This study analyzes how exposure to information delivered by an opposition party facing a longstanding incumbent affects voters in an illiberal democracy, a context where political and media competition is low and the electorate is divided. I estimate the effect of two randomized door-to-door information campaigns on voter behavior in a constitutional referendum to lower constraints on the executive branch. I also test for a polarized electorate among over 220,000 voters reached by the campaigns. The party opposing the referendum delivered information on either i) economy and terrorism related policy outcomes or ii) implications of the institutional change. Using administrative data, I find that each campaign had a zero average effect on vote share, but increased political polarization due to heterogeneous effects. These polarizing ef- fects persist in elections fourteen months after the referendum. I provide suggestive evidence that the negative effect that continues to persist two years later is due to habit formation.

Please sign up for meetings here: docs.google.com/spreadsheets/d/1XJOSHywCIKfSQ2nIJ92bsYS15n_OJgATKDbjRxgQ-zo/edit#gid=0

Link to paper: www.cerenbaysan.com/uploads/1/7/7/3/1773537/polarization_cbaysan_oct2019.pdf