Corporate social responsibility disclosure in Saudi companies: analysing the impact of board independence in family and non-family companies

The study investigates whether family control affects corporate social responsibility practices disclosure (CSRPD) among Saudi Arabian listed companies and whether board independence moderates this relationship. CSR is a salient corporate concern as stakeholders expect transparency regarding employee welfare, environmental impacts, community engagement, governance, and human rights. Prior research links CSR disclosure to financing, audit risk, performance, and investment efficiency, yet disclosure quality and quantity vary. Given Saudi Arabia’s nascent CSR reporting landscape, significant family ownership prevalence, and regulatory emphasis on CSR (e.g., SCCG 2017), the authors examine: (1) whether family control is associated with CSRPD, and (2) how board independence influences the family control–CSRPD nexus. Grounded in stakeholder and socioemotional wealth (SEW) theories versus agency theory, the study’s purpose is to clarify mixed evidence and provide context-specific insights for Saudi Arabia.

The review contrasts stakeholder theory and SEW theory, which predict higher CSR engagement and disclosure by family firms to safeguard legitimacy, reputation, and long-term orientation, with agency theory and the expropriation hypothesis, which suggest family firms may prioritize private benefits over transparency, potentially reducing CSR disclosure. Empirical evidence is mixed: some studies report higher CSR or stakeholder orientation in family firms across various regions, while others document lower CSR performance or disclosure linked to family control. The review also highlights the role of country-level institutions and governance characteristics in shaping CSR behavior. Board independence is discussed as a key governance mechanism that can enhance monitoring, transparency, and responsiveness to stakeholders, though evidence on its direct effect in family firms is mixed. Hypotheses: H1 posits a negative association between family control and CSRPD; H2 posits that board independence positively moderates the family control–CSRPD relationship.

Data comprise Saudi non-financial firms listed on Tadawul from 2016–2021. After excluding banks/insurance and missing data, the final sample includes 837 firm-year observations across 11 sectors. CSR disclosures, family control, and governance data were manually collected from annual reports; financial data came from Thomson Reuters DataStream. Dependent variable (CSRPD quality) was measured via manual content analysis using a self-constructed 37-item checklist across six categories (community, customer, employees, energy, environment, products/services), informed by prior literature, GRI-G4, ISO 26000, and CMA ESG 2022 guidelines. Each item scored 0–3 (0: not disclosed; 1: generic qualitative; 2: detailed qualitative; 3: quantitative). The firm’s CSRPD score equals the sum of item scores divided by 37. Independent variables: Family control measured as percentage family ownership (with robustness to threshold dummies at ≥5%, ≥10%, ≥15%, ≥20%); board independence (BIND) as the percentage of independent directors. Controls include board size, ownership concentration (>5%), institutional ownership, foreign ownership, firm age (ln years), sales growth, ROA, firm size (ln assets), leverage, market-to-book, loss dummy, and beta; plus sector and year fixed effects. Baseline model: CSRPD = β0 + β1 Family + β2 BIND + β3 Family*BIND + controls + sector dummies + year effects + ε. Estimation uses POLS with two-way cluster-robust SEs (firm and year) and winsorization at 1st/99th percentiles. Robustness checks: alternative CSR measure (quantity via dichotomous 0/1 scoring for the same 37 items), alternative family thresholds, subsample analysis by high/low BIND (median split), alternative estimators (Newey-West, PCSE, logistic on CSR median dummy), and endogeneity tests (Heckman two-stage selection and IV-2SLS using ln assets, ln assets squared, and beta as instruments for Family to obtain PrFamily).

- Descriptive statistics: Mean CSRPD quality score 0.991 (median ≈0.919) on a 0–3 scale; board independence averages 47.6%. Family firms exhibit significantly lower CSRPD and lower values across several CSR dimensions; they also have lower BIND, smaller boards, lower ownership concentration, and lower institutional/foreign ownership. They are older, smaller, and have lower sales growth than non-family firms. - Baseline regression (POLS, two-way clustered): Family control negatively associated with CSRPD (β = -0.004, p < 0.001). Economic magnitude: a one standard deviation increase in Family ownership (15.605) corresponds to ~6.29% decrease in CSRPD relative to the mean ([15.605*0.004]/0.991). BIND’s main effect is not significant, but the interaction Family*BIND is positive and significant (β = 0.040, p < 0.05), supporting H2. Among controls, firm size (FSIZE) is positive and significant; foreign ownership (FOWN) is negative and significant; leverage (LEVEG) negative; MTB positive; beta negative. - Alternative CSR measure (quantity): Family remains negative and significant (β ≈ -0.002, p < 0.01); Family*BIND remains positive and significant (β ≈ 0.019, p < 0.05). - Alternative family thresholds (≥5%, ≥10%, ≥15%, ≥20%): The negative association between Family and CSRPD persists; Family*BIND positive and significant across thresholds, indicating the moderating role of BIND is robust. - Subsample by BIND: In low-BIND firms, Family is negatively and significantly associated with CSRPD (β ≈ -0.007, p < 0.01); in high-BIND firms, the association is not significant, indicating BIND attenuates the negative effect. - Alternative estimators: Newey-West, PCSE, and logistic regressions confirm the negative Family effect and positive interaction. - Endogeneity checks: Heckman two-stage and IV-2SLS show Family (or PrFamily) remains negatively associated with CSRPD (e.g., PrFamily β = -0.012, p < 0.01) and the interaction with BIND remains positive and significant (e.g., PrFamily*BIND β ≈ 0.018, p < 0.1), supporting robustness. - Overall: H1 (negative Family–CSRPD association) accepted; H2 (positive moderating effect of BIND) accepted.

Findings indicate that in Saudi Arabia, family control is associated with lower CSR disclosure quality and quantity, aligning more closely with agency theory and the expropriation hypothesis than with stakeholder or SEW-driven predictions. Possible mechanisms include prioritization of family objectives and short-term financial gains, reduced perceived need for external capital and legitimacy, concerns about revealing sensitive information, and lower reliance on CSR disclosure as a control mechanism due to reduced owner–manager information asymmetry. Crucially, higher board independence mitigates this tendency: independent directors enhance transparency, broaden stakeholder orientation, and encourage formal CSR reporting, thereby operationalizing the prosocial potential of SEW by tempering its drawbacks. The moderating role of BIND explains heterogeneity in prior evidence on family firms’ CSR behavior and underscores governance context as pivotal. Policy implications suggest enhancing board independence to improve CSR reporting credibility and stakeholder trust in family-controlled firms.

The study contributes context-specific evidence from Saudi Arabia showing that family control is negatively associated with CSR practices disclosure, while board independence attenuates this effect and promotes CSR transparency. By constructing a tailored 37-item CSR disclosure index and applying extensive robustness tests (alternative measures, thresholds, estimators, and endogeneity controls), the study provides strong support for H1 and H2. Practically, regulators and firms should strengthen board independence and adopt stakeholder-oriented governance to foster credible CSR reporting in family firms. Future research should extend to other Arab countries, explore the unique roles of royal family involvement and other governance features in CSR disclosure, and examine SEW selectivity dynamics over firm life cycles.

- Single-country focus (Saudi Arabia) limits generalizability across the Arab region and other institutional settings. - Measurement relies on manual content analysis of annual reports, which may not capture all informal CSR activities or external communications. - Despite extensive controls and endogeneity remedies (Heckman, IV-2SLS), unobserved factors may remain. - Data availability constraints (datasets available upon reasonable request) may limit replication breadth. - Specific family-control nuances (e.g., royal family presence, family management vs ownership) warrant deeper analysis in future studies.