-level difference is that women are far more likely to choose careers that are flexible or to leave the workplace altogether to start a family. By itself, this difference doesn't likely account for the whole disparity, but there are others as well--women prefer to work with people as opposed to abstract reasoning systems. It could even be true that men and women are on average exactly as qualified to be a Google engineer, but that men have a flatter distribution (more representation at the extreme low and high end of the distribution) and that Google samples from the high end, ergo selecting more males. Greater male variability is common in lots of species (including humans) for lots of different traits. It's also interesting that female participation in STEM is inversely proportional to gender equality; countries with very high gender equality (e.g., Scandanavian countries) have relatively few women in STEM while cultures with less gender equality (the middle east, India, China, etc) have far greater STEM participation. It also seems unlikely that discrimination is the cause since women achieved parity in medicine and law in the '80s and '90s when the fields were notoriously, overtly sexist, but the tech industry can't move the needle despite sinking tens or hundreds of millions into "diversity". It seems entirely likely that innate differences play a significant role.

There will undoubtedly be a lot of (otherwise very intelligent) people who will read "population level gender differences" and interpret it as "Wait, you're saying X is true for literally every woman?! How can you generalize like that?!". Of course, this isn't what I'm saying or what the OP said or what Damore said, so let's not go down that straw man rabbit hole.

OP said "population level gender differences might better explain Google's hiring pipeline" referring not to the population of the US, but the Pool of Engineering students.

So given we aren’t talking about the actual demographics of the regions they hire in, why not narrow it further?