Someone was crowing that Jonathan Haidt's finding that the younger and more liberal you are, the more likely you are to have been diagnosed with a mental health condition was wrong, but claimed that he was unable to show proof because of the conditions of the licence (the licence just says "you may not download, republish, retransmit, reproduce or otherwise use any such content as a stand-alone file").
He was probably hoping that no one would actually look at the data and expose him (he even taunted me with "It will take some expertise to open and process an SPSS file, but surely you're clever enough to do that?"), but too bad, I did (of course, if Haidt had really been lying, it's curious why no one had exposed his academic fraud for years).
His lies were really brazen.
He claimed "This survey didn't even feature the question whether the respondent is a conservative or liberal" and that "it didn't ask the person if they ever went to a doctor to talk about their mental health (which must be a prerequisite for any statistical conclusions like the one you did)"
The first claim is categorically false, because if you download Pew's American Trends Panel Wave 64 dataset, the variable F_IDEO shows the respondent's ideology. It's true that in the Wave 64 questionnaire, they were not asked about their political ideology - but that's because this is panel data and they already have this data from previous waves.
The second claim might be true in a very narrow technical sense, but the question asked (which you can verify in the questionnaire in the PDF in the dataset download) was "Has a doctor or other healthcare provider EVER told you that you have a mental health condition?" But what we're seeking to measure here is not mental health awareness - whether people go to the doctor to talk about their mental health - but mental health state (whether people have mental health conditions).
But the biggest lie is about what the data show. And they indeed confirm Haidt's charts:
- The older you are, the less likely you are to have a diagnosed mental illness
This is interesting because this variable is monotonic - once you have been diagnosed with a mental health condition, you can never truthfully say no to the question: "Has a doctor or other healthcare provider EVER told you that you have a mental health condition?"
You would expect older people to be more likely to have been diagnosed as such in their lifetime since they have been alive longer, yet young people are more likely to answer this question in the affirmative
Of course, the cope is that older people are just in denial, or don't seek medical attention for mental health issues. But if 41% of a population (young, very liberal women) report having a mental health diagnosis, this suggests that if the concept of mental health is to have any meaning at all, there must be overdiagnosis in some form, or it makes a mockery of the concept of mental health. In other words, if everybody is insane - no one is insane.
- The more liberal you are, the more likely you are to have a diagnosed mental illness
Naturally, the cope here is similar to the above. But we see a clear dose-response effect across sexes and ages - the more liberal you are, the more likely you are to have a diagnosed mental illness (with the possible exception of going from Very Conservative to Conservative). So for this cope to work, left wingers need to throw Moderates and Liberals under the bus - not just Conservatives - and claim that they are in denial as well. Apparently only Very Liberal Women are in touch with their own mental health - and everyone else is just suffering from false consciousness.
Furthermore, we know that liberals are less happy than conservatives, and that happiness has an inverse relationship with mental health (the original poster claimed that "Jon Haidt's piece was already debunked in a research paper. And in the most obvious way" but besides never making that claim, the paper he cited also cited many articles about liberals being less happy than conservatives). So the further cope needs to be that conservatives are not just refusing to go to the doctor - they are also in denial about being happy. But of course, any "theory" that can make conservatives look bad is a good theory.
- Women are more likely to have a diagnosed mental illness than men
Across basically all age groups and ideologies, women consistently are more likely to have a diagnosed mental illness than men.
Once again, one cope is going to be similar to the above, but another is going to be about "patriarchy". Yet, the gender equality paradox reveals that increasing gender equality is actually associated with worse female mental health (relative to male).
This supports the hypothesis that grievance politics and left wing ideology drive poor mental health (rather than objective oppression).
For a robustness check, I sliced and diced the data some more.
Looking at the data by race, black non-Hispanic respondents pretty much have similar levels of diagnosed mental illness regardless of ideology. Hispanic conservatives, very conservative and moderate respondents have similar levels of diagnosed mental illness, but liberal and very liberal Hispanics have significantly worse mental health. But it's white people who are really driving the ideology effect.
So for the earlier cope about conservatives being in denial or refusing to go to the doctor to hold, left wingers need to throw black liberals under the bus now, since black very liberal respondents (sample size of 76 notwithstanding) have almost half the mental health diagnosis rate of black liberal respondents. And liberal and very liberal Hispanics also have lower rates of diagnosed mental illness than their white fellow travellers.
Of course, one could come up with further cope about racial disparities in healthcare access but this would work in the opposite direction from the oppression-as-a-cause-of-mental-health-issues claim. The layers of cope are starting to collapse on themselves.
