Beschreibung:
Why has geographical political polarisation increased in recent times? In a simple model, we show that one catalyst is a social learning mechanism whereby policy preferences become more homogeneous within geographical units, yet increasingly heterogeneous between units over time as voters become better informed on the views of those in their vicinity. We study our model’s predictions by taking advantage of the delayed implementation of Brexit and its salience in the UK general elections following the 2016 referendum. We ask whether voters updated their Brexit views after observing the referendum’s local results, and acted upon their new beliefs in the subsequent elections. To estimate causal electoral effects, we exploit the binary media portrayal of local areas as "Leave" ("Remain") depending on whether the local Leave-vote share recorded in the referendum was above (below) fifty percent. Analysing constituency-level longitudinal-data, we document a two percentage-points decrease in the anti-Brexit Liberal Democrat vote share in "Leave-labelled" relative to "Remain-labelled" constituencies, mirrored by an increase for the Conservatives. Our findings constitute novel evidence for contextual information signals causally contributing to geographical polarisation, and have implications for how group-based identities form more broadly