Social Networks and the Geography of Voting: A Spatial Econometric Approach

The thesis will examine the role of social networks in the spatial distribution of electoral outcomes. It will use spatial data analysis to make empirical contributions to a body of related literatures in electoral and economic geography, regional science, political economy, and sociology. The thesis will comprise of an introduction, a background chapter, three empirical chapters, and a conclusion. The background chapter will engage with literature on the effects of online and offline social networks on voting, stressing a research gap on the role of various spatially heterogeneous socioeconomic shocks in these processes. It will also cover methodological considerations in identifying the effects of spatial interactions, and discuss a novel measure of interregional social ties that will be used across empirical chapters. The third chapter will examine the role of social networks and trade-related employment shocks in support for Brexit in the UK. It will look at referendum outcomes in subnational regions, asking whether social proximity to the adverse economic conditions affected the spread of Brexit support. The fourth chapter will examine the effects of social networks on support for anti-immigration platforms during the European refugee crisis. Exploiting data on the allocation of refugees to German districts, it will explore how social proximity to refugee settlement affected voting for the country's foremost nationalist party in the 2017 general elections. The fifth chapter will look at how social networks affected voter turnout during the COVID-19 pandemic. Focusing on the 2020 municipal elections in French departments, it will examine how voter turnout varied by social proximity to COVID-19 hotspots. Overall, the thesis will contribute new evidence on the political spillovers of local socioeconomic shocks, shedding new light on the implications of spatial inequalities in the 21st century.