“Men obstructed from entering female-dominated occupations,” oh my god you guys we did it. Sexism against women is over! The feminists have upended patriarchy and instituted matriarchy. Ha ha, we said we wanted equality but what we really meant was that we wanted more. MORE!
Okay no, let’s list a few facts before we dig into this study:
1) Regardless of this study’s results, I have no doubt that some (if not most) men are regularly discriminated against in some (if not most) female-dominated workplaces.
2) There are many ways to discriminate against a person in a job. This study looked at exactly one.
3) This study was done in Sweden, which, according to the World Economic Forum is the 4th best country for gender equality in the world, just above…I’m sorry, Nicaragua? I…okay that’s like a whole other video I’m just going to put a pin in that one.
I mention the Sweden thing because we live in a US-centric society (especially those of us in the US) and so we tend to assume every study is about us. For context, while Sweden was 4th best for equality, the WEF ranked the US 53rd, just ahead of…Singapore? Really? We beat Singapore? Damn y’all I have a lot of internal biases about various countries’ gender equality issues that I need to examine. Again, not for this video.
Okay, so this study was just published in PLOS One and it looked at pre-existing data from three other studies (all of which include the same main author as the new author) that together submitted more than 3,200 fake job applications from fake men and women to real employers in 15 different industries, some of which were female-dominated and some of which were male-dominated. They found that male-dominated industries were just as positive about female applicants as they were about male applicants, but female-dominated industries were significantly more positive about female applicants than they were about male applicants.
Interesting! So, let’s revisit my earlier points. There are many ways to discriminate against someone in their career. You can tell a child that they are not suited to do a particular job; you can grade them more harshly; you can steer them towards different college majors; you can put them in a major that is dominated by people who are all one gender or race or ethnicity or sexuality that they don’t share; you can treat them poorly in class; you can reject their job application; you can give them a hard interview; you can not hire them; you can hire them but treat them like shit; you can pay them less; you can give them less perks or vacation time; and you can fire them for something you wouldn’t fire someone of a different gender or race or ethnicity or sexuality for.
This study looked specifically at this one: you can reject their job application. This study did not show that women were just as likely as men to be hired in male-dominated industries — only that they were just as likely to be invited for an interview.
Which is still something! A seat at the table is one important step, and it is unconscionable if Swedish men aren’t being offered that opportunity in female-dominated industries.
But…”if.” Does this study prove that that’s what’s happening? Well, it’s a bit more complicated than you may think. Because one-third of this study’s data came from a study on how convicted criminals fare in a job search. Not just “men” versus “women” but “men who were convicted of an assault and are currently doing community service” versus “women who were convicted of an assault and are currently doing community service.”
We already have a more complicated problem to deal with: people may be more likely to view a man convicted of assault as being more violent than a woman who is convicted of assault, and that prejudice may have basis in reality, considering that in the United States about half of all assaults committed by men are “simple assaults” (the least serious charge) compared to 75% of women.
Now add to that the fact that the “female-dominated” industries these ex-convicts applied to included “preschool teacher” and the male-dominated industries included “truck driver.” Gosh, I wonder which one would be more likely to discriminate against a man accused of assault?
And to be clear, that study on ex-convicts in the hiring process is actually quite good. The problem happens when they follow up with this new paper that strips the data without context, and uses it to come to conclusions that aren’t necessarily warranted. In the original study, they also looked at how gender affected the applicants, only instead of just saying “men vs women” in “male-dominated vs female-dominated industries” they also broke it down based on whether or not a job was “high skill” or “low skill”. While they did find that “female ex-offenders face unequal treatment in the hiring process just as male ex-offenders do,” they also found “that discrimination against ex-offenders was most prominent in female-dominated and high-skilled occupations, supporting our conjectures that we derived based on the contact hypothesis. We argued that since criminals are more likely to be men and less educated, the probability of coming into contact with ex-offenders would be higher in a male-dominated population and among low-skilled groups. We therefore conjectured that employers in male-dominated and low-skilled occupations would be more familiar and have higher acceptance towards ex-offenders. In fact, our results showed that ex-offenders were not discriminated against in any of the male-dominated low-skilled occupations included in our study (i.e., auto mechanic and truck driver).”
Boom. Context. It doesn’t make for a great headline, but it does help us understand a complicated problem better.
So that was a full 1/3rd of this new study’s data. Another third came from a study that looked at victims of crime, which was also evaluated in view of gender and also had more context in that it also looked at “high skill” versus “low skill.” They found that female victims were discriminated against in both female- and male-dominated fields regardless of skill level while male victims were discriminated against more in specifically female-dominated high-skill positions.
The last third came from a study on transgender candidates, in which they discarded the transgender candidates and kept the control group. That’s literally the only dataset that should be considered here — they aren’t ex-convicts, they aren’t victims, they’re just people who aren’t volunteering uncomfortable things to potential employers in their introductory emails. “Hey I’d love to be your new preschool teacher, by the way I DID assault someone last year but that’s ancient history. Anyway I had a 3.4 GPA…”
It’s funny because all of these studies share an author, so I’m not trying to say that Ali M. Ahmed of Linköping University is a bad researcher. But I am suggesting that maybe this is one of those cases where “publish or perish” has a real negative impact on the quality of research that ends up making headlines. By pressuring scientists to constantly publish new papers or else get fired, we encourage them to do things like this: going back to your own old data, shuffling it around, dropping context, and seeing if you can come up with something statistically significant to support a new hypothesis to buy you another year or so of not getting hassled for not publishing.
That’s not to say researchers shouldn’t use existing datasets — that happens all the time. But it needs to be done very carefully, with hypotheses and statistical models that are pre-registered so we know that the scientist isn’t just cherry-picking to suit a new purpose.
And again, just to reiterate, I do not doubt that there’s discrimination against men in female-dominated industries. But if we want to fix that, we need better studies to inform policies that will lead to true equality in the workplace.