A Portrait of the Automation Susceptible Individual: Skills-Biased Technological Change and the American Conscience

Abstract

This study considers the relationship between automation potential and the views, beliefs, traits, and experiences of a randomised sample of 26,311 Americans. Previous research on automation has largely attempted to consider the current and potential impact of labour-saving innovations (e.g. digitalisation) on macroeconomic and labour market conditions. There is currently limited literature tracing the views of individuals who are susceptible to automation. This study uses data from the American National Election Survey alongside 22-category automation potential estimates presented by The Brookings Institution. Findings are presented in two parts: the first summarises the time-varying characteristics of automation susceptible individuals along 224 characteristics and views; the second applies a fixed effects OLS model to study the potential impact of elevated automation potential on a narrower set of dependent variables. Additional robustness checks are conducted as well as additional tests with an experimental AI-exposure treatment variable. The chief findings of this approach are that automation susceptible Americans are more likely to be culturally conservative, economically left-leaning, anti-immigration, racist and intolerant, politically apathetic, politically pessimistic, cynical, and despairing.

Keywords: Inequality, Fairness, Shocks, Beliefs, Redistribution, Automation

JEL Classification: D63, D72, D83, D91, H23, J68, O38