Social capital, right-wing populist vote and compliance with Covid-19 restrictions: Local evidence on the restriction of mobility in Germany

[Proposal for Panel 3) Civil society mobilisation OR Panel 5) Discourses/narratives/(mis)information] In early 2020, the Covid-19 pandemic spread quickly and soon affected virtually all countries in the world. The pandemic constitutes an unpredictable external shock, which provides a promising setup for comparative research about the determinants of government policies and citizen behavior and compliance. In Germany, debates in newspapers and social media suggest that areas with higher vote shares for the right-wing populist AfD party faced higher infection and death rates during the second wave of the pandemic. The question whether there is any causal relation between these aspects remains controversial. Infection and death rates are caused by a large variety of factors that are partly beyond human control or not controllable for researchers (such as luck, proximity to high-infection areas, spread of more contagious mutations, testing rates). Therefore, we do not consider these final ‘policy outcomes’. Instead, we focus on the conscious behavior of citizens in response to governments’ policies (‘policy outputs’) against Covid-19. To measure citizens’ compliance with government-imposed restrictions on mobility, we employ fine-grained county-level data on mobility reduction, which are based on mobile phone data. We explore the relation between (a) this measure of actual citizen compliance, (b) social capital, and (c) the vote share for AfD. Based on recent advancements in the literature on social capital, we hypothesize that social capital and social networks can not only have virtuous effects but that they can also spread behavior that is normatively undesirable. Specifically, we show that the density of clubs and private associations in a county (Kreis) is negatively related with the reduction in mobility. Using the density of clubs in the 1930s as an instrument for today’s club density, we aim to show that this relation is robust and causal. The analysis attempts to disentangle this network effect from political preferences for or against the governments’ containment measures. Indeed, counties with a higher vote share for AfD complied less with the government’s mobility restrictions, and partly even increased their mobility compared to the previous year. The effect of AfD preference was much stronger in the second wave of the virus and during the second ‘lockdown’ (in Fall 2020) than during the first (in Spring 2020). This reflects a change in the AfD leadership’s strategy: While the party had demanded stronger protection measures against the virus during the first wave, it changed its position by 180 degrees in the second wave to heavily criticize all government containment measures.