Is College Education Less Necessary with AI? Evidence from Firm-Level Labor Structure Changes

Abstract

ABSTRACT As a general-purpose technology, artificial intelligence (AI) is expected to transform almost all industries and aspects of our society. Thus, it is important to understand the potential changes within the firms related to how AI applications change their labor force. Using a panel dataset with over 1,300 publicly-traded companies in China from 2007 to 2018, we examine the relationship between AI applications and firm labor structure with workers with or without formal college education. The study indicates that AI applications were positively associated with the overall employment as well as the employment of nonacademically- trained workers with no college degrees at the firm level. These associations were more significant in the service sector than in the manufacturing sector. Further causal analysis shows increasing AI applications have a positive effect on a firm’s employment of nonacademically-trained workers and its overall employment but a negative effect on academically-trained workers. We attribute the findings to the technology deskilling effect of AI. The findings suggest that, in response to the potential labor force transformation with increasing AI applications, information-systems research needs to focus on structural changes of labor forces and the implications for preparing human employees to work with AI side by side.