> The underrepresentation of women in academic science is typically attributed, both in scientific literature and in the media, to sexist hiring. Here we report five hiring experiments in which faculty evaluated hypothetical female and male applicants, using systematically varied profiles disguising identical scholarship, for assistant professorships in biology, engineering, economics, and psychology. Contrary to prevailing assumptions, men and women faculty members from all four fields preferred female applicants 2:1 over identically qualified males with matching lifestyles (single, married, divorced), with the exception of male economists, who showed no gender preference.
Encouraging everyone equally is great, but the problem is that many of these advocates seem to operate under the unsubstantiated hypothesis that there are no correlations between gender and career choice except those caused by upbringing. This leads to the idea that an unequal representation is necessarily a sign of a problem that needs to be fixed. I don't think this is supported by the data. We know from biology and other fields that there are natural differences. We also know that in more emancipated societies where people have more opportunity to choose what they want, there is a greater gender imbalance in universities (example: highest % of women earning science degrees is Iran; no western european country nor the US nor Canada is even in the top 20). Perhaps a gender imbalance is a sign of a society where individuals can do what they really want.
"Encouraging everyone equally is great".
We should probably just end the argument on that point we both agree :). The rest will work itself out. If everybody was encouraged and inspired equally and there were still imbalances your point will remain. Unfortunately, that's not true today, so we won't know.
Iran and India have much higher representation of women in STEM fields, and while they have other issues with misogyny, the inculcation of ambition (across gender lines) is very high (I say that having lived in India). The result seems to contradict any "biological" predilections.
But we'll see.
I don't see how it does. If I were to guess I'd say that in those countries there is a greater economic necessity to go into a high earning field in order to earn freedom than in Sweden and Norway. In Sweden and Norway people can follow their passion instead, without having to worry too much about the pay.
It seems inadvisable to me to assume that all differences must be due to socialisation and therefore indicative of a problem that has to be fixed, until the time that research shows this to be the case. Otherwise we risk implementing policies with incentives that make people less happy overall, by incentivising careers that people don't really want to go in. At the very least it should be a warning sign if your metric for gender equality shows Iran at the top and Sweden at the bottom.
Because currently, a 'passion' for STEM fields will tilt your future power prospects. If there is a systematic bias that causes a tilt the other way, it is simply untenable.
Or we could try not generalising outside the scope of an experiment. The experiment doesn't even cover all faculty hiring.
(a) Hiring faculty is a very different thing (and subject to different rules and policies) than getting students succesfully to graduation.
(b) Unless male economists are uniquely meritocratic there's more at play here than personal bias. It seems more likely that the other subjects have a culture of progressively trying to redress historic gender imbalance. Since those preferences seem to occur at a subject/departmental level it seems even more unwise to try to generalise off the back of them.
(c) If you do want to generalise, there are other more generalisable studies that contradict the results of this one.
You can speculate on the reason why the discrimination exists, but calling it "a culture of progressively trying to redress historic gender imbalance" is again a collectivist attitude, as if the treatment of somebody's grandmother somehow confers a debt that must be repaid to their granddaughter but not to their grandson, because the granddaughter is in the same tribe as her grandmother. From an individual perspective it's just discrimination. According to the study an individual human has a 2:1 hiring advantage compared to another individual human, simply because of an attribute that should not significantly factor into the decision. It doesn't suddenly stop being discrimination because the advantaged gender happens to be the female one. By the way, I think this is particularly damaging for the talented women who were hired on merit, because it becomes rational to believe that they they might not have been hired on merit.
> The drop out rates for males are (much) higher.
Seems off-topic, can you explain the relevence of this? There are serious problems with boys education but they don't seem to have a problem with discrimination in the workplace, see [0][1][2][3][4].
> ... simply because of an attribute that should not significantly factor into the decision.
That's an opinion I think there's reasonable evidence to discard, or at least add some nuance to.
Diverse workplaces are more pleasant to work in, and in the case of universities it seems reasonable to believe (based on the work of Carol Dweck & others) that having relevant female role models would improve outcomes for female students in STEM subjects.
Additionally there's a growing body of evidence that documents how diverse workplaces have higher productivity; and gender gaps in entrepreneurship negatively affect GDP (Cuberes & Teignier for a 2015 example http://www.economist.com/news/americas/21661804-gender-equal...)
[0] Female students seen as less competent than identical male students with identical application materials, offered lower starting salaries (Moss-Racusin et al, 2012)
[1] People in gender-incongruent roles penalised more heavily for mistakes (Brescoll, Dawson, & Uhlmann, 2010).
[2] Voluble women perceived as less competent and less suitable leaders, inverse true for men (Brescoll, 2011).
[3] Women who succeed in male-dominated fields percieved as not likeable (Heilman et al, 2004).
[4] Students question the competence of female teachers who evaluate them negatively, less so than male teachers (Sinclair & Kunda 2000).
You said:
> Hiring faculty is a very different thing (and subject to different rules and policies) than getting students succesfully to graduation.
That's why I mentioned it.
> Diverse workplaces are more pleasant to work in
If somebody's opinion was that all male workplaces are more pleasant to work in, would you find that a reasonable basis for discriminating against female applicants?
> Additionally there's a growing body of evidence that documents how diverse workplaces have higher productivity; and gender gaps in entrepreneurship negatively affect GDP (Cuberes & Teignier for a 2015 example http://www.economist.com/news/americas/21661804-gender-equal...)
Not a very convincing study. They just noticed that women work less than men, and calculated the increase in GDP if women worked as much as men. Note that from the data they use it is apparent that women work more in poorer societies. So maybe women work less in richer societies because they can. This is a good thing, because it means the things people do are dictated more by their passion and less by economic necessity.
The other studies are all about perception. I wonder how much of it is due to the effect that I mentioned (if you confer a selection advantage to a group then it becomes rational to believe that post selection individuals in that group are less competent). I'm also a bit wary of publication bias. I've heard from people in the field that there are certain results that are good for your career and certain results that are not.
Do you know of any reliable studies showing actual discrimination of the magnitude that the study I cited showed?
Either knowingly or unknowingly you're abusing both the study's findings and the scientific method as a whole.
We can talk hypotheticals or debate the relative merit of affirmative action, or discuss the flaws in specific studies (because let's face it perfection is hard in social science), but first we need to address why you feel the need to misrepresent the findings of studies. Did you not read the abstract you linked to?
[1] being a link to Steinpreis/Anders/Ritzke (1999)
> You're right that it's biased, but perhaps not in the way you think it is: http://www.pnas.org/content/112/17/5360.abstract
At risk of repeating myself... You've taken a study that looked specifically at tenure-track faculty hiring within four STEM subjects, i.e. a very specific, limited, scope. You then presented that as "evidence" in a general context (in reply to a point about hiring in general being biased against women) in order to argue that the hiring process is biased against men, without any mention of the limited scope of that study.
The way I see it, one of these things is true;
(a) You made a mistake, fair enough, no big deal
(b) You're unfamiliar interpreting studies, fair enough, no big deal
(c) You're misrepresenting science to further an agenda
Given your vociferous objections to affirmative action I'm guessing (c) but happy to give you the benefit of the doubt.
Why does this imply that meritocracy is not great?
I mean, a noble family really did have a greater relative contribution to the economy and to society than a common family (or, at least, could have been said to have a greater contribution by the norms of the time) -- until it didn't, and that's when things changed. It's not like previous societies realized that they're not meritocracies; they utterly and truly believed that their system was the best possible one (and they could prove it!) and that there are people who are less represented in positions of power is just bad luck. No society (that lasted long) has ever said, let's put all the people with real talent at the bottom. It's only when resources were allocated differently due to political action (or natural disaster, like the black death) talent "suddenly" grew where everyone had until then known little talent could be found (and they were right: there was little talent because talent requires cultivation to become effective).
Right, so that's not bad luck. That's systemic bias.
What makes correlated bad luck somehow something to be eradicated (above and beyond ordinary bad luck)?
BTW, it's pretty funny that you call systematic bias a "label", yet bad luck is something that seems sufficiently explanatory to you. It seems to me that "bad luck" is a label that you apply to something that you don't wish to study because you just don't care. It's bad luck that humans can't fly. It's bad luck that people die of the flu. As pretty much everything humans do is try to eradicate things that could be labelled as bad luck, actually using that term is essentially yelling "I don't care about this! I like things as they are!"
I can easily provide reasoning for why study of flu (or other disease) is worth studying: https://news.ycombinator.com/item?id=8369801
Can you do the same (or even provide a coherent definition of) systemic bias? If so, you may convince me it is worth studying. Im unconvinced it is even well defined, since I've never seen a definition that advocates of it consistently apply.
Easily! But just as your reasoning is completely dependent on an axiomatic value (death is bad), so is mine: That many people's freedom to cultivate, explore and express their talents is hindered by those in power (positive claim) is bad because freedom is good (normative claim), therefore we should study how those restrictions work so we can remove them.
But don't get me wrong: the distribution of power will change (it always does), yet inequality will grow again (it always does; a dynamic system that is not perfectly balanced and that includes positive feedback loops will always go out of balance) and this process will be repeated. I know this because pretty much all of history has been this way: inequality grows; social strife grows; people demand change; others first politely and then less so explain to them that that is scientifically/theologically impossible or that things are best as the are; after a lot of struggle things change and power shifts; inequality grows and so on and so forth.
> Im unconvinced it is even well defined, since I've never seen a definition that advocates of it consistently apply.
You are right in the sense that "systemic bias" is a term invented to explain to laymen what "bias" is. To social researchers, it's clear that all social bias (where "social" means "not exploding volcano in one town and not in another") is "systemic" (whatever that means). It really means nothing more than "a result of social dynamics disproportionately directed by those with power", and even that is a definition for laymen because obviously, that's what power is, so it really means "a result of social dynamics". But even that is redundant because every social phenomenon is a result of social dynamics (that's what a social phenomenon means), so it really just means "bias". Personally, I hate the term "systemic bias" because "systemic" may suggest either a bureaucratic bias or something dictated by law or decree.
If all people got equal encouragement, inspiration, and opportunity (from the moment of birth) then we could simply focus at the moment of judgement (which btw, is also currently biased [1]). So that is basically what STEM advocates are working on, making sure women and other disadvantaged minorities have equal encouragement, inspiration and opportunity (specifically to learn) so we can claim to have an equitable meritocracy.
[1] http://advance.cornell.edu/documents/ImpactofGender.pdf