Contextual-level Unemployment and Support for Radical Right Parties: A Meta-Analysis

Who does not think that unemployment triggers radical right voting? Both academics and popular media associate the rising popularity of the radical right with increasing unemployment rates. Radical right parties would profit from unemployment because then in particular blaming migrants for taking away jobs from natives would resonate. And second, because in times of high or increasing unemployment voters are more likely to evaluate the incumbent government of incompetency in dealing with the economic situation. However, the numerous empirical studies resulted in an ambiguous conclusion. Whereas some studies indeed found that high unemployment is associated with high support for radical right parties, most studies did not find this relationship, or even hinted at a negative relationship. We conducted a meta-analysis to gain better insights why this apparent variation in the results exists. Our first goal was to get a better insight in the actual relation between contextual-level unemployment and support for radical right parties, based on multiple samples. Our second goal was to study whether characteristics of an included study moderated the found effect in that study, for instance the researched time and place, applied methodology, and used measurements. To accomplish these goals, we have included about 60 studies that reported an empirical effect of contextual-level unemployment on radical right support. Our results show that the overall effect of unemployment levels on support for the radical right is slightly positive, which is in line with the dominant theoretical frameworks in the literature. However, this effect appears to be random across studies, which indicates that the effect is moderated by characteristics of the included studies. We employed a multivariate meta-regression to test which characteristics were of influence. One of the outcomes is that the effect is stronger in studies that have hypothesized the effect compared to studies that only included unemployment as control variable, possibly indicating publication bias. Furthermore, the effect appeared to be stronger in studies that used data conducted after 2001, when the immigration issue increased in salience. This might imply that unemployment only fosters radical right voting if parties can successfully blame immigrants for increasing unemployment. The effect of contextual-level unemployment was also stronger in studies that used actual voting behaviour as dependent variable compared to studies that used vote intention. Alternative section: Public Opinion and Voting Behaviour in Congested and Contested Political Arena (S55)