Digital transformation in China's listed manufacturers expands, not contracts, labor demand—raising demand especially for highly educated, high-skilled workers. Productivity gains and improved worker digital literacy help explain the rise, though effects differ by ownership, factor intensity and firm size.

Main Finding

Digital transformation in Chinese A‑share listed manufacturing firms (2011–2024) significantly increases firms’ labor demand overall and raises demand specifically for highly educated, high‑skilled workers. The effect operates through higher total factor productivity (TFP) and improved employee digital literacy, and it varies by firm ownership, factor intensity, asset size, and type of digital technology used.

Key Points

• Digital adoption is associated with a net increase in the quantity of labor demanded (contrary to a simple automation‑causes‑job‑loss narrative).
• Demand shifts toward more educated, high‑skill workers—digital transformation complements skilled labor.
• Two mechanisms identify how this happens:
  • Productivity channel: digitalization raises firm TFP, expanding output and labor needs.
  • Skills channel: employees’ digital literacy improves, enabling greater intensity of factor input and higher skilled task performance.
• Heterogeneity:
  • Effects differ by ownership (state vs private), factor intensity (labor‑ vs capital‑intensive firms), and asset size (small vs large).
  • Different types of digital technologies produce substantially different labor‑demand effects.
• Policy relevance: findings point to the importance of workforce adaptation and targeted industrial policies during the digital transition.

Data & Methods

• Sample: Chinese A‑share listed manufacturing firms, years 2011–2024.
• Empirical approach: regression analysis to estimate the impact of firm‑level digital transformation on labor demand outcomes and to test mediation via TFP and employee digital literacy.
• Outcomes analyzed: overall labor quantity demanded and composition by education/skill level (notably high‑skill/high‑education workers).
• Mechanism tests: examined whether changes in TFP and measures of employee digital skills mediate the relationship between digitalization and labor demand.
• Heterogeneity analysis: stratified regressions or interaction terms to assess differences by ownership, factor intensity, asset size, and technology type.

Implications for AI Economics

• Complementarity over substitution: In manufacturing, digital technologies (including AI/digital tools) can complement labor and increase employment demand—especially for skilled workers—so predictions of mass job loss from automation may overstate short/medium‑run displacement.
• Skill bias and inequality risk: Rising demand for highly educated/high‑skill workers implies upskilling and education are central to capturing benefits; without policy, wage and employment inequality may widen.
• Policy targeting: One‑size‑fits‑all labor policies will be inefficient. Policies should:
  • Prioritize digital literacy and reskilling programs targeted to the technologies in use and the affected firm types (SMEs, ownership forms, sectoral factor intensity).
  • Support transitions where technologies are more labor‑substituting or where firms lack complementary capabilities.
  • Encourage adoption paths that reinforce worker complementarities (e.g., human‑in‑the‑loop AI).
• Research priorities: Future AI economics work should disaggregate by specific technologies, follow longer horizons (dynamic effects), include non‑listed and smaller firms, and pursue causal identification to better inform policy design.