Economic shocks occur regularly across European democracies (Eurofound 2017). Despite the prevalence and electoral relevance of such shocks (Colantone & Stanig 2018), the overall results on their impact on voting behavior have been inconclusive (Margalit 2019). I posit that attribution of responsibility could explain the mixed findings. I argue that attribution of political responsibility for sociotropic shocks (e.g. plant closure) is contingent on contextual factors (the type of restructuring event; the affected industry; the number of impacted workers; the sex of workers; the institutional set-up) and individual predispositions (personal job insecurity, nationalism, partisanship). I focus on Central and Eastern Europe, a fertile ground for offshoring (Eurofound 2017) and automation (McKinsey 2018).
The key test involves a conjoint experiment (Hainmueller et al. 2014) conducted on a Romanian sample recruited through Facebook (Boas et al. 2018). I expose voters to hypothetical scenarios triggered by negative restructuring events (e.g. offshoring, automation) and ask them to evaluate the effect of the scenarios on political attribution and voting behavior. I also analyze positive shocks (e.g. job creation due to technological progress). To validate my conjoint analysis, I employ a vignette experiment both in the Facebook survey and a nationally representative sample. Three additional analyses with observational data will allow me to complement my main study: first, I will match municipal-level electoral results and restructuring events in Romania; second, I will match NUTS-2 level election results with different types of restructuring events that took place in the 28 EU countries; third, again at the EU level, I will employ different waves of the European Social Survey and link respondents to the types of restructuring events that have been experienced by their region of residence and/or sector of employment between elections.
With this paper, I propose a theory of attribution of responsibility for economic shocks that explains the mixed results found in the literature. Moreover, I test whether, for economic shocks, voters are more apt to punish than to reward.