Metadehumanization, the perception that another group dehumanizes your own group, is a robust predictor of Americans’ support for anti-democratic norms

Landry, Alexander. 2021. “Metadehumanization Erodes Democratic Norms During the 2020 Presidential Election.” PsyArXiv. April 26. doi:10.31234/osf.io/yj4h6

Abstract: The present research identifies social psychological factors threatening American democracy. Namely, we identify metadehumanization, the perception that another group dehumanizes your own group, as a robust predictor of Americans’ support for anti-democratic norms. Both immediately before and after the 2020 US Presidential Election, American political partisans perceived that their political opponents dehumanized them more than was actually the case. Partisans’ exaggerated metadehumanization inspired reciprocal dehumanization of the other side, which in turn predicted their support for using anti-democratic means to hurt the opposing party. Along with extending past work demonstrating metadehumanization’s corrosive effect on democratic integrity, the present research also contributes novel insights into our understanding of this process. We found the most politically engaged partisans held the most exaggerated, and therefore most inaccurate, levels of metadehumanization. Moreover, despite the socially progressive and egalitarian outlook traditionally associated with liberalism, we found that the most liberal Democrats actually expressed greatest dehumanization than Republicans. This suggests that political ideology can at times be as much an expression of social identity as a reflection of deliberative policy considerations, and demonstrates the need to develop more constructive outlets for social identity maintenance.

---

Partisan Differences in (Meta)dehumanization. Moore-Berg et al. (2020a) found roughly equal levels of both metadehumanization and dehumanization between Democrats and Republicans. Consistent with this, independent samples t-tests revealed that Democrats and Republicans did not differ in their metadehumanization (Time 1: p = .63; Time 2: p = .20), nor in their metaprejudice (Time 1: p = .20; Time 2: p = .52). However, in our sample, Democrats expressed greater dehumanization of Republicans than vice-versa (Time 1: M_diff = 11.14, t(858) = 4.86, p = .001, d = .35; Time 2: M_diff = 7.65, t(858) = 3.41, p = .002, d = .24). Democrats also expressed greater prejudice (Time 1: M_diff = 15.30, t(858) = 6.54, p = .001, d = .45; Time 2: M_diff = 13.21, t(858) = 5.61, p = .001, d = .39) and spite (Time 1: M_diff = 0.39, t(853) = 4.72, p = .001, d = .33; Time 2: M_diff = 0.31, t(853) = 3.68, p = .002, d = .26) than Republicans.