Infectious Disease and Authoritarianism

I am a great admirer of Jordan Peterson and agree with him on most things.  I certainly sympathize with the panic that seized him when his wife was diagnosed with terminal cancer. They were a pair since childhood so the loss was maximal. I myself was pretty upset at the recent loss of a long relationship but I cannot imagine how I would have felt if my  relationship had stretched so far back.

I had support that meant I had no need to turn to anxiolytic drugs but I certainly understand that he did.  As is always the danger, use of such drugs can induce dependence and it is rather heroic that he fought so hard to defeat that dependence.  One hopes that he is back to full health soon.

I am not sure how recent is his article below but it does indicate a mind not at its best. For a start, I can locate no article that is as he describes it. I think he is voicing a garbled memory of a well-known article about parasite load: "Pathogens and Politics: Further Evidence That Parasite Prevalence Predicts Authoritarianism". Parasite load is high in Africa so a number of theories about it have been circulated -- e.g. here.

What it is NOT, however is a disease.  It is not of viral or bacterial origin.  It is caused by invertebrates. So Peterson's recollection that the study caused disease is incorrect.  It is also incorrect that the study was about psychological authoritarianism.  It was about political authoritarianism.

Finally, what can we make of the high correlation Peterson reports.  He is generally pretty good on statistics but it  would seem that he has not heard of ecological correlations. In statistics, “ecological” correlations have nothing to do with environmentalism.  They are "ecological" ones in Robinson's (1950) sense -- i.e. the units for analysis were not indivisible. As Robinson shows, such correlations can easily be beguilingly high, particularly where the units for analysis are few, and such correlations are not estimates of individual correlations.

As Menzel (1950) has pointed out, however, ecological correlations do not have to be estimates of individual correlations to be of interest. Ecological correlations tend to tell us more about broad processes than details within such processes.

The correlation between national IQ levels and national income levels reported by Lynn and Vanhanen are ecological correlations and H.C. Lindgren’s finding of a -.61 correlation between high income and voting for Richard Nixon in the the 1972 U.S. Presidential election is another example. It implied that richer and more highly educated people MUCH preferred the way-out Leftist McGovern

So Peterson is judging the correlation he reports by irrelevant criteria.  An ecological correlation of .7 is mundane, not striking


Lindgren, H.C. (1974) Political conservatism and its social environment: An analysis of the American Presidential election of 1972. Psychological Reports, 34, 55-62.

Menzel, H. (1950) Comment on Robinson's "Ecological correlations and the behavior of individuals" American Sociological Review 15, 674.

Robinson, W.S. (1950) Ecological correlations and the behavior of individuals. American Sociological Review 15, 351-357.

There was a paper published in PLOS ONE about a year ago. They were looking at [the following issue]. Let’s say I assessed your political attitudes—I could do that with, say, an authoritarian belief scale, because authoritarianism has been studied quite a bit since the end of World War II. Nobody really knew what to do with it in relation to personality, but it doesn’t matter; you can assess it with a reasonable degree of accuracy.

These people did two things: they did a cross-country survey and then within-country surveys. So if you were looking at a phenomenon, you could look at the country level—US vs. Canada—or you could go into the US and then you could look at a state level. And it’s nice to do the analysis of both levels, to see if it replicates itself across the two different conceptual strata.

And what they found was mind-boggling. It’s Nobel-prize-winning stuff, as far as I’m concerned.

The correlation between the prevalence of infectious disease in a locale and the degree to which authoritarian beliefs were held in that locale was about point 7. It’s like, you never see that in the social sciences. That’s higher than the correlation between IQ and grades, which is about as good as we ever get in terms of prediction.

So it’s like, really? It’s that high? And one of the things that implies is that one of the ways to get rid of authoritarian attitudes, assuming that you want to get rid of such things, is through public health.


No comments:

Post a Comment

All comments containing Chinese characters will not be published as I do not understand them