Artificial intelligence and informal employment: evidence from China’s agricultural transfer population

Artificial intelligence (AI) is profoundly reshaping labour markets, but its relationship with informal employment remains unclear. This study focuses on the agricultural transfer population (ATP), a major group engaged in informal employment in China. Using data from the China Labour-force Dynamics Survey (CLDS), we systematically examine the association between AI and informal employment among the ATP and explore potential channels. We find that AI is significantly and positively associated with informal employment among the ATP. The observed association is primarily consistent with the job-creation and empowerment channels, while evidence for the substitution channel is insufficient. Heterogeneity analysis further shows that the positive association is more pronounced among female and younger ATP workers and those employed in the service sector. After distinguishing between two types of informal employment, we find that AI is positively associated with informal self-employment, whereas its association with informal wage employment is not statistically significant. In addition, the positive association is weaker among the ATP who have received skill training, particularly those with lower levels of education. This study not only provides new empirical evidence on the relationship between AI and employment structure, but also offers important policy implications for promoting the urban integration of the ATP.