A Cross-national Comparison of Eight Populist Attitudes Scales

Political events of recent years, culminating with Brexit and Donald Trump’s election in 2016, have brought much attention to studying the charm of populism. Many have tried to explain its appeal at the individual level by conceptualizing populism as an attitude, and using batteries of survey questions to measure it. In order to provide valid substantive answers, however, it is essential to check whether and which of these scales actually work in capturing this attitude. This paper is the first attempt at comparing various currently published or used populism scales in comparative perspective. We test the psychometric properties and the cross-national validity of eight batteries of populism items: Castanho Silva et al. (forthcoming), Hawkins, Riding, and Mudde (2012), Akkerman, Mudde, and Zaslove (2014), Stanley (2012), Elchardus and Spruyt (2014), Schulz et al. (forthcoming), Oliver and Wood (2016), and the module at the wave 5 questionnaire of the Comparative Study of Electoral Systems (2016). For that, we use data collected online with respondents from ten countries in Europe and the Americas, with ca. 250 participants per country. While most batteries do predict support for parties considered populist, and are associated with theorized correlates of populist attitudes (such as low political trust), our results show most of them have non-invariant items, meaning they should not be used in cross-country comparisons and models. Our results also indicate that some widely used batteries might overestimate levels of populist attitudes, and that they fail to capture the full breadth of the concept. We conclude with suggestions of scales and items that produce more reliable and fine-grained results.