Autism, Libertarianism and the Political Fringe
In 2018 the Duke University historian, Nancy Maclean, was asked what motivates the thinking of libertarians and public choice theorists, economists who apply economic models to the study of governments. She said that their apparent lack of empathy and desire to reduce human societies to mathematical models is indicative of autism.
Although the professor’s comments created a backlash in the media, the idea that libertarians are autistic is a popularly held view. A simple Google search for “Are Libertarians autistic?” will lead you to many online forums with people making the same argument. This opinion is not without merit. A decade ago, Psychologist Jonathan Haidt found in a survey that libertarians score higher on systemising traits and low on empathising traits, which is a key characteristic of autism.
Despite libertarians being apparently autistic, no one has surveyed libertarians to see how many are actually autisitc. But this can be done. Scott Alexander, of the blogs Slate Star Codex and Astral Codex ten, runs near annual surveys of thousands of his readers, which asks them questions such as whether they are autistic or what their political affiliation is. I use the 2020 survey which is the most recent. Older versions allow people to write in their political ideology, rather than simply choosing it from a list, making this data more difficult to deal with.
I use two measures of autism from the survey. The first measure is whether the respondent has a medical diagnosis of autism. The second measure is a broader indicator of having autistic traits and includes everyone who has a medical diagnosis, has autistic family members or thinks they might have autism.
In the SSC survey we see that Libertarians are comparatively not that autistic and in a similar range to Liberals. It should be noted that the SSC survey is not an unbiased sample of the population - Scott Alexander’s readers have very high rates of autism, 4.2% have a medical diagnosis compared to a US average of 2.2%. And 31% of surveyed readers have a medical diagnosis or autistic family or think they might have autism. Libertarian readers have about average rates of autism for SSC readers. With mainly autistic readers, I suspect this sampling bias has hidden Libertarian’s true autism advantage of Liberals and Conservatives.
Another explanation is that over the last ten years we have seen the Libertarian to Alt-Right pipeline, with many people changing affiliation. It would not surprise me if Alt-Right groups have taken the most autistic libertarians, since an ideology most predicated on the importance of racial differences is sure to attract people who are low in empathy and awareness of social norms and also fascinated by racial statistics. Economist Tyler Cowen, who has followed the Libertarian movement since running the Austrian Economics Newsletter at 18, says the movement is now dead. He notes that the outside of the establishment think tanks, the libertarians at home have typically moved into ‘unsavoury direction’. He also says smart autistic types on the internet have moved away from big-L Libertarianism and too holding a more eclectic broader view, including ideas further to the right.
So perhaps most of the non-establishment, autistic Libertarians are just no longer Libertarians?
This brings me onto the most obvious pattern in the above charts. Those with more extreme ideologies appear to have higher rates of autism. This is true on the left and right with both Marxists, and the Alt-righters and Neoreactionaries being around 40% autist on the self-diagnosis measure. Given autism has been described as ‘the extreme male brain’, I reproduce the above charts at the end of the blog but only with male survey respondents. All the patterns still hold.
Given autistics like hard logic and care little for social desirability bias, it makes sense that they could be attracted to extreme ideologies which propose simple answers with clear logical frameworks (eg. human biodiversity or dialectical materialism.) In Britain, people who are so extreme as to be convicted for being potential terrorists, typically all have medically diagnosed autism and have come from backgrounds such as police officer and Cambridge Maths PhD - another infamous mathematician comes to mind. An individual who ran a campaign to free the Cambridge Mathematician is a bus driver, who spends his spare time running a blog about buses.
DISCLAIMER - I’m not saying all autistics are terrorists or that Alt-Righters, Neoreactionaries or Marxists are terrorists. But the correlation between autism and fringe ideologies and the prevalence of autism amongst people convicted for being potential terrorists, might suggest autistics could have a tendency for being extreme.
We can test this hypothesis by looking at rates of autism across a 1 (Far left) to 10 (Far Right) political spectrum in the SSC survey.
On the self-diagnosed autism measure it would seem people on the more extreme ends of the political spectrum really do have higher rates of autism. However, when we look at only those with medically diagnosed autism we find autism is only associated with being extreme left wing.
The medical diagnosis pattern corresponds to the findings that left wing people tend to have higher rates of mental illness. However, it has been suggested that right wing people who believe in strength rather than victimhood and weakness, are less likely to get diagnosed as having a mental illness or may even underreport their unhappiness. My guess is that since autism is associated with being smarter and more logical, right wingers are happier to publicly identify as autistic, even if they don’t want to go a doctor to get verified. If you go on right wing forums you will find the label of autism frequently being used for banter, but also as a compliment. Given my experience with people on the fringe right and my high scores on the Autism Quotient (test yourself here), I suspect autism is associated with being generally radical, not just left wing.
That’s the end of the blog post. Below I’ve copied AQ distributions from autistics and the general population so you can interpret your own score. The source Is Simon Baron-Cohen’s book Autism: The Facts. After that I present the graphs shown above, but only using male survey respondents to show gender confounds are not distorting the results.