New paper: “Your Health vs. My Liberty”

How might messaging, reasoning, and philosophical beliefs predict people’s responses to pandemics? Michał Białek and I started wondering about this a few months ago. So we ran some experiments to find out. Our pre-registered hypothesis was wrong, but the other findings were really interesting. Before I get to the findings, consider making some predictions: ask yourself how you expect the following variables to correlate with compliance (or non-compliance) with public health officials’ recommendations such as mask-wearing and sheltering in place:

  • Flatten the curve graphs: Related to compliance? More or less compliance?
  • Reflective reasoning: Related to compliance? More or less compliance?
  • Mathematical competence: Related to compliance? More or less compliance?
  • Economic conservatism: Related to compliance? More or less compliance?
  • Social conservatism: Related to compliance? More or less compliance?
  • Libertarianism: Related to compliance? More or less compliance?
  • Effective altruism: Related to compliance? More or less compliance?
  • Utilitarian sacrificial harm: Related to compliance? More or less compliance?
  • Belief in God: Related to compliance? More or less compliance?
  • Religiosity: Related to compliance? More or less compliance?
  • Compatibism about free will: Related to compliance? More or less compliance?
  • Identifying as White: Related to compliance? More or less compliance?
  • Identifying as a man: Related to compliance? More or less compliance?

1. Identifiable Victim Effect

Have you heard of the identifiable victim effect? For the uninitiated, it’s the finding that people are more motivated to victims of crises when they are portrayed as individuals compared to when the victims are portrayed statistically. As far as I know, no one has tested the identifiable victim effect during a global crisis. 

So Michał Białek and I decided to run the experiment. We expected to find an identifiable victim effect: public health messaging featuring individual victims would motivate more mask-wearing, sheltering in place, etc. than public health messaging about statistical victims—i.e., “flatten the curve” graphs.

We were wrong.

Contrary to the identifiable victim effect: Public health compliance and perceived threat of pandemics was mostly unaffected by the presence (or absence) of identifiable victims.

2. Reasoning

Intuitively, you might think that some people will be more swayed by flatten the curve graphs. In particular, you might think that people who reason more carefully—especially about math—will be more swayed by data visualizations than by pictures of individuals. To test this, we gave participants in all experimental conditions reflective reasoning and numeracy tests.

Counterintuitively, neither reflection nor numeracy test performance reliably predicted compliance or threat when controlling for messaging and other factors such as philosophical belief. 🤷‍♂️

3. Philosophical belief

When we control for messaging, reasoning, and a bunch of philosophical preferences, things get interesting. We did not detect links between compliance with public health recommendations and typical political preferences, utilitarian acceptance of sacrificial harm, belief in God, beliefs about free will, or metaphilosophical beliefs.

  • The best predictor of compliance? Effective altruist preferences
  • The best predictor of non-compliance? Libertarian preferences

Was that what you predicted before you started reading? Tell us on Twitter!

So what else predicted compliance with public health recommendations?

  • Beliefs about science
  • Religiosity (but not belief in God)

4. Conclusion

Overall, the data seem to suggest that non-compliance with public health recommendations may not be a matter of ineffective messaging, reasoning, so much as certain beliefs about morality and science.

Of course, there are lots of other interesting descriptive stats in Table 3 that might be related to other research. E.g., contrary to some prior research, neither gender nor ethnicity correlated with public health compliance. Which correlations are you most interested in? Tell us on Facebook.

This research involved two experiments on a total of 998 English-fluent people from April 15-20 and May 7-11. You can find a free copy of the pre-review paper on PsyArxiv at http://psyarxiv.com/5tjaf or—as with all of my papers—on my CV at byrdnick.com/cv.

Related

Published by

Nick Byrd

Nick is a cognitive scientist at Florida State University studying reasoning, wellbeing, and willpower. Check out his blog at byrdnick.com/blog