Chinese listed firms that adopt AI pay their executives more, driven by productivity gains and easier financing; stronger corporate governance materially curbs this pay uplift.
As artificial intelligence is elevated to a national strategic level and deeply integrated into the real economy, its core role in driving industrial upgrading and economic growth has become increasingly prominent. Executives, as key decision-makers in corporate strategy, have compensation levels that not only relate to agency costs and incentive effectiveness but also influence the fairness and efficiency of internal income distribution within enterprises against the backdrop of technological change. Based on data from A-share listed companies from 2007 to 2023, this study employs textual analysis to construct a firm-level indicator of AI application and systematically examines the impact of AI on executive compensation. The findings are as follows: First, AI significantly increases the compensation levels of executives in listed companies. This conclusion remains robust after addressing endogeneity concerns through methods such as instrumental variable approaches and conducting a series of robustness tests. Second, mechanism tests reveal that AI indirectly promotes the growth of executive compensation through three pathways: alleviating financing constraints, enhancing enterprise total factor productivity, and increasing the level of executive human capital. Third, the level of corporate governance plays a negative moderating role between AI and executive compensation, suggesting that sound internal governance mechanisms can effectively curb the potential expansion of managerial power and the capture of excess returns by executives during technological change.
Summary
Main Finding
AI adoption at the firm level raises executive compensation in Chinese A‑share listed companies (2007–2023). The positive relationship is robust to endogeneity controls (including instrumental variable approaches) and multiple robustness checks. Mechanism tests show the effect operates indirectly via eased financing constraints, higher total factor productivity (TFP), and increased executive human capital. Strong corporate governance weakens the AI → pay link, indicating governance can limit managerial rent capture during technological change.
Key Points
- Sample and scope: A‑share listed firms, 2007–2023.
- AI measure: firm‑level AI application indicator constructed via textual analysis of corporate disclosures.
- Core result: firms with greater AI application pay their executives more.
- Endogeneity: result survives instrumental variable approaches and other robustness tests.
- Mechanisms (mediation):
- Alleviation of financing constraints — AI makes investment/expansion easier to finance, enabling higher pay linked to larger or riskier projects.
- Productivity gains — AI increases firm TFP, which is associated with higher compensation.
- Executive human capital — AI raises the skill/market value of managers, lifting their pay.
- Moderation: stronger internal corporate governance reduces the magnitude of AI’s effect on executive compensation, consistent with governance curbing power expansion and rent extraction.
Data & Methods
- Data: Panel of Chinese A‑share listed companies, 2007–2023.
- AI indicator: constructed at the firm level using textual analysis of corporate disclosures (e.g., filings/annual reports) to capture AI application intensity.
- Empirical strategy:
- Baseline panel regressions linking AI indicator to executive compensation, with standard firm controls and fixed effects.
- Endogeneity addressed via instrumental variable approaches (details as reported in the study) and a battery of robustness checks.
- Mechanism analysis via mediation tests assessing financing constraints, TFP, and executive human capital as channels.
- Moderation analysis testing interaction between AI adoption and corporate governance measures to assess governance’s dampening effect.
- Outcome: executive compensation (aggregate or top‑management pay); controls likely include firm size, performance, leverage, industry and year effects (as standard in this literature).
Implications for AI Economics
- Distributional consequences: AI adoption can raise upper‑tail earnings within firms by increasing executive pay, affecting intra‑firm income distribution and potentially overall inequality.
- Role of institutions: Corporate governance materially shapes how AI’s economic gains are allocated — strong governance reduces managerial rent extraction and can promote more equitable/efficient sharing of AI benefits.
- Mechanisms to model: Financing friction alleviation, productivity improvements, and human capital revaluation are key channels linking AI adoption to compensation outcomes; these should be incorporated into theories of AI diffusion and firm behavior.
- Policy and managerial implications:
- For regulators: strengthen governance, disclosure, and compensation oversight to limit rent capture and ensure AI gains support productivity and broad welfare.
- For firms: align incentives and governance structures to ensure AI investments translate into firm value rather than disproportionate managerial pay increases.
- For labor and training policy: expect rising demand for managerial and technical human capital; invest in upskilling to spread gains more broadly.
- Research directions: quantify effect sizes across industries; identify the most effective governance mechanisms; evaluate welfare implications at aggregate level and interactions with labor market frictions.
Assessment
Claims (10)
| Claim | Direction | Confidence | Outcome | Details |
|---|---|---|---|---|
| Firm-level AI adoption raises executive compensation in Chinese A-share listed companies (2007–2023). Wages | positive | high | Executive compensation (aggregate or top-management pay) |
0.48
|
| The positive AI → executive pay relationship is robust to endogeneity controls, including instrumental variable approaches, and to multiple robustness checks. Wages | positive | high | Executive compensation (effect of AI on pay remains positive under IV and robustness checks) |
0.48
|
| The firm-level AI application indicator is constructed via textual analysis of corporate disclosures (e.g., filings/annual reports) to capture AI application intensity. Adoption Rate | null_result | high | AI application intensity measure (text-derived) |
0.48
|
| AI adoption alleviates financing constraints, and this channel contributes to higher executive compensation. Firm Productivity | positive | medium | Financing constraints (mediator) and executive compensation (outcome) |
0.29
|
| AI adoption increases firm total factor productivity (TFP), and higher TFP is associated with higher executive compensation. Firm Productivity | positive | medium | Firm total factor productivity (mediator) and executive compensation (outcome) |
0.29
|
| AI adoption raises executives' human capital/market value, which contributes to higher compensation. Skill Acquisition | positive | medium | Executive human capital/market value (mediator) and executive compensation (outcome) |
0.29
|
| Stronger internal corporate governance weakens the AI → executive pay relationship, consistent with governance limiting managerial rent capture during technological change. Governance And Regulation | negative | medium | Interaction effect on executive compensation (AI × corporate governance) |
0.29
|
| Data consist of a panel of Chinese A-share listed companies covering 2007–2023. Research Productivity | null_result | high | Sample period and coverage (data description) |
0.48
|
| The empirical strategy uses baseline panel regressions with standard controls (e.g., firm size, performance, leverage) and fixed effects to estimate the AI → pay relationship. Wages | null_result | high | Executive compensation (estimation target in regressions) |
0.48
|
| AI adoption can raise upper-tail earnings within firms (executive pay), with potential implications for intra-firm income distribution and aggregate inequality. Inequality | positive | low | Upper-tail earnings / intra-firm income distribution (interpretive implication) |
0.14
|