Anti-intellectualism and the COVID-19 pandemic

What leads people to reject public health advice like mask wearing and social distancing? Using surveys of over 27,000 Canadians and a blend of cross-sectional, panel and experimental analyses, we highlight the importance of anti-intellectualism – a generalized mistrust of experts and intellectuals

Like Comment
Read the paper

Note this blog post relates to: Merkley, E., & Loewen, P. J. (2021). Anti-intellectualism and the mass public's response to the COVID-19 pandemic. Nature Human Behaviour.

As the pandemic accelerated in March 2020, my postdoc supervisor and co-author, Peter Loewen, approached me about running weekly surveys of adult Canadians for the Media Ecosystem Observatory. We wanted to track the attitudes and behaviours of Canadians over the course of the COVID-19 pandemic to see how they adapted as one of the most disruptive periods in modern history unfolded.

I had just finished publishing my dissertation, part of which was related to the concept of anti-intellectualism – the generalized mistrust of experts and intellectuals. I became interested in this topic as I was completing my dissertation at the University of British Columbia, which was broadly focused on the question of why citizens often fail to listen to consensus expert advice. I needed an angle for a third paper for the dissertation and I wasn’t happy with what I had at that point. I came across Richard Hofstadter’s Anti-intellectualism in American Life in Powell’s Books in Portland, Oregon, and I was off to the races.

My dissertation work found its way to Public Opinion Quarterly. In that article I show that anti-intellectualism is related to opposition to positions of scientific consensus (e.g. climate change, safety of nuclear power and GMOs, etc.) and that providing people with signals of scientific consensus on these questions is unpersuasive to anti-intellectuals and may even produce a backfire effect. I had a hunch that anti-intellectualism might be of importance in shaping how Canadians responded to the COVID-19 pandemic.

So, we put a question battery into the Media Ecosystem surveys to track anti-intellectualism, along with an assortment of COVID-19 behaviours and attitudes, like risk perceptions, social distancing, COVID-19 misperceptions, and news consumption. This would allow us to observe the relationships between these concepts over the pandemic. And, as we show in our Nature Human Behaviour article, our measure of anti-intellectualism is strongly linked to these attitudes and behaviours above and beyond the effects of left-right ideology and science literacy.

However, we weren’t happy settling for these cross-sectional, correlational analyses. There are a few important limitations. First, it is possible that trust in expert communities is partially a function of people’s beliefs related to COVID-19. Second, our behavioural measures, like social distancing compliance, are self-reports, which may not accurately reflect behaviour due to problems of recall or social desirability.

Finally, and most importantly, these analyses didn’t allow us to directly observe how anti-intellectualism shaped citizen behaviour in response to expert advice. A gold standard approach would be to randomly assign subjects communications from experts, and perhaps to prime anti-intellectualism through the random assignment of elite rhetoric as well. However, we believed that with the unprecedented salience of the pandemic and the subsequent rapid ascent of expert-led communication across the media landscape, it meant that the vast majority of potential respondents were pre-treated with information from experts.

In our work we were able to exploit a panel component of our surveys to provide more evidence of causal direction. We show that past levels of anti-intellectualism are associated with future levels of COVID-19 risk perceptions, social distancing, misperceptions, and news consumption after accounting for the past values of these outcomes and other confounders.

We also include the results of a pair of modified conjoint experiments, with pre-registered replications, where we randomly assigned survey participants – not messages from experts – but headlines, embedded in news story profiles, that either featured COVID-19 content (or not) or COVID-19 news content with or without an expert cue. We show that people prefer to read stories about COVID-19 and prefer news reports about COVID-19 that feature experts, but that these tendencies are much less true for anti-intellectuals. This design has the advantage of both avoiding problems of pre-treatment and providing behavioural evidence of the role of anti-intellectualism in COVID-19 information search.

And, most importantly, we take advantage of changing expert recommendations regarding the use of medical and non-medical masks in April and May of 2020. We see that anti-intellectualism – rather than ideology or science literacy – is associated with within-respondent change in mask usage, where uptake occurred principally among those highly trusting of experts. So it really does seem like anti-intellectualism is fundamentally connected to dynamics in how citizens respond to expert advice.

Taken together, our results highlight the importance of trust in fostering public concern related to the pandemic and compliance with public health directives. The effects we observe stand head and shoulders above factors like ideology or science literacy in Canada. We should expect this to be the case in a context where experts are at the forefront of political communications.

The implication is that improving compliance with public health directives requires more than just top-down messaging from expert sources. A sub-set of the public is deeply skeptical of these sources, so appeals to their authority are unlikely to be of much use in persuading these citizens. Top-down expert communication needs to be paired with a bottom-up community driven approach that mobilizes other trusted messengers to reinforce critical public health recommendations from expert communities, like the importance of social distancing, mask wearing, and vaccination.

Ultimately, we need a lot more research on the concept of anti-intellectualism. What are its sources and consequences and how does this vary cross-nationally? What role does elite rhetoric play in fostering or activating anti-intellectual sentiment? Are there ways we can built trust in expert communities, or persuasive strategies to reach individuals that distrust experts and expertise? As the COVID-19 pandemic has shown us, the stakes are high.  

"COVID-19 Anti-Lockdown Protest in Vancouver, May 3rd 2020" by GoToVan is licensed under CC BY 2.0

Eric Merkley

Postdoctoral Fellow, University of Toronto