Enhancing the corporate social & environmental responsibility of Chinese energy enterprises: A view from the role of management compensation incentive

China’s energy sector, a major driver of global energy demand and CO₂ emissions, faces increasing pressure to balance energy provision with environmental protection. CSR encompasses actions beyond legal requirements that improve social welfare, including employee, environmental, customer, and stakeholder-related responsibilities. In China, CSR measurement is supported by databases such as RKS and Hexun. Prior studies show multiple external and internal determinants of CSR and mixed effects of CSR on firm outcomes. Managerial characteristics (e.g., CEO traits, governance, compensation) are known to influence CSR/CER, and there may be bidirectional causality between CSR and executive pay. This study focuses on the causal impact of management compensation (salary and equity incentives) on CSR and CER for Chinese energy enterprises, addressing potential endogeneity via instrumental variables. Research questions: (1) What are the causal effects of management compensation incentives on CSR and CER? (2) How do salary versus equity incentives differ in their effects? (3) How do effects vary across sub-industries, ownership types, and regions? (4) What are implications for energy enterprises’ CSR/CER? The study contributes by identifying causal effects using IVs at the enterprise level and by focusing on China’s energy sector, central to carbon peak/neutrality goals.

CSR research documents varied external (political, economic, cultural) and internal (risk management, leadership, strategy, culture, finance) determinants of CSR. CSR can enhance performance, reduce cost of debt, spur innovation, and improve reputation and customer trust, with implications for competitive advantage and stock price risks. Managerial traits (marital status, confidence, CEO inside debt, board structure, gender mix) and compensation structures also matter. Evidence on the link between compensation and CSR/CER is mixed: cash pay can positively affect environmental performance while equity ownership may reduce it; pay restrictions can reduce CSR; nonlinear (inverted U) relations are reported. Conversely, CSR may influence executive pay via profitability improvements, though some studies find negative relations. These mixed findings imply potential two-way causality. Theoretical framing: agency theory suggests higher cash compensation can motivate executives and may lead to CSR overinvestment (reputation building, risk management), while equity incentives align managers with shareholder value, potentially discouraging CSR with high costs and delayed payoffs—especially given China’s volatile stock market. Hypotheses: H1 salary incentives positively affect CSR & CER; H2 equity incentives negatively affect CSR & CER.

Empirical strategy: panel data of Chinese listed energy enterprises (2010–2021) across eight sub-sectors (electric power, power generation equipment, electrical grid, gas, oil & natural gas, coal, energy equipment, environmental protection). Dependent variables: CSR total score and CER (environmental responsibility sub-index) from Hexun (0–100 scale; CSR weights: shareholder 30%, employee 15%, supplier/customer/consumer 15%, environmental 20%, social 20%). Key independent variables: management salary incentive (log of total executive compensation) and equity incentive (management shareholding ratio). Managers include directors, supervisors, and senior managers. Controls: CEO duality, total liability ratio (Lev), Tobin’s Q, ROE, firm age (years since IPO), firm size (log employees), shareholder concentration (HHI of top ten shareholders), largest shareholder proportion (Top1), number of senior managers. Baseline model: firm and year fixed effects with interactive fixed effects to absorb initial characteristics: CSR(CER)_{it} = β0 + β1 Incentive_{it} + X_{it}γ + θ_i + μ_t + ε_{it}. Endogeneity and reverse causality are addressed via instrumental-variable (IV) regressions with 2SLS and interactive fixed effects. Salary IV (IV_S): interaction between industry-level total executive salary payment and the firm’s initial (first five years) ROA rank (Bartik-style). First stage: Salary_{it} on IV_S plus controls, firm FE, year FE, and an interaction of initial firm financial performance with year FE. Equity IV (IV_E): interaction between industry management shareholding ratio and the firm’s initial (first five years) controlling shareholder ownership rank; first stage analogous with interaction of initial ownership and year FE. Second stage regresses CSR(CER) on fitted Salary or Equity with the same fixed effects and controls. Robustness checks include: adding further controls (proportion of independent directors, cash ratio, growth rate of owners’ equity); redefining incentives per manager (per_Pay, per_Share); subsample excluding post-2018 due to CER rating change; balanced panel estimations; reduced-form and exogeneity checks where endogenous variables remain significant and IVs do not, supporting exclusion restrictions. Data sources: CSR/CER from Hexun; firm financials and governance from CSMAR; industry classification from Wind. Sample: 3,272 firm-year observations after excluding ST/*ST*/PT and pre-listing data (2010–2021).

- Baseline FE and IV results: Salary incentives significantly increase CSR; equity incentives significantly decrease CSR. IV estimates suggest larger absolute effects than FE, implying prior underestimation due to endogeneity. - Magnitudes (CSR): IV coefficient on Pay ≈ 15.213; on Share ≈ −0.774. Authors’ interpretation: a 1% increase in total management salary increases CSR by about 15 points; a 1% increase in management shareholding reduces CSR by 0.774 points. - Magnitudes (CER): IV coefficient on Pay ≈ 5.645; on Share ≈ −0.245. Authors’ interpretation: a 1% increase in total management salary increases CER by 5.645 points; a 1% increase in management shareholding reduces CER by 0.245 points. - Control variables: CEO duality is consistently negative for CSR/CER; higher leverage reduces CSR; some evidence that higher Top1 increases CSR/CER; greater shareholder concentration (HHI) often negative. - Instrument validity: First-stage coefficients for IV_S and IV_E are strong; underidentification and weak IV tests reject weak instruments; endogeneity tests often significant, supporting IV use. - Industry heterogeneity: No significant difference in salary incentive effects across groups, but equity incentive’s negative effect is stronger in the electricity and environmental industry than in conventional energy/equipment manufacturing (significant negative interaction with EL_EN). - Ownership heterogeneity: Salary incentives have a greater positive effect on CSR in state-owned enterprises (SOEs) (significant interaction with SOE); equity incentive effects do not significantly differ between SOEs and non-SOEs. - Regional heterogeneity: No significant regional differences between eastern coastal and inland firms in the effects of salary or equity incentives on CSR/CER. - Robustness: Results remain consistent when adding extra controls, redefining incentives per manager, restricting to pre-2019 due to CER rating change, and using balanced panels. Reduced-form tests suggest IVs affect outcomes only through the endogenous variables.

Findings align with agency theory and the overinvestment hypothesis: higher salary incentives motivate managerial effort and can increase engagement in CSR/CER, potentially due to reputational benefits and risk management considerations. Conversely, equity incentives align managers more tightly with short-term shareholder value, discouraging CSR/CER investments that have high costs and delayed or uncertain payoffs, particularly in a volatile policy-driven stock market. The stronger negative equity effect in electricity and environmental industries suggests heightened sensitivity where environmental externalities and regulatory scrutiny are greatest. The larger positive salary effect in SOEs is consistent with their broader social objectives and looser focus on profit maximization, coupled with information asymmetries that make CSR investment an appealing reputational strategy for managers. Overall, disentangling salary and equity components clarifies their opposing roles in shaping CSR/CER behaviors in energy firms.

This paper identifies the causal impacts of managerial compensation structures on CSR and CER in Chinese energy enterprises. Using IV strategies on 2010–2021 panel data, salary incentives are shown to significantly promote CSR and CER, while equity incentives significantly deter them. The negative equity effect is more pronounced in electricity/environment industries, and salary incentives have a stronger positive effect in SOEs; effects do not differ by region. Policy and managerial implications: firms should emphasize performance-linked salary incentives for managers and avoid excessive managerial equity stakes that may suppress CSR/CER; governments should strengthen governance of SOEs and consider limiting managerial shareholding ratios while developing complementary incentive systems to support CSR/CER goals. Future research may incorporate additional compensation components and governance factors, alternative CSR/CER metrics, and extended datasets to validate and generalize these findings within and beyond the energy sector.

- External validity: The sample is limited to Chinese listed energy enterprises (2010–2021); results may not generalize to unlisted firms, other sectors, or countries. - Measurement: CSR/CER measures rely on Hexun ratings; CER scoring methodology changed in 2018 (addressed via subsample tests), which may affect comparability across years. - Endogeneity and instruments: Although multiple robustness and exogeneity checks support the IV strategy, identification rests on assumptions about the exogeneity of Bartik-style instruments constructed from industry aggregates and initial ranks. - Data structure: Unbalanced panel with missing values in some years (mitigated by balanced-panel robustness checks) may still introduce selection concerns. - Variable scaling/interpretation: Salary incentive is modeled as log compensation; interpretations of marginal effects depend on functional form and may be sensitive to scaling assumptions.