Algorithm Advancement and Workforce Dynamics: Evidence from On-Demand Delivery Market

Status: Under Review
Coauthors: Yiyu Huang (Fudan), Xianghua Lu (Fudan)

Abstract

This paper examines how algorithm advancements impact worker productivity and reshape labor market dynamics in gig economy. Using comprehensive rider-level data (2019-2023) from a major on-demand delivery platform, we develop and estimate a structural model of labor participation decisions, accounting for worker heterogeneity. We find that algorithm advancements have a significant skill-equalizing effect in gig platform, disproportionately benefiting less-skilled workers through two mechanisms: diminished returns to specialized skills and reduced barriers to market participation. This leads to substantial shifts in workforce composition with increasing participation and retention among less-skilled workers while reducing engagement from high-skilled workers. Our findings demonstrate that algorithmic impacts on labor markets depend critically on worker heterogeneity, competitive dynamics, and strategic labor supply responses. These insights have important implications for platform design, workforce development policies, and understanding how technologies reshape employment patterns in flexible labor markets.