Assessing the Effect of Task Automation in Labor Markets: Case of IT Services Industry

Abstract

Recent studies indicate that automation, enabled by Artificial Intelligence and associated technologies, has a profound impact on society — the prominent one being the potential displacement of workers resulting in possible unemployment and a decrease in wages. In this work, we assess the impact that task automation has on employment and wages in the Information Technology (IT) services industry. We use a triad of methodologies, including a detailed survey of experts in the industry to get their views on automation and its potential impacts, followed by agent-based analytical modeling and empirical verification using datasets from O*NET available from the U.S. Department of Labor. The key contribution of this work is an occupation mobility pathway structure using the gathered empirical data connecting the occupations in the IT services industry based on skill proximity. The paper also proposes a taxonomy of IT services tasks based on their automatability. The simulation results using the constructed agent-based model indicate that unemployment rates are lower when the displaced workers retrain in tasks of occupations with skill proximity, and the benefits are significant for workers in high-risk occupations. Accordingly, we propose policy prescriptions on automation, work policies, and labor welfare.