Robots, China and Polls: Structural Shocks and Political Participation in the US

What is the effect of labour market adjustment to automation on political participation? We study the consequences of the introduction of industrial robots across US commuting zones on voter turnout in US counties between 2000 and 2016. We first replicate prior results showing negative effects of exposure to robots on employment and household incomes at local labour markets and then show that an increase in the exposure by one robot per thousand workers leads to a 0.64 percentage point lower voter turnout at US presidential elections. We contrast this result with the effect of the exposure to Chinese imports, for which we do not find a negative effect on political participation. Using individual level data we document that people at risk of automation are 15% percent more likely to abstain. To understand why the effect is not uniform, we conduct an online survey experiment. We find that the nature of the shock matters beyond the mere economic consequences. While the government is seen as instrumental in addressing the trade shock, it is perceived less effective in the case of automation. Our finding highlight an important behavioral aspect of the political economy of technological change.