Close to Home: The Effect of Proximity to Violent Protests on Hong Kong's 2019 Electoral Outcomes

Abstract: This paper aims to investigate the effect of violent pro-democracy protests on electoral support for pro-democracy candidates in the case of Hong Kong's 2019 District Council election. More specifically, we estimate the size of an effect resting on three theoretically and contextually informed microfoundations: violence in protests, geographical proximity of protests to voters' homes, and temporal proximity of protests to election day. We argue that these three microfoundations are all incorporated in our designed proxy measure of MTR (metro) closures, allowing us to test their aggregate effect on constituency electoral outcomes. The analysis is pursued through an ordinary least squares method of multiple linear regression as well as a probit regression. Results show that the effect is statistically significant and robust across several model specifications. We conclude that geographical proximity to violent pro-democracy protests, within a specified time frame before election day, indeed has a positive marginal effect on support for pro-democracy candidates in the case of Hong Kong 2019. However, this effect should be viewed in relation to what is, in all likelihood, a dramatic shift in baseline support for pro-democracy candidates.