People who are most fearful of genetically modified foods think they know the most about them but actually know the least

There are few subjects where a larger gap exists between public opinion and expert opinion than people's views on foods, like corn or wheat, that have been genetically manipulated to, for example, increase crop yields or bolster pest resistance. Experts generally view so-called GM foods as totally safe to consume, while the public is suspicious of them - and this divide is massive.

One Pew Research Center survey found that just 37 per cent of the American public believed GM foods are safe to eat, compared with 88 per cent of members of the American Association for the Advancement of Science (public attitudes are similarly negative in the UK, with a 2014 poll finding that 40 per cent of adults felt the government should not promote GM foods, compared with 22 per cent in favour, and the rest unsure).

Unlike some subjects where this divide between layperson and expert opinion is heavily mediated by politics, such as climate change caused by human activity - in the US and elsewhere, conservatives are far less likely to believe in it than are liberals and climate scientists - the GM-food divide doesn't really have a political dimension: Liberals, centrists, and conservatives are all about equally likely to have what are, from the point of view of experts, unfounded fears about the safety of GM foods.

To better understand the source of these fears, a team led by Philip M. Fernbach, a professor at the Leeds School of Business at the University of Colorado Boulder, surveyed nationally representative samples in America, Germany and France, and other online participants, about their views on both GM foods and climate change, tested their knowledge on these subjects by asking them to answer factual questions, and also asked them to gauge their perceived level of knowledge on those subjects.

The headline finding from the study, published as a letter in Nature Human Behaviour, is neatly summed up by its title: "Extreme opponents of genetically modified foods know the least but think they know the most." That is, on average, the more vehemently a given respondent said they were opposed to GM foods, the fewer questions about the subject they answered accurately, and the higher they rated their own knowledge.

This can be seen as a subject-specific version of the Dunning-Kruger effect, or people's tendency to be ignorant about their own ignorance. The finding held across the samples in different countries (though in France, the zoomed-out finding that supporters of GM foods knew more about the subject than opponents wasn't statistically significant, as it was elsewhere):

via Fernbach et al, 2019

These are dispiriting findings. The people who know the least think they know the most, and suffice it to say some of them are likely taking an active part in spreading overblown fears about GM foods, worsening the problem.

One possible explanation for the results is that people with the strongest opposition to GM foods are concerned about matters unrelated to scientific knowledge, such as being opposed for moral reasons.

However, the researchers looked into this and found that "73% of respondents cited food safety/health concerns as their reason for opposition…Extreme opponents were actually more likely to cite food safety/health concerns than moderates and the main results replicate when we restrict the analysis to the subset of participants citing food safety/health concerns."

Things get even more interesting upon examining Fernbach et al's findings with regard to anthropogenic climate change, knowledge, and perceived knowledge. Because there, while the researchers found the same general pattern, it wasn't statistically significant: "Contrary to GM foods and gene therapy we found no significant effects of knowledge—objective or self-assessed—on extremity of climate change attitudes."

Why might that be? The researchers think it's because climate change has become politicised whereas GM foods have not. "The null effect [for climate change] has been attributed to 'cultural cognition,'" the authors write, referring to a concept most commonly associated with Dan Kahan of Yale. "When an issue becomes polarized, people's attitudes reflect affiliation with their ideological group and not individual knowledge. That is, individuals subscribe to whatever their in-group beliefs, regardless of how much they know about the issue."

Moreover, on this and other politicised topics, extreme ideology might sap people's ability to truly integrate new information, and recognise the merit of counter-arguments, in a manner that nudges their beliefs.

While the "traditional view in the public understanding of scientific literature is that public attitudes that run counter to the scientific consensus reflect a knowledge deficit," in reality this view, itself, appears to be inaccurate, at least when it comes to politically charged issues: As people get more knowledgeable, their beliefs on such controversies don't necessarily change in the predictable manner one would expect if there were a direct tie between knowledge and belief unmediated by ideology.

That might be because strong ideological beliefs cause people to simply filter and interpret new information in a manner which reinforces those beliefs. "Okay, based on this research it does look like the planet is warming, but that's part of a natural cycle that has nothing to do with humans."

As Fernbach and his colleagues note at the end of their paper, even for less-politicised issues like GM foods, their findings suggest that improving public awareness of basic scientific consequences might be more complex than previously realised, since those holding onto the most severe forms of misinformation are also least likely to seek out more facts or be open to hearing the other side. "This suggests that a prerequisite to changing people's views through education may be getting them to first appreciate the gaps in their knowledge." Which is a whole other task — and a very important one.

- Jesse Singal (@JesseSingal) is a contributing writer at BPS Research Digest and New York Magazine, and he publishes his own newsletter featuring behavioural science talk. He is also working on a book about why shoddy behavioral-science claims sometimes go viral for Farrar, Straus and Giroux.