Elections and domestic violence
研究2020年美国总统选举后，特朗普支持率高的选区家庭暴力报告增加，支持率每高20个百分点，暴力增加8%，且效应在重新计票完成后才显现。
Does post-election anger contribute to domestic violence? We study the 2020 US Presidential contest between Trump and Biden. Mapping block-level address data for domestic violence reports from as many of the 100 largest US cities as possible (20) to electoral precincts, we estimate diff-in-diff regressions to determine whether the pre-election vs. post-election difference in a precinct’s level of domestic violence is related to the precinct’s level of support for the losing candidate. We find that a 20-percentage point increase in the Trump two-party vote share correlates with an 8% increase in post-election domestic violence. Moreover, the effect is delayed until the recount was completed, suggesting that the effect observed is due to the resolution of the Presidential election rather than the act of voting, cessation of ads, or results of concurrent electoral contests.