An interesting study missed my snake-eye view in early June. It's really a group of studies, done in different ways, but all on the same topic: How married heterosexual men with stay-at-home wives differ from married heterosexual men who have wives who work (in the labor market) when it comes to their views about women in the workforce.
Turns out that the former men express more sexist sentiments about women than the latter group of men. The crucial question, for the researchers, is the chicken-and-egg problem: Do men with more sexist values choose to have marriages where the wife stays at home? Or does having a wife who stays at home make men more sexist in their opinions about women at work?
Perhaps both are true at the same time?
All the sub-studies are worth having a look at, but I'd like to address only one of them, the fourth study in the paper*:
Studies 1-3 focused on passive attitudes of men in different marriage structures toward women. In Study 4, we examined whether these men would actively engage in actions that would prevent women in the organization from advancing their careers. A second goal of the study was to examine our hypothesis using a sample of men who might be accustomed to making important decisions—managers. To this end, we conducted a controlled quasi- experiment using male managers that were married and working full time. In this study, we examined if compared to men from modern marriages, men from traditional marriages deny qualified female employees opportunities for promotion (Hypothesis 4).
I picked this study for closer scrutiny because the subjects in it are managers which means that they might have real power to influence how their subordinates fare. Also, this study is a pretty neat experiment. It consists of telling the 232 male accountants in the study that they are to judge the desirability of a candidate for a fictitious but fantastic MBA program (on full pay) and a consecutive promotion at work. The study subjects were asked to either recommend or not recommend the candidate for admission in the program, and they were told that accuracy is very important.
I bet you can guess what the researchers did next. That's right. Roughly half of those 232 accountants were given a package of information about a candidate called Diane Blake, the rest were given the same package but the candidate's name was David Blake.
I love these types of studies because they control for the possibility that David might, in fact, be the better candidate. Given that the information about Diane is the same as the information about David, an unbiased assessment should find the two candidates equally good.
And indeed that is what happened in the current study. A test attempting to establish whether the study subjects thought one person was more qualified than the other person showed no difference. Thus, the study subjects did not attribute lower skills or ability to Diane or David.
Given this, it is fairly shocking that men with stay-at-home wives were significantly less likely to regard Diane a good fit with the MBA program than David. Remember that this doesn't seem to arise from the kind of sexism where Diane is viewed as less able (even though she cannot be, given the same information for both candidates).
So what's going on here? It could be that the men with stay-at-home wives believed that the fictitious Diane would be more likely to quit for family reasons than the fictitious David. But as both pretend-people already had 25 years of work experience this seems unlikely to be the main explanation.
Whatever the proper explanation, this study suggests that managers with certain gender values could harm the career prospects of women. Those values correlate with having what the researchers call "a traditional marriage."
*Ever since I got the Mac I've had trouble with linking to pdf articles. Let me know where I can find the link, please. Also note that the methods in this paper come across as somewhat basic (off the statistics textbook) to me. This is less of a problem because the different studies use different methods. Thus, the findings are unlikely to be caused by simple methodological problems. But one should be careful about not imputing too much power to the findings.