Artificial Intelligence, Labor Market Frictions, and the Effectiveness of Monetary Policy in Stabilizing Employment

Abstract

The rapid development of artificial intelligence (AI), combined with labor market matching frictions in China, has profound implications for labor market employment and wage dynamics, while also altering monetary policy mechanisms for employment stabilization through its labor substitutability and capital-biased attributes. Existing studies, however, have not explored the interplay between AI’s employment substitution effects and wage-enhancing effects within a unified theoretical framework, nor adequately examined how AI influences the effectiveness of monetary policy in stabilizing employment.

This paper constructs a dynamic stochastic general equilibrium (DSGE) model incorporating heterogeneous AI shocks and Diamond-Mortensen-Pissarides (DMP) search-matching frictions. Through mechanistic analysis, we elucidate the micro-level channels through which AI affects employment and monetary policy regulation efficacy. Building on this foundation, we conduct numerical simulations based on China’s economic realities to analyze how general-purpose AI (GPAI) and capital-biased AI (CBAI) affect employment and monetary policy effectiveness. Our findings reveal that AI influences employment through both the substitution and income effects, but the income effect lags behind the substitution effect, resulting in short-term employment displacement and exacerbating the coexistence of “high growth, high investment, and low consumption” in the economy. The capital-biased nature of AI shifts enterprise behavior from “hiring for expansion” to “investing for expansion” under accommodative monetary policy, rendering monetary policy effective in promoting economic growth but ineffective in stabilizing employment. Additionally, intensified labor matching frictions elevate recruitment costs, further weakening enterprises’ motivation to hire under accommodative monetary policy. Incorporating employment objectives into monetary policy frameworks involves a trade-off between “employment stabilization” and “growth stabilization”: while it may partially mitigate AI-driven excessive labor substitution under accommodative monetary policy, investment and output would relatively decline.

Compared to existing literature, this paper’s innovations are primarily reflected in: First, it transcends the prevailing focus on AI’s substitution effects by formalizing both substitution and income effects within a unified framework, while theoretically unraveling the microeconomic mechanisms of AI’s impacts on employment and monetary policy effectiveness. Second, it examines the effectiveness of incorporating employment objectives into expanded monetary policy rules to address excessive employment substitution under AI development when AI creates short-term employment substitution and exacerbates divergence between output and employment gaps, thereby further clarifying feasible paths for strengthening monetary policy regulation.

This research reveals specific mechanisms through which AI affects employment, wages, and monetary policy effectiveness. On one hand, it helps guide AI development from “labor substitution” toward “human-machine collaboration,” while reducing labor market matching frictions through multiple approaches to mitigate excessive employment substitution; on the other hand, it necessitates continuous improvement of social security systems to address short-term employment substitution, while strengthening monetary policy objective management and coordinating multiple measures.