Zombie Beliefs - 0, Reality - 2

The hits keep on coming for ideological zombie beliefs.

In this instance, the firmly established belief that women are systematically and persistently discriminated against because of their sex. As always, for the statistically impaired, the fact there may be no evidence of institutional bias does not mean that men and individual women don't face or experience a number of biases and discriminations. It only means that to some degree that they cancel out (biases for and against) and/or that differences in outcomes are traceable to causes other than bias.

In this case, the belief out of postmodernist/critical theory ideology is that women suffer institutional negative discrimination based on their sex. The claim manifests in a belief that women are paid less for the same work (refuted for thirty years), that there is a campus rape crisis fostered by academia's toxic masculinity, and that women are selected for jobs, prizes, and opportunities at a lower rate than is explainable by other factors.

From No evidence of any systematic bias against manuscripts by women in the peer review process of 145 scholarly journals by Flaminio Squazzoni, Giangiacomo Bravo, Pierpaolo Dondio, Mike Farjam, Ana Marusic, Bahar Mehmani, Michael Willis, Aliaksandr Birukou, and Francisco Grimaldo. From the Abstract:

This article examines gender bias in peer review with complete data on 145 journals in various fields of research, including about 1.7 million authors and 740,000 referees. We reconstructed three possible sources of bias, i.e., the editorial selection of referees, referee recommendations, and editorial decisions, and examined all their possible relationships. In line with previous research, we found that editors were sensitive to gender homophily in that they tended to match authors and referee by gender systematically. Results showed that in general manuscripts written by women as solo authors or co-authored by women are treated even more favorably by referees and editors. This is especially so in biomedicine and health journals, whereas women were treated relatively less favorably in social science & humanities journals, i.e., the field in which the ratio of female authors was the highest in our sample. Although with some caveat, our findings suggest that peer review and editorial processes in scholarly journals do not penalize manuscripts by women. However, considering the complex social nature of gender prejudices, journals should increase gender diversity among reviewers and editors as a means of correcting signals potentially biasing the perceptions of authors and referees.

I love the last line of the catechism. Might be summarized as Even though we find no empirical evidence of discrimination, we recommend that we keep creating positive discrimination on behalf of women just in case. If that is not an insulting and patriarchal white knight attitude, I don't know what is. We can't find any evidence of discrimination but the poor dears need looking after.

Bah. Bring back Classical Liberalism where evidence counts, people are granted full natural rights and equality and are judged based on the content of their character and not their sex or color.

This one is even more interesting. From The persistence of pay inequality: The gender pay gap in an anonymous online labor market by Leib Litman, Jonathan Robinson, Zohn Rosen, Cheskie Rosenzweig, Joshua Waxman, and Lisa M. Bates. Another study based on an anonymous market where discrimination based on sex cannot occur and yet where there are gender differential outcomes despite the infeasibility of discrimination based on gender. From the Abstract:

Studies of the gender pay gap are seldom able to simultaneously account for the range of alternative putative mechanisms underlying it. Using CloudResearch, an online microtask platform connecting employers to workers who perform research-related tasks, we examine whether gender pay discrepancies are still evident in a labor market characterized by anonymity, relatively homogeneous work, and flexibility. For 22,271 Mechanical Turk workers who participated in nearly 5 million tasks, we analyze hourly earnings by gender, controlling for key covariates which have been shown previously to lead to differential pay for men and women. On average, women’s hourly earnings were 10.5% lower than men’s. Several factors contributed to the gender pay gap, including the tendency for women to select tasks that have a lower advertised hourly pay. This study provides evidence that gender pay gaps can arise despite the absence of overt discrimination, labor segregation, and inflexible work arrangements, even after experience, education, and other human capital factors are controlled for. Findings highlight the need to examine other possible causes of the gender pay gap. Potential strategies for reducing the pay gap on online labor markets are also discussed.

Once again with the ideology - they find no discrimination, only differences in outcomes based solely on individual choices. Even though there are no identified barriers and even though people's outcomes are driven by their choices, the researchers make the normative argument that more research is needed to ensure that everyone has the same outcomes regardless of their personal choices and preferences. The totalitarian mind is relentless in its opposition to variety and freedom.

This second study is comparable to the research from Uber a couple of years ago. There is no human intervention in the dispatching of Uber drivers. The system does not know or care about gender or race of driver. Outcomes are entirely and only driven by driver choices and actions.

In such a non-discrimination environment, men still earn more than women. My recollection was that it was about 7%. Not because they are men and favored but because 1) they stick with Uber longer and understand the dispatching algorithm better, 2) because they drive just a little bit faster than women, and 3) they make more revenue optimum ride choices (where and when to drive).

In other words, there is disparate impact entirely and solely attributable only to individual choices.

Together, these studies suggest that there is a 7-11% variance between male and female market outcomes which cannot be attributable to discrimination. That is a useful boundary to know.

Some years ago I looked at the data across multiple forms of market achievement (promotions, awards, income, etc.) and found persistent gaps of about 15-30%. In most fields, women in the US make up 15-30% of the top earners, top promotions, top award winners, etc. And that was far better than every other country. Wage gaps generally ran from 10-30% depending on the field.

However, when you normalized the data to take into account half a dozen obviously causal factors such as years of experience, hours worked per week, degree of specialization, choice of profession, etc. the gaps all shrank to almost nothing. For example, the percentage of men and women working in a law firm working 70 hours a week, having worked for the firm for ten years, being on call for all hours and all location work, and with a degree from a premier university was identical. I.e. if you had those attributes, then you were virtually certain to be promoted regardless of whether male or female. The problem being that far fewer women chose to stay with a single firm for ten years, work 70 hours a week or were available for all work, all hours, and all locations.

But in the income data on wage gaps, there was always a 2 - 6% variance. The big factors erased most of the gap and the remaining gap was so small that it suggested that adding a further variable to the big obvious ones would simply reduce the gap further. But it obviously left open the possibility that there is some small degree of discrimination.

When the original claim is that discrimination causes a 40% wage gap and you ID five obvious explanatory variables and they do indeed shrink the gap to 5%, it is tempting to simply say that there is no discrimination and if you keep adding variables then the gap will shrink further. Which is likely true but until it is actually done, the case remains open. Gathering clean, reliable, consistent data is hard work and expensive and so there is not much of an incentive to spend more to prove less and less.

But it is a nagging loose end.

These two studies suggest that there is always going to be a 7-11% fudge factor to income gaps that have nothing to do with discrimination. If true, that potentially explains the small residuals in the earlier work.