The spread of extremist attitudes is a critical issue in democratic societies. However, the direct measurement of extremist attitudes may be biased by social desirability. Participants may conceal their extremist attitudes in surveys to avoid stigmatization, as such attitudes contradict the democratic principles upheld by society. Previous research relies on direct survey items that do not account for social desirability bias.
To my knowledge, this study is the first to address this gap by measuring social desirability bias in right-wing, left-wing, and religious extremist attitudes, examining which groups are most affected by this bias, and testing whether factors found in previous extremism research are relevant after controlling for social desirability bias. The study includes 6,200 respondents from Germany, Great Britain, and the Netherlands and combines direct items of right-wing, left-wing, and religious extremist attitudes with list experiments.
I show that social desirability bias is highest in religious extremist attitudes (~11%-35%). In contrast, social desirability bias remains lower in left-wing extremist attitudes (~10%) and right-wing extremist attitudes (~9%). Across all forms of extremist attitudes, men and younger respondents are most prone to social desirability, while educational attainment has ambiguous effects on social desirability bias. After controlling for social desirability, many factors found in previous studies show no or inconsistent correlations with extremist attitudes (e.g., political alienation). Only authoritarian personality traits are significantly correlated with all forms of extremist attitudes.
Therefore, future research on extremist attitudes should test findings by considering the bias induced by social desirability and including indirect measurement approaches.