Who Supports Regulating Automation?: Identifying “Luddites” Preference

• In Event: (Post-)Pandemic Politics

Abstract

What makes people support regulating labor-saving technology adoption in the economy? While recent technological progress in robotics and artificial intelligence (AI) increases economic efficiency, it also deepens public anxieties about job displacement and inequality over the last decade. While existing studies have examined whether higher 'automation risks' of individuals lead to higher support for economic redistribution or social protection, empirical findings show mixed shreds of evidence. Further, they have not explored how sociotropic rationale regarding 'what is good or bad for society' affects policy preferences, going beyond the self-interest of individuals. My research fills this gap by interrogating how public perception about the social impacts of automation defines support for restrictive (i.e., levying 'automation tax' on employers) and compensatory measures (i.e., unemployment benefits and skill training) against the negative consequences of automation.

I conduct an online survey experiment from British and Swedish populations, where the survey respondents are embedded in a different labor market setting that affects automation-driven employment anxieties. The study posits that public concerns about deepening inequality lead to higher support for restricting labor-saving technologies in the workplace and expanding compensatory measures, while people are less willing to impose regulations on automation when they see automation brings shared benefits for employees. I identify the ‘distinctiveness’ of public perception on automation-driven economic change. As the 19th century Luddite movement in Britain illustrates, empirical findings show that people are willing to restrict the change at the expense of slowing down 'progress' in automation if it deepens inequality. My study presents two policy implications. First, public policies that assuage public anxieties about job-displacing automation are effective when they are combined with policies that promote an economic safety net in the labor market, which includes unemployment benefits and skill training/education. This implies that a compensatory approach is more effective than a restrictive measure in addressing public anxieties toward automation. Secondly, deepening economic inequality can be detrimental to economic growth by undermining favorable assessment of automation-driven change, which may impede technological progress in the economy. It shows that taking advantage of new labor-saving technologies can be maximized only when their benefits are shared widely rather than resulting in capital-biased distribution, suggesting that economic efficiency and equitable sharing are compatible.

In a nutshell, I demonstrate that technology is not destiny. Instead, how people navigate their hopes and fears from automation-driven economic change is dependent on politics concerning ‘who gets what' from the double-edged transformation.

Author

• Jaewook Lee, McGill University