Make It Great Again: The Relationship Between Populist Attitudes and Nostalgia

Abstract

Populist movements typically endorse a pessimistic view that blames the "elites" for societal problems. Why is this populist worldview so appealing to many citizens? We propose that populism is associated with nostalgia: A bittersweet feeling defined as a sentimental longing for a better past. We tested this idea in three preregistered studies. Study 1 revealed that both personal and collective nostalgia (i.e., referring to either personal memories or a shared national history) were associated with populist attitudes. Moreover, the nostalgia measures mediated a link between collective angst and populist attitudes. Studies 2 and 3, then, were experiments designed to investigate the causal order between nostalgia and populist attitudes. In Study 2, a manipulation of nostalgia could not establish a causal effect on populist attitudes; however, a measure of nostalgia was again correlated to populist attitudes. In Study 3, we tested the reverse causal order by exposing participants to either a populist or pluralist speech. Results revealed that exposure to the populist speech increased both personal and collective nostalgia. In all studies, these effects emerged independent of political orientation. Apparently, feelings of nostalgia are closely associated with populist attitudes and may help explain why citizens find a populist worldview appealing.

Personality traits, democratic support, and authoritarian nostalgia

“…Studies have found a close link between nostalgic sentiment and rightist populism worldwide, raising concerns for the mobilizing effects of nostalgia in the midst of the new wave of autocratization (Elçi 2021; van Prooijen et al. 2022). While the distinction between authoritarian nostalgia and authoritarianism is difficult to draw, empirical results from this article suggest that nostalgic voters retain distinct characteristics from conservative or radical‐right voters.…”

Those were the what? Contents of nostalgia, relative deprivation and radical right support

“…To test the differential effects of group-based and personal nostalgia, I summarize them into separate indices. While the two group-based nostalgia items resemble operationalizations in previous research (e.g., Smeekes, 2015;Smeekes et al, 2014;Van Prooijen et al, 2022), a confirmatory factor analysis (CFA) using the R-package lavaan (Rosseel, 2012) Control variables. I control for seven demographic background variables: gender (male, female, other), age (continuous), highest education, occupation status, gross household income, origin (Dutch or other) and place of residence (1 [Extremely urban] -5 [Not urban]).…”

“…Hence, a growing body of literature documents populists’ nostalgic rhetoric (e.g., Elgenius & Rydgren, 2022), nostalgia's changes over time (Heath et al., 2022) or its link to acculturation preferences (Smeekes & Jetten, 2019) and radical right ideology (De Vries & Hoffmann, 2018; Smeekes et al., 2021). Experimental evidence by Van Prooijen and colleagues (2022) shows that people become more nostalgic after populist, compared to pluralist, rhetoric.…”

“…The data have two merits. First, they allow me to quantitatively map the prevalence of various nostalgia contents, which, compared to previous work differentiating personal and group‐based nostalgia (e.g., Van Prooijen et al., 2022), is more detailed and derived from a representative sample. Second, compared to previous cross‐sectional analyses on the relationship between nostalgia and radical right support (e.g., De Vries & Hoffmann, 2018; Smeekes et al., 2021), the panel design facilitates better causal conclusions.…”

“…Likewise, nostalgia and radical right voting may be consequences of omitted variables. Arguably, panel data cannot keep up with experimental designs’ causal evidence (e.g., Van Prooijen et al., 2022). Nevertheless, combining a large, representative sample with various control variables and fixed‐effect regressions with clustered standard errors, reduce the risks of omitted variable bias and selection effects better than cross‐sectional designs.…”