Exploring Regional Context Effects: Populist Party Support in Post-Socialist Czechia and Eastern Germany
Martin Refisch, Josef Bernard, Andreas Klärner
The debate on the geographies of discontent suggests a causal influence of regional deprivation, often referred to as the 'left behind' regions, on support for populist parties. Indeed, analyses using regionally aggregated data have found correlations between a region's economic, social and demographic characteristics and populist support. The strength of these associations depends on the country, the political parties and the operationalisation of left-behindness used. However, operating at an aggregate level does not allow conclusions about effects on individual preferences, as contextual effects cannot be distinguished from compositional effects.
In order to be able to estimate the relevance of contextual living conditions for populist party support, we conducted multi-level analyses for western Germany, eastern Germany and the Czech Republic. Contextual data were obtained from national statistical offices, the European ARDECO database and INKAR. We collected a wide range of economic, social and demographic indicators and reduced them using exploratory factor analysis, which provides insights into the structure of regional inequalities. The resulting dimensions are (i) economic prosperity, (ii) social exclusion and (iii) relative expansion. In addition, we build on previous studies by also testing for the effects of (iv) population density and (v) the proportion of the population aged over 65. These contextual data are merged with national survey data from the German General Social Survey 2021 and the Czech Household Panel Study 2023. We focus on binary dependent variables indicating whether the respondent would vote for the Alternative for Germany (AfD), the ANO or the SPD if an election were held in the week of the interview. In addition, a number of socio-demographic control variables are also included in the models.
The initial finding of our logit multi-level models is the identification of the location of the variance. With regard to the Alternative for Germany (AfD), only 5.8 per cent of the variance in eastern German AfD support can be attributed to contextual factors, while 94.2 per cent is explained by individual-level characteristics. The value is comparable for the SPD (5.7 per cent) and even lower for the ANO (3.4 per cent). The second main finding is derived from the regression coefficients and their significances. Support for the SPD and ANO is most likely when living in regions with low economic prosperity. However, the SPD is also more supported in rural areas, while the ANO receives the greatest support in regions which experience high social exclusion rates as well as population and job losses. The relative dynamics also increases the probability of voting for the AfD in eastern Germany, as does a high share of elderly inhabitants. This differs from western Germany, where economic prosperity is the most relevant dimension.
In conclusion, context effects were identified, but their influence is weak and differs by party and region. Therefore, a multidimensional concept of left-behindness is essential for understanding regional variations in populist voting. Additionally, the geography of discontent arguments can assist in understanding the factors that drive populist support in Czechia and eastern Germany. However, focusing on the individual level is more promising for the explanation of populist support.