Experimentally induced changes in authoritarian submission as a response to threat

Abstract

Authoritarianism is best conceptualised by three attitudinal clusters: Aggression, Submission, and Conventionalism. Once considered a fixed characteristic, recent observational research has demonstrated how the dimension of submission can fluctuate in response to COVID-19 threat as a means of maintaining collective security. However, this effect has not been investigated with other forms of threat, nor has it been supported experimentally. In the present study, we sought to test observational findings by priming 300 participants with either a COVID-19 threat, a domestic terrorism threat, or a non-threatening control. Levels of authoritarianism were tested before and after presentation of a prime and then the difference between the two measures could be compared between prime conditions. Results from a Bayesian multivariate regression analysis informed by observational findings suggested that participants who experienced the COVID-19 or terrorism primes reported higher levels of authoritarian submission after the prime compared to before the prime, relative to those who experienced the neutral control prime. In contrast, the aggression subfactor did not seem to elicit any change in response to threat, and the conventionalism subfactor showed a response only to the terrorism prime. We concluded that two different forms of societal threat could elicit changes in specific dimensions of authoritarianism over a very short time span. We caution against the common practice of treating authoritarianism as a unidimensional construct without careful consideration.

Figure 1

Density plots of the two threat measures by each prime condition. COVID-19 threat demonstrated a ceiling effect, whereas terrorism threat yielded a spread of values across its entire range.

Figure 2

Likelihood, prior, and posterior distributions of posterior samples for each prime condition relative to the control prime. The black dot marks the mean, and tails within each shape denote the 95% CI based on the ETI defined above. Horizontal line gives reference to zero, implying no effect of the prime with each authoritarian factor.