Hacker News Reader

Comment by cousin_it | Hacker News Reader

Truth isn't the only thing that can hurt, though. Ignorance and malice can hurt too, especially when combined... Let me try to explain how I understand this issue.

When some subgroup is underrepresented in some area even though there's no statistical difference in innate ability, many people tend to automatically blame discrimination. I don't see how that can be right. For example, you could notice that Asians are overrepresented in US universities. The theory concludes that either all US universities are racist in favor of Asians (obvious bullshit), or Asians are innately smarter than whites (more obvious bullshit). Hmm. Isn't it more likely that the theory is mistaken and there can be other factors at work?

Same goes for the small number of women programmers. Sorry, you don't get to attack me for sexism until you have data proving sexism is the real culprit. Otherwise it's like blaming witchcraft.

View on HN

We have data. A brief sampling: http://www.nber.org/papers/w15853.pdf, http://dl.acm.org/citation.cfm?id=1508924, http://www.advancingwomen.com/awl/fall2002/FERREIRA/FERREIRA..., http://www.scribd.com/doc/18370009/Sources-of-Financing-for-...

There is plenty more out there, all suggesting that women are promoted differently, treated differently, more likely to experience active gender-based harassment and are more likely to be excluded from the camaraderie and informal networking that leads to job satisfaction. Don't assume that just because you are ignorant everyone else is too. Some of us have been engaging with, measuring, quantifying and attempting to change these destructive culturally-sanctioned behaviors for a very long time.

Only the first of those four articles deals with women in IT jobs. It is behind a paywall but I found a freely available copy here: http://www.mcgill.ca/files/economics/leavescience_all.pdf . It's a nice article. I don't think it supports laying the blame on sexism, though.

The paper focuses on something called the "exit rate", that is, how many people with a certain education are working at a job unrelated to their education. The average exit rate over all professions and both genders is 21%. Female engineers have a pretty low exit rate of 12.9% - that is, 12.9% of women who studied engineering now work in other areas. Male engineers, on the other hand, have an extremely low exit rate of 9.8%.

The author then tries to explain the difference between "pretty low" for women and "extremely low" for men. She observes that the differentials in exit rate, across all fields including engineering, can be mostly predicted by the gender makeup of the field. Quote from the paper's conclusion:

> The results would appear to point to problems for women specic to the engineering profession. However, I show that the excess exits of women trained as engineers, as well as their excess exits for pay and promotion reasons, are no larger than would be expected given the share of men in the field: the higher the male share, the greater the excess female exits from the field, both in total and for pay and promotion reasons.

Then the author hypothesizes that improving mentoring and reducing "possible discrimination" might help reduce the gap, without giving any specific arguments.

It seems to me that the really important figure in the paper is 9.8%, not 12.9%-9.8%. That is, more males are behaving abnormally by staying programmers despite possible problems with pay and promotion, while the behavior of female engineers is closer to the average rational human being. What do you think?

Also, I'd be really interested in seeing what you think are the best papers proving that the lack of women of IT is due to sexism. Taking apart random vaguely related papers is fun, but I don't have infinite time.

If you don't have time to bother to learn about a topic, why do you assume you are more knowledgable on it than people who do spend their time studying the topic at hand? Your theory boils down to "Tide goes in, tide goes out. Never a miscommunication. You can't explain that." You can "take apart" as many papers as you want to maintain your ignorance, but all you do is make yourself irrelevant to the discussion.

If you want to "prove" that the disparity is due to men and women experiencing different cultural dynamics in male-dominated spaces (which is why research regarding male-dominated scientific fields and financial firms are relevant to the question at hand) all you need is the Census Bureau's labor statistics. The interesting research questions are what specific dynamics are at play and how we can disrupt them.

Lest you assume my contemptuous response to your willful ignorance can be ignored due to a [citation needed] tag, one paper that perhaps offers more of the background you are looking for would be the FLOSSPOLS overview of their gender-related findings in Open Source communities: http://www.flosspols.org/deliverables/FLOSSPOLS-D16-Gender_I... For background on the mechanisms involved on a micro level I'd suggest the book Thinking, Fast and Slow, and on a more macro level specifically regarding gender, I recommend Pink Brain, Blue Brain.

If you want to claim that the highly statistically-unlikely and culturally-variable gender disparities in pay, promotion and participation are not due to sexist dynamics, you need to provide both an alternative theory and evidence to support it. Given that the only evidence required to prove that sexist behavior is present in technical spaces are the comments to this article, the burden of proof is on those who disregard the published research on the topic and claim those behaviors are either isolated aberrations or have no effect on women's willingness to be employed in the field.

About the labor statistics: if you say they're all I need, it seems that you dismiss the argument I made above, that disparity of outcome does not imply discrimination all by itself. Why?

About the paper: it looks much more sloppy to me than the previous one you linked. This quote from p.12 is at the root of my dislike:

> Unlike the survey data, where claims to generality can be made via mathematical operations, generalising from ethnographic work is done via interpretation as in the humanities.

The humanities haven't been scientifically tested, so we don't know if their method leads to truth. For example, on pp. 16-17 the authors note that many women are computer users but few become IT professionals - and immediately conclude, from that fact alone, that hostile work environment must be the reason. It's fishy that there are no intermediate steps of reasoning and no alternative hypotheses. There are many such examples in the paper.

About the books: I'm already a big fan of Kahneman. Why specifically do you mention his book as support? Pink Brain, Blue Brain seems to deal with the development of children rather than with sexism in mostly-adult fields like IT, and if its conclusions are right, then sexism of male IT workers is not required to explain the small number of women in IT because differences in development are already enough. Why are you mentioning it as support?

Also, do you disagree with what I said about the previous paper? It's a little frustrating when people abandon lines of argument to start new ones.

There is an interesting issue in the philosophy of science regarding the "hard center" of theories. Basically any prediction about a reasonably complicated situation will rely on a large number of premises.

If you happen to be committed to one premise more than the others, you will always blame one of the other premises when an observation doesn't fit your hypothesis.

In the case of sexism there are so many confounding factors that you'll never be able to say with certainty that sexism is the root cause, only that there is a good chance that it is.

If there is a good chance that something bad is happening then it seems like you could reasonably be expected to produce a non-zero amount of effort in avoiding it.

Reply on news.ycombinator.com