Environmental Liability Disclosure and Quality Sustainability Reporting: Evidence from Listed Agricultural Firms in Kenya

Sustainability reporting quality remains a critical challenge for agricultural firms listed on the Nairobi Securities Exchange (NSE), with many firms failing to meet international Environmental, Social, and Governance (ESG) standards despite growing stakeholder pressure for environmental accountability. Limited empirical evidence exists on whether disclosure of environmental liabilities such as asset restoration obligations, pollution clean-up liabilities, environmental fines and penalties translates into higher-quality sustainability reports among Kenyan agricultural firms. This study examined the influence of environmental liability disclosure (ELD) on quality sustainability reporting (QSR) of listed agricultural firms in Kenya, guided by Legitimacy, Stakeholder, Institutional, and Disclosure theories. An explanatory research design with a longitudinal panel approach was adopted, covering all seven listed agricultural firms (complete sector census) over 2015–2024, yielding 70 maximum firm-year observations. Secondary data were obtained through systematic content analysis of annual reports, financial statements, and ESG filings. QSR was measured using a validated eleven-dimension index (Relevance and Reliability sub-dimensions, maximum score 21), while ELD was measured using financial ratios. Data were analysed using descriptive statistics, correlation analysis, and panel regression (OLS), following poolability and Hausman specification tests. Findings revealed that ELD had a negative, statistically insignificant relationship with QSR (β = -0.0042, p = 0.954), leading to non-rejection of the null hypothesis. Descriptive statistics showed a right-skewed ELD distribution (mean = 0.2743, SD = 0.3372, N = 58) driven by episodic liability recognition rather than a sustained disclosure trend, while correlation analysis confirmed the weakest and least consistent association of ELD with QSR among the four green accounting dimensions examined (Pearson r = 0.020; Spearman ρ = 0.002). This non-significance persisted across the bivariate model, the direct-effects model (Model 2, R² = 0.3681), the ELD×Environmental Management Accounting interaction (Model 4), and the full mediation model (Model 8, R² = 0.4407), indicating that the result is robust rather than an artefact of model specification. The study concludes that environmental liability disclosure, in its current voluntary and inconsistent form, does not meaningfully drive sustainability reporting quality among Kenyan agricultural firms. The study recommends mandatory environmental liability disclosure requirements aligned with GRI standards, development of Kenya-specific environmental accounting standards, and stronger regulatory enforcement to enhance disclosure credibility and comparability.