Job Exposure Rate by Generative AI, Risk of Unemployment, and Policy Preferences: Evidence from Experiments

• In Event: Public Policy and Economic Wellbeing from Employment to Retirement

Abstract

The rapid development of generative AI poses significant risks to white-collar jobs. This study investigates how information regarding job risks associated with generative AI influences subjective job security and policy preferences. We conducted a survey experiment in China, randomly presenting participants with data on the exposure rates of white-collar jobs to generative AI. Our findings reveal that treated respondents both perceive their own and other white-collar workers’ jobs as more susceptible to replacement. They show reduced support for government subsidies aimed at enhancing AI development and increased support for progressive social policies, including continuing education, enhanced unemployment benefits, and universal basic income. While treated groups are not more likely to support for education that emphasize on social skills or STEM skills, they are more likely to say that they hope their children would become bureaucrats in China so that their children would have a stable job. These results have important implications for policymakers as they navigate the intersection of technological advancement and labor market dynamics. 

Author

• Menghan Shen, Teachers College, Columbia University
  Presenting Author