Chinese manufacturers that pair digital upgrades with green transformation pay workers closer to their marginal product — firms with higher digital–green synergy exhibit significantly lower labor-market monopsony, driven by stronger labor demand, scale expansion, and investment, with the digital dimension having the larger effect.

Main Finding

The paper finds that a firm-level digital–green synergistic transformation (coordinated advancement of digitalization and greening) significantly reduces firms’ labor-market monopsony power in Chinese manufacturing. This improvement in the distributional relationship between labor and capital operates mainly through (1) expanding labor demand, (2) expanding production scale, and (3) raising investment. Effects are stronger for state-owned enterprises and larger firms; both the digital and green subsystems individually restrain monopsony power (digital having the stronger single effect). The paper finds no empirical support for a moderating role of corporate governance.

Key Points

• Question addressed: Does coordinated digitalization and greening (the digital–green synergistic transformation) make factor income distribution more equitable by weakening firms’ monopsony power over labor?
• Conceptualization of distributional outcome: labor market power (a firm’s ability to pay wages below workers’ marginal product) — i.e., the degree worker remuneration departs from marginal contribution.
• Main result: Higher levels of digital–green synergy → lower firm-level labor-market monopsony (improved share of surplus to labor).
• Mechanisms identified:
  • Labor demand channel: synergy increases overall and high-skill labor demand (new businesses, productivity, green upgrading), raising the elasticity of labor supply facing firms.
  • Production-scale channel: synergy enables product-market expansion and geographic/segment diversification, diluting local hiring monopsony.
  • Investment channel: higher digital+green investment expands capacity and complementary demand for skilled labor (capital deepening with derived labor demand).
• Heterogeneity:
  • Stronger distributional improvement among state-owned enterprises (SOEs).
  • More robust estimates for larger firms.
  • Both digital and green subsystems matter separately; digital alone has a larger effect than green alone.
• No support found for corporate governance strengthening the effect.
• The authors also examine product-market markups as a complement to ensure the effect is originating in the labor market rather than via product-market power.

Data & Methods

• Sample: Chinese A-share listed manufacturing firms, 2007–2023; unbalanced panel with 836 firms and 11,815 firm-year observations.
• Data sources:
  • Financial and production data: CSMAR.
  • Patent data: CSMAR research patent database.
  • Environmental disclosures: CSR reports and annual reports.
• Sample processing: exclude abnormal listings, non-positive inputs/outputs, firms with <8 employees, and observations missing key variables; continuous variables winsorized at 1st/99th percentiles.
• Key empirical measures:
  • Digital–green synergistic transformation index: constructed from firm-level digital indicators and environmental indicators using an entropy-weight method and a coupling coordination model (to capture coordination/synergy, not only separate levels).
  • Firm labor-market power (Markdown): measured via a production frontier / cost-minimization approach that extends De Loecker & Warzynski (2012) and follows Yeh et al. (2022), identifying the extent wages fall short of labor’s marginal product under imperfect labor-market competition.
• Identification strategy: firm-level panel regressions linking the synergistic index to measured labor-market monopsony, with tests of transmission channels (employment, scale, investment) and heterogeneity analyses. The paper also inspects product-market markup as a complementary outcome.
• Limitations noted by authors: sample restricted to listed manufacturing firms that disclose both digital and environmental information (so representativeness is limited); sample-selection and data-availability constraints.

Implications for AI Economics

• Measurement matter: This paper emphasizes measuring distributional impacts via firm-level labor-market power (markdowns) rather than raw employment or wage levels — a useful approach when studying AI/digital adoption and inequality because monopsony captures appropriation of worker surplus.
• Complementarity of digital and green investments: AI/digital adoption should be analyzed jointly with complementary investments (e.g., green transition, capital deepening). Synergies can produce distributional gains if they expand labor demand and scale rather than only substituting labor.
• Channels to watch in AI research:
  • Labor-demand composition: AI may displace routine tasks while increasing demand for high-skilled roles; the net effect on monopsony depends on whether derived demand and hiring scale outpace displacement.
  • Scale and diversification effects: digital-enabled market expansion or geographic reach can dilute local monopsony power.
  • Investment-induced derived demand: AI/automation investments that are complementary to high-skill labor can increase labor’s marginal product and bargaining leverage.
• Policy relevance: Results suggest policy should encourage coordinated digital and green investments and support channels (green finance, upskilling, competition in hiring markets) to ensure AI-driven productivity gains are shared with workers.
• Research takeaway: Empirical studies of AI’s distributional effects should account for imperfect labor-market competition (monopsony) and explore heterogeneous effects by firm size, ownership, and sectoral complementarities (e.g., green transition).

If you want, I can: - Extract the paper’s empirical specification(s) and variable constructions in more detail (e.g., the exact form of the coupling coordination metric or the markdown estimation). - Draft policy recommendations tailored to AI and green-tech adoption based on these findings.