Environmental, social and governance reporting, audit committee characteristics and readability of annual reports: evidence from an African emerging economy

It is the responsibility of the management of organisations to prepare financial and annual reports and, more importantly, ensure they are in the right form for stakeholders' use. The general-purpose financial reports provide varied information to all kinds of stakeholders depending on the information needs of a user. Moreover, the content of these reports plays a significant role in informing capital market participants' investment and economic decision-making, and also plays a role in how efficiently and effectively capital markets function. Notwithstanding the preceding relevance of the content of financial reports, the conceptual framework of financial accounting/reporting as per the International Financial Reporting Standards (IFRS) requires certain attributes to be achieved/present if financial reports are to perform their intended roles or serve their intended purpose (Birt et al., 2017). Relevance and reliability are fundamental qualities financial reports are to possess, while enhancing qualities include timeliness, verifiability, understandability and comparability (Nobes and Stadler, 2015). It is worth mentioning that none of these qualities function in insolation, although fundamentally, all reports are first and foremost expected to be relevant (i.e. have predictive as well as confirmatory value) and must be reliable (i.e. free of error, neutral and completely contain all material information).

At the centre of any financial report and its content making a difference (in terms of being relevant and reliable) is its ability to be understood (Jonas and Blanchet, 2000). From the preceding, it stands to reason that a financial/annual report that cannot be understood (incomprehensible) can be robbed of its relevance and reliability (Smith and Smith, 1971). From literature, the understanding is that a more readable financial/annual report is one that has a higher probability of being understood (Seifzadeh et al., 2021; Telles and Salotti, 2024). Notwithstanding the demand on firms to ensure the interests of different stakeholders, including the economic goals of shareholders, are met, it is imperative for organisations to produce financial and annual reports that are readable since it has implications for the kind of decisions that are made by stakeholders.

Consequently, given the espoused importance of financial/annual report readability, there have been efforts to study those factors that impact financial/annual report readability. From literature, some key issues that have the ability to affect the readability of annual reports include business strategy (Lim et al., 2018; Habib and Hasan, 2020), culture (Noh, 2021), corporate social responsibility (CSR) performance (Bacha and Ajina, 2020), board independence (Rahman and Kabir, 2024), board secretaries (Sun et al., 2023), political corruption (Xu et al., 2022), product market competition (Rahman et al., 2024), female board participation (Ginesti et al., 2018) and earnings management (Shauki and Oktavini, 2022). Altogether, the insight drawn from these prior works is that multiple issues can either impede or improve the readability of annual reports of firms, hence the need for firms to widen their tentacles to identify what helps and what does not help in promoting readability and understandability of financial reports.

The global sustainability crisis has pushed non-financial disclosure/reporting conduct into the spotlight of governments, regulators, researchers and policymakers alike, largely as a response to mounting concerns over climate change and environmental degradation. Out of this shift, environmental, social and governance (ESG) practices have emerged as a central theme in organizational discourse, built on the premise that firms attending to these dimensions help ensure a business environment capable of meeting the needs of both present and future generations. As awareness of ESG's importance has grown, so too has stakeholder pressure for firms to disclose their ESG activities transparently within annual reports. This growing demand has in turn spurred a wave of scholarship examining whether such disclosures actually matter, with findings that remain far from uniform. Existing work has linked ESG disclosure and performance to a range of outcomes which includes financial performance (Friede et al., 2015), the quality of financial reporting and investment efficiency (Ellili, 2022), sustainable corporate growth (Oprean-Stan et al., 2020), broader economic growth (Hassani and Bahini, 2022), the cost of debt (Raimo et al., 2021), market dynamics and investment efficiency (Xue, 2025) and even portfolio outcomes (Bermejo Climent et al., 2021).

Yet, despite this expanding body of research, a clear gap remains, especially in developing economies like Ghana, concerning whether ESG disclosure shapes how readable annual reports actually are, and whether the characteristics of audit committees play a role in that relationship. This article sets out to close that gap using data from Ghanaian listed firms to assess how ESG reporting affects annual report readability, while also considering audit committee attributes as a possible internal governance mechanism that conditions this link. Although prior studies offer useful insight, none fully settle the matter. Ndegwa (2024), for example, found that sustainability disclosures improve the clarity of financial statements and also moderate how board diversity and earnings management relate to statement readability. Elsewhere, Xu et al. (2025) showed that firms with stronger ESG performance tend to produce more readable analyst reports, and Shimamura et al. (2025) found a positive relationship between readability and ESG scores, alongside a negative one between readability and volatility in ESG ratings. Country-level findings tell a similarly mixed story, where in France, for instance, stronger CSR performance corresponds with more readable annual reports (Bacha and Ajina, 2020), and in the United States, CSR performance is associated with clearer CSR reporting (Wang et al., 2018). By contrast, Ben-Amar and Belgacem (2018), studying Canadian firms, found that better social performance was tied to more convoluted MD & A disclosures, suggesting firms may sometimes use complexity strategically as a form of impression management.

These contradictions raise a deeper question about how ESG disclosures truly affect the comprehensibility of annual reports, particularly in emerging markets where both reporting practices and governance structures can vary widely. This uncertainty draws attention to the possible moderating role of internal governance structures, especially audit committees, an area that remains only partially explored. Velte (2018a) found that audit committees with financial and sustainability expertise tend to produce more readable integrated reports, with the combination of both types of expertise proving more effective than either alone. In a separate study, Velte (2018b) also found that greater gender diversity on UK audit committees improved the clarity of key audit matters. Likewise, Zheng et al. (2025) reported that audit committee size, audit quality, gender diversity and board independence all contributed positively to the readability of Pakistani annual reports. Even so, while these studies affirm that audit committee characteristics matter for reporting quality broadly, none has directly tested whether such characteristics moderate the ESG-readability relationship, despite the audit committee's well-established role at the heart of firms' financial reporting and accounting processes.

Beyond these unresolved tensions, many studies are rooted in institutional contexts quite different from Ghana's own governance and regulatory landscape, with this study drawing further motivation from local evidence uncovered by Gyasi and Owusu-Ansah (2018) and Gyasi (2019). The work by Gyasi and Owusu-Ansah (2018) examined SSNIT's annual reports from 2011 to 2015 and established that the documents were generally hard to understand and grew even less readable over time. In addition, the work by Gyasi (2019) focused on four Ghanaian banks between 2013 and 2016 and found that dense vocabulary and overly long sentences considerably weakened report clarity. Given that both organizations operate under strict regulatory oversight, these findings point to a readability problem that may run deeper across Ghana's corporate reporting landscape as a whole. Seeking to understand what drives these patterns, Gu and Dodoo (2020) studied Ghanaian listed firms and found a positive association between firm performance and readability. Despite the preceding, little is known about whether organizational or governance-related factors like audit committee characteristics help explain the effect of ESG disclosure on readability.

Overall, three gaps are identified from the extant literature concerning the subject matter under investigation that this research attempts to bridge. First, there are conflicting results from empirical studies on the connection between ESG disclosures and the readability of annual reports, as studies have found a positive correlation (Bacha and Ajina, 2020; Wang et al., 2018; Xu et al., 2025), a negative relationship (Ben-Amar and Belgacem, 2018) and some a mixed one (Shimamura et al., 2025). Secondly, even though audit committee attributes have been proven to influence annual reporting readability directly (Velte, 2018a, b; Zheng et al., 2025), there are no studies that have investigated their moderating effects on the relationship between ESG disclosure and reporting readability. Third, there is limited research on ESG disclosures and their readability, with several of the existing evidence on advanced countries and Asia with little attention paid to the relationship between ESG disclosures and readability in emerging African economies such as Ghana, where ESG reporting is nascent and governance structures differ markedly (Gyasi and Owusu-Ansah, 2018; Gyasi, 2019; Gu and Dodoo, 2020). The current research seeks to address these research gaps by analysing the relationship between ESG disclosures and readability in Ghanaian listed companies, and exploring whether audit committee characteristics moderate the relationship.

Informed by the above-mentioned gaps, this research aims to achieve two major objectives/goals. First, this research seeks to analyse the impact of ESG disclosure on the readability of annual reports of companies listed on stock exchanges in Ghana. Second, this research aims to determine whether certain features of audit committees (namely, gender diversity, meeting frequency and size) moderate the impact of ESG disclosure on annual report readability. Addressing the stated goals, this research provides new empirical findings on the joint impact of non-financial disclosure and corporate governance on reporting quality by way of reporting readability. Country-specific research such as this matters because regulatory maturity, governance norms and sustainability reporting practices vary widely across national contexts. While ESG disclosure is well established in many advanced economies, it remains nascent in developing contexts like Ghana. Studies from such settings offer valuable insight into how policy and regulation shape corporate ESG reporting, informing future policy refinement. The intuition behind the linkage is that as non-financial reporting often introduces technical language, narrative and visuals that make reports bulkier and harder to follow, this can potentially undermine their usefulness by impairing their readability. Building on this, the study also examines whether audit committee attributes moderate the ESG disclosure-readability relationship.

The literature has long emphasized governance mechanisms, particularly the board of directors, in driving firm performance through their role in shaping internal controls. Boards rely on specialized committees, and the audit committee stands out as among the most influential (Abu and Jaffar, 2020). Audit committees are typically responsible for overseeing financial reporting integrity, safeguarding resources, maintaining internal controls and supporting audit functions (Buallay and Al-Ajmi, 2020; Zadeh et al., 2023). Given this oversight role, audit committees may plausibly shape how ESG disclosure affects readability. This is because even if ESG reporting adds length and complexity that makes reports denser and bulkier, a well-resourced audit committee could help offset this through its influence on the reporting process. Accordingly, this study examines how audit committee gender diversity, meeting frequency and size each moderate the ESG-readability relationship. The study's main contribution lies in exploring this conditioning effect within an African emerging economy context.

The rest of the manuscript is structured into literature review, methods and data, result presentation and discussion and conclusion and recommendations.

The study draws on the obfuscation hypothesis in discussing the effect of ESG reporting on annual report readability. The premise of the hypothesis is that managers may increase the complexity of disclosures by way of adding excessive details, technical jargons or non-financial narratives, which has the ability to result in lengthy annual reports and lower the ease with which stakeholders can read and interpret reports (Courtis, 1998; Hassan et al., 2019). The purpose of ESG reporting, which is to provide transparency on ESG practices of firms, may introduce lengthy qualitative disclosures. This introduction of increased qualitative disclosures may not only crowd out financial information and make reports bulky, unreadable and understandable, but may also serve as a tool that can divert the attention of stakeholders from less favourable financial outcomes. In effect, drawing on the obfuscation theory, the study argues that ESG reporting has the potential to impair the readability of annual reports.

To explain the moderating effect of audit committee characteristics in how ESG reporting impacts report readability, the study draws on the agency theory and the resource dependence theory. Boards serve as the channel through which the activities of management are monitored to ensure alignment with the interests of capital providers of firms, which is consistent with the position of the agency theory (Meckling and Jensen, 1976; Bonazzi and Islam, 2007; Boivie et al., 2016). One of the key committees through which boards function is the audit committee, which serves as a key monitoring mechanism that mitigates managerial opportunism and other unethical practices, particularly those related to information and data reporting. From the preceding understanding, the study argues that the audit committee and its associated characteristics serve as a tool that ensures that managerial opportunism is minimized by ensuring ESG disclosures are not used as a tool for obfuscation. The study argues that larger audit committees, increased and frequent meetings and gender diverse membership help to strengthen oversight over management, which compels management of firms to present ESG information in a much clearer, readable and understandable form. In addition, with the resource dependence theory (Pfeffer and Salancik, 2003) highlighting that diverse and well-structured boards and committees are critical for positive outcomes, the study further advances that a well-structured and diverse audit committee ensures the onboarding of critical expertise, perspectives and networks that improve the quality of disclosures. The presence of both females and males on the committee with varied backgrounds and traits as well as frequent committee engagements/meetings, will ensure that management gets the needed oversight and advice that ensures the integration of ESG information in the annual reports that support, rather than impair, annual report readability.

This governance-outcomes link is consistent with a growing empirical literature linking audit committee characteristics specifically to ESG-related outcomes across diverse institutional settings. Pozzoli et al. (2022) show that audit committee independence and expertise positively shape ESG performance among European Union firms, while Belouadah (2026) and Masmoudi and Alsmady (2025) report similar positive effects of independence on ESG disclosure quality and performance in Saudi Arabia and France, respectively. The preceding works, for instance, lend direct support to the resource dependence and agency-theoretic reasoning underlying this study's expectation that audit committee size, meeting frequency and gender diversity will similarly shape the quality of ESG-related disclosure, extending it further by proposing that these attributes condition not just the quality or extent of ESG disclosure, but its readability specifically.

Empirical literature shows that ESG and CSR reporting have implications for different firm outcomes, and a small but growing body of work now speaks directly to whether ESG reporting makes annual reports and related disclosures more or less readable (i.e. understandable). Overall, the existing body of work reveals that the literature does not converge on a single answer or issue. One cluster of studies reports a positive association between ESG/sustainability performance and readability. Xu et al. (2025) found that analyst report readability responds positively to firm ESG performance, and trace this to two mechanisms: greater information accessibility and greater analyst effort, with the effect strongest among firms in polluting industries, firms with opaque financial information and state-owned enterprises. Ndegwa (2024) similarly shows that sustainability reporting enhances the readability of financial statements, while Bacha and Ajina (2020) found a positive relationship between CSR performance and annual report readability. In the same vein, Bifulco et al. (2025), using a Fog index approach, found that stronger environmental performance is associated with more readable nonfinancial disclosures, a result they interpret through signalling theory where firms with genuinely good environmental performance have an incentive to disclose clearly in order to distinguish themselves from weaker performers. Notably, Bifulco et al. (2025) found no equivalent relationship for the social and governance dimensions, which is an important qualification often lost when ESG is treated as a single composite construct. Similarly, Kaur et al. (2025) find that firms with strong ESG performance tend to produce more detailed and readable disclosures under the environmental and social pillars specifically, while poor ESG performers publish shorter, less-readable narratives – directly reinforcing the signalling-consistent cluster, while also showing that the effect operates unevenly across ESG's sub-dimensions rather than uniformly across the composite construct. Yan (2024) studied Chinese listed firms and adds a further nuance where it was found that more comprehensive environmental disclosure was associated with higher MD&A readability and a more positive disclosure tone, and this relationship was partly indirect, operating through improved earnings quality rather than the environmental disclosure itself. Cao et al. (2025) and Shimamura et al. (2025) extend the positive-association finding to a different outcome variable and a different measurement approach respectively: the former show that more readable ESG reports reduce ESG rating divergence by lowering information asymmetry (though this effect weakens for firms already benefiting from high media attention or location in low-carbon pilot cities), while the latter, using a generative-artificial intelligence (AI)-based readability measure, find a positive correlation between readability and ESG scores that does not hold when readability is measured using conventional text-based (word-feature) indices. This last point is informative in that it suggests that some of the disagreement in the literature may be an artefact of measurement text-based readability formulas (e.g. Fog, Flesch) and semantic or AI-based readability measures do not necessarily capture the same construct, so studies using different instruments are, in effect, answering slightly different questions.

A second cluster of studies reports the opposite pattern, or at least complicates the “more ESG, more readable” narrative. Ben-Amar and Belgacem (2018) found that stronger corporate social performance is associated with greater MD & A textual complexity, consistent with the possibility that managers use CSR opportunistically, embedding sustainability narratives to obscure rather than clarify. Hu et al. (2024) and Li (2025) approach the question from the angle of greenwashing rather than performance per se, and both find that lower disclosure readability is associated with a higher degree of greenwashing, implying that readability itself may function as a signal of disclosure integrity, but that firms with something to hide have an incentive to keep disclosures complex. Huang et al. (2025) offer a further qualification relevant to reconciling these two clusters: they show that ESG ratings matter more for firm value when ESG reports are more readable, but that this moderating role of readability is itself weakened by firms' growth potential and institutional ownership. In other words, the value of readability is not constant across firms but depends on the information environment the firm already operates in, a point that helps explain why single-country or single-index studies produce different effect sizes and, in some cases, different signs.

Recent evidence from the wider ESG-performance literature, while not addressing readability directly, reinforces the view that ESG disclosure's effects on firm outcomes are rarely linear or straightforward. Works like Ghosh et al. (2023), which focused on Indian listed firms, found a U-shaped relationship between environmental disclosure and financial performance, with benefits only emerging once disclosure crosses a specific threshold. Similarly, Ghosh et al. (2024) found a non-linear effect of carbon performance on financial performance, though disclosure practices alone showed no such effect. Extending the logic espoused by the preceding works is that of Agarwala et al. (2024) who disaggregated ESG into its ESG components, finding a U-shaped relationship for the environmental component alongside linear-positive relationships for the social and governance components. Although these studies examine financial rather than readability outcomes, they collectively suggest that the relationship between nonfinancial disclosure and firm-level outcomes is highly sensitive to how disclosure is measured, disaggregated and thresholded, a caution equally relevant to interpreting mixed findings in the ESG-readability literature reviewed above.

Further sharpening the ESG-readability picture are recent related works which went beyond financial performance. For instance, Kaur et al. (2025) find that firms with strong ESG performance tend to produce more detailed and readable disclosures under the environmental and social pillars specifically, while poor ESG performers publish shorter, less readable narratives, directly reinforcing the signalling-consistent cluster, while also showing that the effect operates unevenly across ESG's sub-dimensions rather than uniformly across the composite construct. Further insight is provided by Sánchez-Hernández et al. (2026), who reported that readability gains are not purely a function of firm ESG performance but can also be investor-driven, as they established that firms with socially responsible investment ownership produce less linguistically complex CSR reports, an effect that is stronger among environmentally sensitive firms and when the investor is European, pointing to external stakeholder pressure as a further mechanism shaping disclosure clarity. Linking the preceding works with the non-linear, threshold-based findings of Ghosh et al. (2023, 2024) and Agarwala et al. (2024) on ESG disclosure and financial performance, the indication is that whether ESG disclosure clarifies or obscures firm reporting depends jointly on firm-level ESG performance and investor composition, factors this study's Ghanaian setting differs from markedly, given its comparatively thin institutional investor base and voluntary disclosure regime.

In sum, three points emerge from a critical reading of this literature rather than a simple listing of findings. First, the direction of the ESG-readability relationship appears to depend on which mechanism dominates in a given setting: a signalling mechanism (genuine performers disclose clearly to differentiate themselves) versus an obfuscation mechanism (weaker or opportunistic performers use ESG narrative to complicate disclosure and manage impressions). Studies using signalling-consistent settings (e.g. environmental performance in Bifulco et al., 2025; analyst-driven scrutiny in Xu et al., 2025) tend to find positive associations, while studies focused on complexity, textual tone or greenwashing incentives (Ben-Amar and Belgacem, 2018; Hu et al., 2024; Li, 2025) tend to find the reverse. Second, composite ESG measures may mask offsetting effects across the ESG components, as Bifulco et al. (2025) demonstrate directly. Third, almost all of this evidence, centring on Chinese, European and cross-country analyst-based samples, comes from developed or large emerging markets with comparatively mature ESG reporting infrastructures and enforcement regimes. However, the relevance of either the signalling or the obfuscation mechanism to a developing-economy, weaker-enforcement context such as Ghana remains untested.

Against this backdrop, and given that Ghanaian firms operate in an environment of comparatively voluntary, less standardized ESG reporting and weaker regulatory scrutiny, conditions under which the obfuscation hypothesis (that nonfinancial disclosure is more readily used to complicate rather than clarify annual reports, through excessive detail and technical jargon) is arguably more likely to dominate over a signalling mechanism that depends on the market's capacity to reward clear disclosure, the following hypothesis is proposed for testing:

Given the acknowledgement that the audit committee plays an integral role in the reporting processes of organisations (Ghafran and O'Sullivan, 2013), the study argues that although ESG reporting could impair the readability of annual reports, audit committee attributes (representing audit committee quality and effectiveness) may underlie the nexus between ESG reporting and annual report readability. The audit committee has oversight responsibility for internal controls, internal audit and accounting and financial reporting systems (Abbott et al., 2010; Zain et al., 2006). Drawing on Huang et al. (2025)'s finding that the value of readability is conditional on a firm's broader governance and information environment, it is plausible that the same logic extends to the production of readability: firms with stronger audit committee oversight may be better placed to prevent ESG disclosure from degenerating into the kind of unfocused, jargon-heavy narrative that the obfuscation hypothesis anticipates.

Building on this, and following existing literature (Arif et al., 2021; Buallay and Al-Ajmi, 2020), the study argues that a larger audit committee is more likely to possess the expertise, experience and knowledge needed to ensure that only relevant ESG issues are reported, supporting more readable annual reports. Likewise, more frequent audit committee meetings should allow ESG reporting matters to be more thoroughly discussed and deliberated, so that only issues of genuine priority to users are disclosed (Arif et al., 2021; Bravo and Reguera-Alvarado, 2019). Gender diversity on the audit committee is expected to work in the same direction: a more diverse committee draws on a wider range of expertise and perspectives in deciding what and how to report, reducing the risk that users are presented with excessive or irrelevant nonfinancial detail (Bravo and Reguera-Alvarado, 2019; Rezaei et al., 2022). Because women are often characterised as more stakeholder-oriented and risk-averse, their presence on the audit committee is expected to favour disclosure of only stakeholder-relevant nonfinancial issues, resulting in less cluttered and more readable annual reports. In sum, while the study advances through the lens of the obfuscation hypothesis that ESG reporting is likely to impair annual report readability by loading reports with excessive nonfinancial content, it argues that audit committee size, meeting frequency and gender diversity, as governance mechanisms grounded in agency theory and resource dependence theory, can counteract this tendency by filtering out disclosure that is not genuinely relevant to users, resulting in less verbose and lengthy reports which can aid readability.

The preceding theoretical reasonings are supported by a number of rapidly growing empirical literature linking audit committee characteristics to ESG disclosure quality and performance, even though only a small part of it examines readability directly. Evidence from Saudi Arabia (Belouadah, 2026), the European Union (Pozzoli et al., 2022), the ASEAN-5 region (Nurhandika et al., 2025) and France (Masmoudi and Alsmady, 2025) consistently finds that audit committee independence positively predicts ESG disclosure quality or performance, while findings on financial expertise and tenure are more mixed. In this regard, some works establish that expertise is positively associated with ESG outcomes (Belouadah, 2026; Pozzoli et al., 2022), while others find no significant effect (Nurhandika et al., 2025; Sihombing and Nurhaliza, 2025), and tenure is negatively associated with ESG disclosure quality across several contexts (Belouadah, 2026; Pozzoli et al., 2022; Masmoudi and Alsmady, 2025). Audit committee meeting frequency and size also feature prominently in the literature, though with less consistent results with studies like Buallay and Al-Ajmi (2020) finding that meeting frequency is positively associated with sustainability reporting among GCC banks, and Sihombing and Nurhaliza (2025) similarly find committee size, independence and meeting frequency all positively related to ESG performance among ASEAN-5 firms, whereas Ruziwa et al. (2025) find no significant effect of meeting frequency on ESG reporting quality among JSE-listed firms.

In addition to the preceding direct effects established, the extant literature reveals that audit committees can condition the governance-ESG relationships. For instance, Adnan et al. (2026) find audit expertise and independence moderate the ESG-financial performance relationship across twelve culturally diverse countries, Ma et al. (2024) show the audit committee moderates the relationship between board gender diversity and ESG disclosure among Chinese firms, and Metwally et al. (2025) find that audit committee characteristics strengthen the positive relationship between ESG disclosure and firm value in Saudi Arabia. One work that closely relates to the present study is the one by Gutiérrez Ponce et al. (2025), who established that corporate governance characteristics, including board independence and accounting expertise, directly and moderately shape the readability of the chairman's statement among Jordanian listed firms, offering rare direct evidence that governance attributes condition narrative readability specifically, rather than ESG performance or disclosure quality more broadly. Overall, this literature offers strong cross-country support for the general proposition that audit committee attributes shape the quality of ESG-related governance outcomes, even though the specific question of whether these attributes moderate the ESG disclosure-readability relationship, as this study investigates, remains largely untested outside of Gutiérrez Ponce et al.'s (2025) narrower focus on chairman's statements. Accordingly, the following hypothesis is proposed:

The study uses a correlational panel design to examine relationships/associations between variables, drawing on annual reports of 30 Ghana Stock Exchange (GSE)-listed firms from 2000 to 2023. The sample size was informed by data availability. For the period under consideration, the GSE had approximately 37 firms listed. However, the final sample of 30 was arrived at after excluding firms that had fewer than 4 consecutive years of annual report availability. The resulting panel is unbalanced, reflecting delistings, listings and gaps in annual report availability (particularly for scanned pre-2010 reports, as noted below), yielding the firm-year ranges (T = 13.6–17.8 years, n = 18–25 firms) reported for the moderating variables. Given the panel structure, fixed effects were selected as the estimator, supported by several diagnostic tests (Table 1). The Breusch–Pagan LM test showed the firm-specific variance component was zero and statistically insignificant (chibar²(01) = 0.00; p = 1.000), indicating random effects was not preferred over pooled ordinary least square (OLS), though this alone does not confirm OLS efficiency, since that depends on the error structure.

The Hausman test was therefore run to compare fixed and random effects, strongly rejecting the null of coefficient equivalence across all four specifications (p < 0.001), indicating unobserved firm characteristics correlate with ESG disclosure, audit committee characteristics and other regressors, confirming fixed effects as the appropriate baseline. Further diagnostics showed the standard fixed-effects assumptions were violated: Modified Wald tests confirmed groupwise heteroskedasticity (p < 0.001) in all models, Wooldridge tests indicated first-order serial correlation and Pesaran tests detected cross-sectional dependence in some specifications. To produce reliable inference despite these issues, the study uses the Driscoll–Kraay fixed-effects estimator (xtscc, fe), which is robust to heteroskedasticity, autocorrelation and cross-sectional dependence. All estimations were conducted in Stata using the xtscc command for Driscoll-Kraay standard errors with the fixed-effects option. Variance inflation factors (VIFs) were computed for all regressors in each specification to check for multicollinearity. All VIF values fell below the conventional threshold of 10, indicating multicollinearity was not a material concern despite the inclusion of interaction terms.

Fixed effects are also theoretically justified given that sample firms likely have stable, firm-specific traits (regulatory environment, industry, founding ownership structure) that could confound results. Controlling for these isolates within-firm variation, reducing omitted variable bias. While the Hausman test statistically supports FE, it does not identify the specific source of these effects; the theoretical reasoning above fills that gap. Since fixed effects relies solely on within-firm variation, the moderator needed meaningful within-firm movement; a moderator varying mainly between firms would be absorbed into the firm fixed effect, leaving little identifying variation for the interaction term. Meetings were expected to show the most within-firm variability (driven by firm-specific demands, regulation and audit complexity), gender diversity was expected to vary via director turnover and committee size was expected to vary least, given its typical stability under governance policy. This was confirmed: within-firm standard deviations were 0.46, 1.09 and 0.75 for gender diversity, meetings and size, respectively, representing 40.5%, 29.4% and 46.1% of total variance, all substantial (average T = 13.6–17.8 years, n = 18–25 firms), supporting inclusion in the fixed-effects framework with no indicator requiring a limitation caveat (see Table VII, Appendix). The empirical models follow Abdelazim et al. (2025) and are given as follows:

Because audit committee gender diversity, meeting frequency and size are conceptually distinct governance attributes rather than sub-components of a single index, each is entered into Equation (2) separately, together with its corresponding interaction term with ESG disclosure, producing three separate estimations of the moderation model rather than one specification with three simultaneous interaction terms. This avoids the multicollinearity that would arise from including three correlated interaction terms in a single equation and allows the moderating role of each audit committee characteristic to be assessed independently. To further mitigate multicollinearity between the constituent terms and their interactions, ESG disclosure and each audit committee characteristic were mean-centred prior to constructing the interaction terms, following standard practice (Aiken and West, 1991).

Equation (1) above estimates the effect of ESG disclosure on financial report readability, while equation (2) estimates the moderating effect of audit committee characteristics in the nexus between ESG disclosure and financial reporting readability. In both equations, RREAD stands for financial report readability while ESGdisclosure stands for environmental, social and governance disclosure. In equation (2), AuditComCha stands for audit committee characteristics and includes audit committee size, audit committee meetings and audit committee gender diversity. The control variables in the model includes Bdsize stands for board size, boardinde stands for board independent directors, Leverage stands for financial leverage, SoFirm stands for firm size, ROA represents return on assets, Loss represents loss reported by a firm in a particular year and firmage represents firm age.

The concept of annual report readability represents the ease with which investors, analysts, creditors and regulators can read and comprehend disclosed financial information. This matters beyond style as unreadable reports impair users' ability to extract decision-relevant information (Lawrence, 2013), raises information processing costs (Bloomfield, 2008; Lehavy et al., 2011), deepen information asymmetry (Li, 2008) and are linked to higher audit fees, lower analyst following and greater stock price crash risk (Lehavy et al., 2011; Kim et al., 2019; Abdelazim et al., 2025). In emerging markets, where institutional monitoring is weaker, readability carries added significance (Lara et al., 2017), making it particularly relevant in the Ghanaian context.

Traditional readability metrics, such as the Fog Index (Gunning, 1952), Flesch Reading Ease and Flesch-Kincaid Grade Level, rely on sentence length and word complexity, but these were calibrated on general prose, not financial disclosures (Loughran and McDonald, 2014). In technical reports, complexity often reflects necessary precision rather than poor communication (Bloomfield, 2008), and these indices can also be manipulated by simply shortening sentences. This study instead measures readability by document length (page count), since longer reports are harder for readers to process fully, consistent with reader interest theory (Chall, 1958) and evidence that length directly impairs information transmission (Luo et al., 2018). Page count does not fully capture textual clarity, so file size serves as a robustness check. Following Abdelazim et al. (2025), Loughran and McDonald (2014), Luo et al. (2018) and Cho et al. (2022), readability is proxied by page count in the audited annual report (transformed as −1 × ln(Pages), making higher values an indicator of greater readability). This choice is also partly practical as many GSE annual reports from 2000 to 2010 are scanned PDFs, making them largely unsuitable for word/sentence extraction and Fog/SMOG computation.

Endogeneity is partly addressed in three ways: fixed effects exploit within-firm variation to partly mitigate reverse-causality concerns; time-varying confounders (ROA, FirmLoss, firm size) are controlled directly to address omitted variable bias, and system GMM offers a further check against simultaneity, since this cannot be fully ruled out within fixed effects alone. Two additional issues are worth noting in relation to the obfuscation theory, where it is argued that managers lengthen reports to conceal poor performance (Li, 2008; Bushee et al., 2018), actually reinforce the readability measure rather than undermining it and this motive is controlled for via ROA and FirmLoss. Separately, since report length reflects both operating complexity and managerial discretion, firm size is controlled to isolate the discretionary component that is the focus of this study. An alternative dependent variable, PDF file size (logged, × −1), is used for robustness checks, following Loughran and McDonald (2014) and Cho et al. (2022), given the two length measures typically converge (Cho et al., 2022; Abdelazim et al., 2025).

The key independent variable is firms' ESG disclosure, constructed using a two-step procedure consistent with existing literature (Luh et al., 2024; Naveed et al., 2021), which relied on the Global Reporting Initiative (GRI) framework. First, a dummy-coding approach identified the presence of disclosures across the three ESG dimensions. Environmental reporting was assessed across five categories: greenhouse gas (GHG) emissions, energy usage, energy mix, water usage and environmental operations, each scored 1 if disclosed in a given year, otherwise 0. Social reporting was evaluated across seven areas: training and education, CEO pay ratio, gender pay ratio, employee turnover, gender diversity, non-discrimination, and health and safety, scored 1 if any appeared, otherwise 0. Governance reporting was assessed using nine categories: board diversity, board independence, ethics and anti-corruption, data privacy, disclosure practices, external assurance, regulatory compliance, stakeholder engagement and general corporate governance measures, each scored similarly. The annual ESG score was calculated by averaging the relevant dummy values, ranging from 0 to 1, with values closer to 1 indicating more extensive disclosure.

Audit committees are recognized as integral to firm success, particularly regarding internal controls, internal audit, financial reporting systems and external audit appointment (Alzeban, 2020; Phornlaphatrachakorn, 2020; Bananuka and Nkundabanyanga, 2023). Three indicators measure audit committee characteristics: gender diversity (number of female directors on the committee), meeting frequency (total annual audit committee meetings) and committee size (total directors on the committee).

The control variables were selected on the basis that each has been shown in prior literature to be associated with both firms' ESG/CSR reporting behaviour and their broader disclosure or reporting characteristics, making them plausible confounders of the ESG disclosure-readability relationship if omitted. Board size and independence, for instance, are established correlates of both CSR/ESG disclosure decisions (Rahman and Kabir, 2024; Ginesti et al., 2018) and reporting quality more generally, while firm size, leverage, profitability and loss status are routinely linked to both the extent of voluntary disclosure and the length/complexity of annual reports (Bacha and Ajina, 2020; Ndegwa, 2024; Bifulco et al., 2025). Their inclusion therefore isolates the effect of ESG disclosure on readability net of factors that could otherwise drive both variables simultaneously.

Board size, board independence, financial leverage, firm size, firm profitability, loss dummy and firm age are included as control variables, congruent with existing literature (Bacha and Ajina, 2020; Ndegwa, 2024; Bifulco et al., 2025). Bigger boards may possess the expertise for quality reporting decisions, so a positive relationship with readability is expected [(proxied by number of directors (Luh, 2025a)]. Independent directors, detached from daily operations, are expected to positively influence readability through stronger oversight (proxied by ratio of non-executive directors [(Luh and Kusi, 2023)]. Debt covenants incentivize clearer reporting for creditors, so leverage (total liabilities/total assets (Luh, 2025a) is expected to positively relate to readability. Firm size's effect is ambiguous: larger firms have more resources for efficient reporting, but face coordination challenges and greater operational complexity that could impair clarity, so either a positive or negative relationship is expected [(proxied by natural log of total assets (Luh, 2026b)]. Profitability (ROA) is expected to positively influence readability, as profitable firms can better invest in reporting systems, while loss-making firms, being cash-constrained, are expected to show a negative relationship [(measured as profit before tax divided by total assets (Luh, 2025b)]. Finally, firm age is expected to positively relate to readability, as older firms accumulate the expertise to identify and report only what is relevant (measured as years since incorporation (Luh, 2026a). See Table 2 for the measurement of variables.

Table 3 presents the descriptive statistics. Annual report readability (report length) averaged −4.014, ranging from −5.394 to −1.792, while ESG reporting averaged 0.173, ranging from 0.07 to 0.4. On audit committee characteristics, female representation averaged roughly 1 director (0.791), with a range of 0–3; meeting frequency averaged about 5 per year (4.5), ranging from 2 to 16; and committee size averaged about 4 directors, ranging from 0 to 7. In terms of control variables, on average, about 8 directors were on boards for the period under consideration, with roughly 6 being non-executive. In addition, the average financial leverage stood at 0.673 while firm size, profitability (ROA), loss-making and firm age averaged 15.641, 0.071, 0.241 and 37, respectively.

Table 4 reports pairwise correlations to check for multicollinearity among the independent variables. Following Lehmann's (1998) threshold of 0.9, all correlation coefficients fell below this level, indicating no multicollinearity, a result further confirmed by the VIF reported in Table 3.

Tables 5 and 6 presents regression estimates showing the effect of ESG disclosure and audit committee attributes on annual report readability. Table 5, Column 1, excluding any audit committee moderator, isolates the unconditional effect of ESG disclosure/reporting and serves as the baseline against which the moderated specifications in Columns 2–4 should be compared. The results reveal a negative and statistically significant relationship between ESG disclosure and readability (β = −1.7427, p < 0.05), supporting H1. This aligns with the obfuscation hypothesis (Courtis, 1998; Hassan et al., 2019), which holds that qualitative, non-financial narratives introduced through ESG reporting can crowd out financial information and increase report bulk rather than transparency. It is also consistent with Ben-Amar and Belgacem (2018), who found corporate social performance positively associated with MD&A textual complexity, interpreting this as evidence that managers may engage with CSR/ESG opportunistically to complicate disclosures, and with the greenwashing findings of Hu et al. (2024) and Li (2025), who linked lower disclosure readability to greater greenwashing, suggesting that poorly governed ESG communication can obscure rather than inform. This baseline result contrasts with the more optimistic literature like Xu et al. (2025), Ndegwa (2024), Bacha and Ajina (2020) and Bifulco et al. (2025), which found positive associations between ESG/CSR performance and disclosure readability. The findings suggest that, in the Ghanaian context and prior to accounting for governance quality, ESG reporting behaves more consistently with obfuscation-oriented findings than transparency-enhancing ones, reinforcing that this relationship's direction remains context-dependent and had not previously been tested in an African emerging-market setting.

Beyond statistical significance, the economic magnitude of this baseline effect merits attention before turning to the moderated results, since this is what the moderation analysis is ultimately measured against. Given the dependent variable is scaled as -ln(Pages), the coefficient of −1.7427 (SE = 0.6278) implies a substantial proportional increase in report length for a one-unit increase in ESG disclosure/reporting. The sample mean of ESG disclosure is 0.173 (SD = 0.053, range 0.07–0.40). Translating the coefficient into economic terms, a one-standard-deviation increase in ESG disclosure (0.053) is associated with an increase in ln(Pages) of 1.7427 × 0.053 ≈ 0.092, which corresponds to an approximate 9.7% increase in report length (calculated as [eˆ0.092 − 1] × 100%). Given the sample's relatively narrow range of ESG disclosure scores (0.07–0.40), an effect of this size at the mean is economically meaningful and provides a useful benchmark against which the subsequent moderated effects can be evaluated.

Columns 2–4 in Table 5 show that audit committee gender diversity, meeting frequency and size moderate the ESG-readability relationship, consistent with agency theory (Meckling and Jensen, 1976) and resource dependence theory (Pfeffer and Salancik, 2003): a well-resourced audit committee constrains managerial opportunism and channels the expertise needed to integrate ESG content clearly rather than as filler. Assessing this moderation's real impact requires examining the net conditional effect of ESG disclosure at given moderator level (β₁ + β₂M), since the interaction coefficient alone only shows how the ESG effect changes as the moderator increases, not where it turns or how that point compares to the sample's actual distribution.

For gender diversity (Column 2), the ESG main effect becomes insignificant (β = 0.4682, not significant.) once the moderator and interaction are introduced, while the interaction term is positive and significant (β = 1.7694, p < 0.05). Combined, these coefficients imply a net ESG effect of 0.4682 + 1.7694 × ACBDiv, which is positive across the entire observed range of audit committee gender diversity (ACBDiv: mean = 0.791, SD = 0.727, min = 0, max = 3 female directors) rather than emerging only past some threshold: the point estimate is positive even at zero female directors, though not statistically distinguishable from zero at that point, and strengthens as female representation on the committee rises. At the sample mean of roughly one female director (0.791), the implied net ESG coefficient is approximately 1.87, comparable in magnitude to and opposite in sign from the baseline ESG effect reported in Column 1. This aligns with Bravo and Reguera-Alvarado (2019) and Rezaei et al. (2022), who argue gender-diverse committees bring the expertise and stakeholder orientation needed to ensure only relevant, well-framed non-financial information is reported. These results extend this by showing the moderating effect is not merely attenuating but appears to fully reverse the sign of the baseline obfuscation effect even at modest levels of committee diversity, with the relationship growing stronger as the number of female directors increases towards the observed maximum of three.

For committee size (Column 4), the ESG main effect strengthens to −3.4746 (p < 0.01), while the interaction term is positive and significant (β = 0.5863, p < 0.10), implying a threshold of roughly six directors (3.4746/0.5863 ≈ 5.9). This threshold sits close to the upper end of the observed distribution (mean = 3.62, SD = 1.11, min = 0, max = 7), meaning only audit committees near the largest sizes actually present in the sample are able to fully offset ESG disclosure's negative effect on annual report readability; committees at or below the sample mean remain firmly in the range where ESG disclosure continues to reduce readability. This supports Arif et al. (2021) and Buallay and Al-Ajmi (2020), who argue larger committees provide the expertise depth needed to ensure only relevant ESG issues are reported. These findings sharpen this claim: full neutralization of ESG disclosure's readability cost requires committee sizes towards the top of the observed range rather than merely above average, and marginal increases from the sample mean (e.g. four to five members) are unlikely to be sufficient, a distinction the reviewed literature has not quantified.

For meeting frequency (Column 3), the coefficients warrant closer scrutiny before joining the same narrative as size and diversity. The ESG main effect is positive and insignificant (β = 1.5613, n.s.), while the interaction term is negative and significant (β = −0.7655, p < 0.05), opposite in sign to the other moderators. Solving for the net ESG effect (1.5613 - 0.7655 × ACMeetings) shows it turns negative once meeting frequency exceeds approximately two meetings per year (1.5613/0.7655 ≈ 2.04), essentially at the very bottom of the observed range (mean = 4.551, SD = 2.017, min = 2, max = 16). At the sample mean, the implied net effect is roughly 1.5613 - (0.7655 × 4.551) ≈ −1.92, larger in magnitude than the unmoderated baseline effect itself, and it grows more negative still towards the upper end of the distribution. This means increasing meeting frequency moves the net relationship more negative across nearly the entire empirical range, contradicting expectations from Arif et al. (2021) and Bravo and Reguera-Alvarado (2019) that frequent meetings enable better deliberation on ESG issues, which leads to the preparation of annual reports that are less complex and bulky, ensuring readability. This result suggests meeting frequency here reflects reactive governance: committees convene more often because existing disclosures are already problematic, rather than more frequent meetings improving communication and readability. This is an interpretation the literature has not directly tested, and one that this article's findings are positioned to raise as a notable contribution. The nuanced implication is that size and diversity offset the obfuscation effect, while meeting frequency does not, and given that the reversal point sits near the minimum of the observed distribution, it appears to compound the effect across virtually the full range of committee activity observed in the sample.

These findings underscore the audit committee's monitoring role, consistent with its positioning as a central oversight mechanism of organisations (Abbott et al., 2010; Ghafran and O'Sullivan, 2013). Effective audit committees, particularly diverse, adequately sized ones are better positioned to scrutinize narrative disclosures and enhance reporting quality, aligning with resource dependence rationale. In Ghana, firms strengthening audit committee structures can produce clearer ESG-integrated reports, though the threshold values suggest incremental governance improvements alone may not suffice; committees need meaningfully high size or diversity levels before the offsetting effect materializes, a nuance beyond the largely directional reviewed literature (Velte, 2018a, b; Mohammadi and Naghshbandi, 2019; Bravo and Reguera-Alvarado, 2019; Arif et al., 2021; Rezaei et al., 2022). Evidence on the ESG-report discourse as investigated by this study, if well positioned against prior evidence from other developing and emerging economies, helps clarify what is distinctive about the Ghanaian result rather than merely idiosyncratic to it. In this regard, lending support to the findings in this study as far as the moderating effect of audit committee characteristic is concerned is Zheng et al. (2025), who studied Pakistani listed firms, and similarly found that audit committee size and gender diversity positively predicted annual report readability, a directional finding consistent with the size and diversity results reported here, though their study examined these attributes as direct predictors of readability rather than as moderators of an ESG disclosure effect, making a like-for-like comparison of magnitudes difficult. More directly comparable in construct is the worky by Ndegwa (2024), who studied Kenyan listed firms and established that sustainability disclosure improved financial statement readability outright, a positive baseline association that contrasts with the negative baseline ESG-readability relationship documented here for Ghana. Also of relevance to this study is the evidence by Adhariani and du Toit (2020) from Indonesia, which established that sustainability reports exhibited persistently low readability and a pattern of isomorphic language across industries, which they attributed to a belief among firms that complex language impresses rather than informs stakeholders. This is a pattern more consistent with the obfuscation-oriented baseline finding reported here than with Ndegwa's positive Kenyan result.

This three-way divergence across African and Southeast Asian emerging markets is notable and can better inform policy, firm practices and further research. It suggests that even within broadly similar regulatory and institutional contexts, the direction of the ESG-readability relationship is not uniform and may instead depend on factors this study did not directly test, such as the maturity of each country's ESG reporting mandate, the extent of standardization in disclosure formats or differences in how readability itself was measured. This is particularly of importance because Ndegwa (2024) and Adhariani and du Toit (2020) use text-based readability formulas, whereas this study relied on document length. Evidence from developed markets adds a further point of contrast rather than convergence as Nilipour et al. (2020), specifically with reference to New Zealand, found that sustainability reporting readability improved by only 6.5% over a decade despite substantial growth in reporting volume, interpreting this stagnation as indicative of obfuscation risk even in a mature, well-regulated reporting environment. This suggests that Ghana's negative baseline finding is not simply an artefact of weak institutional development, since even firms operating under stronger regulatory oversight exhibit comparable readability stagnation. This reinforces a broader point raised earlier in the literature review, which is to the effect that comparisons of ESG-readability findings across countries are complicated by measurement heterogeneity as much as by genuine contextual or institutional difference. In effect, Ghana's negative baseline relationship should be read as evidence for one plausible pathway among several observed in comparable emerging-market settings, rather than as the definitive African or developing-economy pattern.

Overall, comparing the unconditional (Column 1 in Table 5) and conditional effects (Columns 2–4 in Table 5) shows the negative ESG-readability relationship isn't eliminated by ESG reporting itself, but is contingent on audit committee oversight strength and only above specific, quantifiable thresholds of size and diversity. This distinction between statistical and economically meaningful moderation is the central contribution intended here, extending the literature's directional argument into a threshold-based empirical account of how much governance strength is required before the offsetting effect occurs. The results discussed in Table 5 are largely consistent with the system GMM estimations presented in Table 6, which mainly help to reduce challenges linked with endogeneity that is prevalent in the majority of econometric estimations.

Table VII in Appendix reports average marginal effects of ESG disclosure on readability at representative values of each audit committee attribute, providing a direct empirical check on the threshold estimates derived analytically above. The pattern for gender diversity and committee size corroborates the algebraic thresholds closely: the ESG effect is statistically indistinguishable from zero at low levels of each moderator and becomes significantly positive (diversity) or loses its negative significance (size) only as the moderator approaches its upper range, consistent with thresholds of roughly one female director and six committee members, respectively. For meeting frequency, the marginal effect is a statistically insignificant near-zero at the observed minimum (dy/dx = 0.030, p = 0.968) but turns significantly negative from four meetings onwards and continues to deteriorate through the observed maximum (dy/dx = −10.687 at 16 meetings, p = 0.010), reinforcing that more frequent meetings are associated with a progressively larger, not smaller, obfuscation effect across virtually the entire empirical distribution. The moderation effect of the audit committee attributes as discussed above is also illustrated in Figure I, II and III in Appendix.

The article investigated the moderating effect of audit committee characteristics in the relationship between ESG reporting and readability of annual reports, using data from 30 firms listed on the GSE between 2000 and 2023. The sample included both financial and nonfinancial firms; since all listed firms operate under common regulation, the study deemed it appropriate to include both categories. The study finds that although ESG disclosure negatively affects annual report readability, audit committee features can help ensure readable reports. Specifically, the presence of female directors on audit committees, increased meeting frequency and larger committee size tame ESG disclosure's readability-reducing effect. This suggests that while ESG disclosure cannot be ignored in today's business environment, firms need gender-diversified audit committees holding sufficient meetings with adequate directors to produce readable, understandable annual reports.

In terms of policy, the findings imply that regulators, including the Institute of Chartered Accountants (ICAG) Ghana, the Securities and Exchange Commission (SEC) Ghana and the GSE, should strengthen corporate governance codes to ensure well-structured audit committees, if reporting quality and readability are to be achieved. The study recommends that governance frameworks emphasize gender diversity within audit committees, promote adequate committee size, and prescribe and continue to encourage minimum meeting frequencies to enhance monitoring effectiveness and readability. Additionally, ESG reporting frameworks in contexts like Ghana should include guidelines ensuring sustainability disclosures are communicated clearly without impairing readability. Governance policies should also mandate continuous training programs for audit committee members to build capacity in ESG assurance and effective communication of complex information.

In terms of practice, firms should recognize ESG reporting as integral to transparency and legitimacy, while acknowledging that excessive or poorly structured disclosures can impair readability. Audit committees should be empowered to play a stronger oversight role in ensuring ESG/non-financial disclosures are presented clearly and concisely. Committee appointments should include both males and females, since this enriches deliberations and introduces broader perspectives that promote disclosure clarity. Increasing meeting frequency provides opportunities for deeper review and discussion, ensuring reports are prepared with greater clarity. Where possible, boards should maintain larger audit committees, since this allows a broader expertise pool to address complex reporting issues. Finally, firms should adopt internal readability assessments as part of report preparation to ensure reports remain informative and user-friendly.

Overall, the findings support the obfuscation hypothesis, which holds that managers may use complex disclosures like ESG information to obscure unfavourable details or manage stakeholder perceptions, thereby reducing readability. However, since audit committee attributes mitigate this tendency, this underscores governance oversight's relevance in constraining obfuscation. The study's findings also lend support to the resource dependence theorists, who argue that audit committees represent valuable organizational resources providing the expertise, knowledge and external linkages needed to enhance reporting processes and clarity. Finally, the findings are consistent with the position of the agency theorists by demonstrating that well-functioning audit committees, proxied by gender diversity, size and meeting frequency, serve as effective monitoring mechanisms that minimize managerial discretion and ensure ESG information is reported transparently, credibly and understandably. Essentially, these findings show that robust audit committee structures not only enhance governance quality but ensure sustainability reporting contributes to, rather than detracts from, annual report readability.

Despite its contributions, this study has limitations. First, readability is proxied solely by page count and file size, a choice grounded in prior literature and necessitated by earlier GSE annual reports being scanned PDFs that preclude reliable text extraction. However, page count and file size reflect document length and size, not textual or linguistic complexity; two similarly long reports could still differ in sentence structure, vocabulary, clarity or coherence. The readability measure should thus be read as a length-based proxy rather than a comprehensive textual one. Second, the study covers 30 firms from a single stock exchange market over the period 2000–2023. While offering valuable insight into an underexplored African market, this limited scope constrains generalizability to firms operating under different regulatory regimes, capital markets and institutional environments. Ghana's governance dynamics and disclosure practices may not translate to other African or developed markets with more mature ESG frameworks, so conclusions beyond this context should be extended cautiously, pending further validation.

Building on these limitations, several research avenues are proposed. First, cross-country comparative analyses could examine whether audit committee characteristics' moderating role holds consistently across other African markets, emerging economies or combinations of emerging and developed markets helping establish whether these findings reflect Ghana's specific institutional environment or a more generalizable governance dynamic. Second, future research should employ alternative or complementary readability measures beyond page count, including computational textual analysis (Fog Index, Flesch-Kincaid Grade Level or financial-context-specific algorithms) where machine-readable text is available, or NLP approaches capturing linguistic and semantic complexity more directly. Triangulating length-based and text-based measures would help establish whether present findings reflect genuine readability effects versus document length alone.

Third, given that ESG reporting practices, regulatory expectations and audit committee structures are evolving rapidly, longitudinal designs tracking how the ESG-readability relationship and its moderation by audit committee characteristics change over time as ESG reporting matures and standardizes in emerging markets would be valuable. Such studies could examine whether specific ESG standards or regulatory mandates produce measurable shifts in these relationships. Collectively, pursuing these directions would extend and refine the current study's contributions while addressing the geographical and methodological boundaries acknowledged above.

The supplementary material for this article can be found online.