Bridging the gap: the role of assurance quality in reducing sustainability decoupling

In recent decades, global challenges such as climate change, biodiversity loss, environmental degradation, human rights inequalities, excessive water consumption and gender imbalances have intensified social and political scrutiny of corporate activities, prompting companies to reconsider and reshape their business practices (Laine, 2024; Carungu et al., 2025). In response, policymakers and regulators have increasingly promoted international initiatives and policy frameworks, including the Paris Climate Agreement, the European Green Deal and Climate Law and the United Nations (UN) 2030 Agenda. These initiatives have encouraged companies to shift their priorities from short-term profit maximization toward long-term value creation, thereby accelerating the transition toward a more sustainable, resource-efficient and low-carbon economy (United Nations, 2015; O’Dwyer and Unerman, 2020; Carungu et al., 2025).

This transformation has also reshaped accounting practice, with sustainability reporting becoming an increasingly institutionalized dimension of corporate disclosure (Argento et al., 2019; Jackson et al., 2020; Grossi et al., 2021). Sustainability reporting aims to meet the growing information needs of investors and other stakeholders regarding firms’ social and environmental performance (Hahn and Lülfs, 2014; Jackson et al., 2020; Gebhardt et al., 2023). In this context, voluntary reporting frameworks – such as those promoted by the Global Reporting Initiative (GRI) and the International Integrated Reporting Council (IIRC) – have proliferated alongside regulatory developments moving toward mandatory sustainability disclosure, including Directive 2014 / 95/EU (the Non-Financial Reporting Directive, NFRD) and, more recently, Directive 2022 / 2464 (the Corporate Sustainability Reporting Directive, CSRD) (Talbot and Boiral, 2018; Nicolò et al., 2025a; Raimo et al., 2025a).

Despite these developments, a growing body of empirical evidence suggests that sustainability reports are often far from transparent, reliable and decision-useful sources of information for investors and other stakeholders (Michelon et al., 2015; Talbot and Boiral, 2015, 2018). In many cases, companies use sustainability reporting as a reputational tool to project the image of responsible corporate citizens committed to sustainable development (Boiral, 2013; Hahn and Lülfs, 2014; Jiang et al., 2023). This may involve selective, incomplete or biased disclosures that emphasize positive environmental, social and governance (ESG) outcomes while minimizing or concealing negative impacts, as well as narrative strategies aimed at shaping or influencing stakeholders’ perceptions (Cho et al., 2015; Marquis et al., 2016; Talbot and Boiral, 2018). By presenting an idealized representation of their sustainability practices that does not necessarily reflect actual performance, companies seek to maintain legitimacy and mitigate public concerns, particularly when operating in environmentally sensitive industries (Boiral, 2013; Cho et al., 2015; Nicolò et al., 2025b). As a consequence, an increasing gap may emerge between what companies actually do in terms of ESG practices and what they disclose in their sustainability reports – a phenomenon commonly referred to as sustainability decoupling (Sauerwald and Su, 2019; García-Sánchez et al., 2022a; Talpur et al., 2024). Such corporate hypocrisy, arising from the misalignment between corporate assertions and actual performance (Shklar, 1984), has generated growing frustration and skepticism among users of sustainability reports, concerns that recent regulatory initiatives have only partially alleviated (Talbot and Boiral, 2018; Velte, 2023; Nicolò et al., 2025a).

In response, a growing stream of research has examined sustainability decoupling practices from different perspectives. Several studies have explored how companies disclose negative ESG events and controversial issues (e.g. Boiral, 2013; Hahn and Lülfs, 2014; Cho et al., 2015; Talbot and Boiral, 2015; Roszkowska-Menkes et al., 2024), while others have investigated the relationship between ESG performance and disclosure to identify potential symbolic or legitimacy-seeking behaviors (e.g. Hummel and Schlick, 2016; Papoutsi and Sodhi, 2020; Nicolò et al., 2025b). However, considerably less attention has been devoted to identifying the determinants of sustainability decoupling. The limited studies addressing this issue have mainly focused on internal governance mechanisms, such as board attributes (e.g. Sauerwald and Su, 2019; García‐Sánchez et al., 2020; Eliwa et al., 2023; Gull et al., 2023a), the composition of sustainability committees (Gull et al., 2024) and CEO power (e.g. Shahab et al., 2022; Gull et al., 2023b). While these contributions provide valuable insights into the internal governance dynamics that may influence decoupling practices, they offer only a partial and incomplete understanding of the phenomenon. In particular, the potential role of external monitoring and accountability mechanisms in mitigating sustainability decoupling remains largely unexplored. As a result, current knowledge provides limited evidence on whether and how external oversight tools may constrain the gap between what companies disclose and what they actually do. This gap is particularly relevant in light of the increasing regulatory emphasis on transparency, verification and accountability in sustainability reporting. Consequently, a significant research gap persists regarding the influence of external factors on firms’ sustainability decoupling practices.

Among the potential external mechanisms capable of addressing this issue, sustainability assurance has attracted increasing scholarly and regulatory attention. Assurance is an independent service provided by professional accountants, auditors or specialized consultants, aimed at enhancing the credibility and transparency of ESG information disclosed in sustainability reports (Moroney et al., 2012; Ballou et al., 2018; Boiral et al., 2019). In this sense, external assurance may represent an important instrument for mitigating sustainability decoupling and strengthening stakeholder trust and corporate legitimacy (Martínez-Ferrero and García-Sánchez, 2017; García-Sánchez et al., 2022a).

Nevertheless, empirical evidence on the relationship between external assurance and sustainability decoupling remains limited. Sauerwald and Su (2019) found that external assurance reduces corporate social responsibility (CSR) decoupling, whereas Braam et al. (2025) showed that CSR decoupling is significantly mitigated when firms obtain reasonable assurance combined with either a broad assurance scope or an assurance provider belonging to the auditing profession. In contrast, García-Sánchez et al. (2022a) found no significant relationship and highlighted that only specific assurance-provider characteristics – such as being an accounting firm, possessing greater experience and having sectoral specialization – contribute to reducing CSR decoupling. This mixed evidence echoes part of the literature suggesting that assurance may sometimes function merely as a symbolic mechanism aimed at seeking legitimacy rather than substantively improving reporting practices (Talbot and Boiral, 2013, 2018; Michelon et al., 2015). Accordingly, as Hummel et al. (2019) argued, assessing assurance quality – that is, the depth and rigor of the assurance process – is essential to obtain more meaningful evidence on the ability of assurance to improve the credibility of ESG disclosure and mitigate decoupling practices.

Drawing on legitimacy theory, this study addresses this research gap by investigating whether and how assurance quality mitigates ESG decoupling. To this end, the analysis relies on a sample of 717 European companies drawn from the LSEG Data and Analytics database that issued third-party-assured sustainability reports over the period 2014–2023, yielding an unbalanced panel of 6,925 firm-year observations. In line with prior studies (Hawn and Ioannou, 2016; García-Sánchez et al., 2022a), ESG decoupling is measured through a score ranging from −1 to + 1, calculated as the normalized difference between ESG disclosure (20 external items) and ESG performance (18 internal items), both obtained from the LSEG Data and Analytics database. A panel Tobit regression model is then estimated to examine the relationship between assurance quality – captured through five dimensions reflecting the scope, methodology and level of detail of the sustainability assurance process – and ESG decoupling.

This study provides empirical evidence that higher-quality external assurance is associated with lower ESG decoupling. This finding is consistent with the substantive perspective of legitimacy theory, suggesting that external assurance – provided by independent professional accountants, auditors or private consultants – reinforces firms’ pursuit of legitimacy by enhancing the credibility and reliability of ESG information disclosed in sustainability reports. The robustness of these results is confirmed through a battery of tests using alternative ESG decoupling measures, which yield consistent findings.

Building on this evidence, this study contributes to the sustainability decoupling literature by shifting the focus from internal governance mechanisms to external determinants, highlighting assurance quality as a key factor in mitigating the gap between corporate disclosure and performance. It also advances the assurance literature by examining multiple, and still underexplored, dimensions of assurance quality – such as scope, content, level, methodology and applied standards – thereby providing more detailed insights into how assurance practices influence ESG credibility. Finally, by adopting a legitimacy theory perspective, this study extends its application to the domain of sustainability assurance, demonstrating that high-quality assurance can operate as a substantive mechanism that enhances transparency and reduces ESG decoupling.

The remainder of this study is organized as follows. First, the literature review is outlined, followed by the theoretical background and hypotheses development. Next, the research methodology is presented, followed by the illustration and discussion of the main findings. Finally, concluding remarks are offered, along with implications, limitations and avenues for future research.

In recent years, sustainability reporting has gained momentum, primarily in response to growing public and national concerns over social and environmental issues, as well as the resulting changes in stakeholder needs (Tschopp and Nastanski, 2014). As with financial reporting, providing a transparent, fair and accurate representation of firms’ ESG-related activities and impacts has become pivotal to meeting investor demand and bolstering stakeholder confidence in the sustainability of corporate business (Talbot and Boiral, 2018; Schwoy et al., 2025). According to the KPMG, 2022 Global Chief Executive Officer (CEO) Outlook (KPMG, 2022), 69% of CEOs acknowledge growing stakeholder demand for greater transparency and enhanced reporting on ESG issues, while 72% expect stakeholder scrutiny of matters such as climate change and gender equality to intensify further in the coming years. Similarly, a global survey conducted by Ernst and Young (2020) found that 91% of investors reported that ESG performance plays a pivotal role in their asset allocation and selection decisions, and 72% conduct a structured and methodical evaluation of ESG disclosures. Accordingly, the rate of sustainability reporting among the world’s largest 250 companies has reached 96% (KPMG, 2022, 2024).

However, both the gray and academic literatures have underscored that investor and stakeholder appetite remains unmet due to persistent concerns about the quality and reliability of ESG information disclosed by many companies (Ernst and Young, 2020; Schwoy et al., 2025). Despite voluntary (e.g. GRI, IIRC) and mandatory (e.g. NFRD, CSRD) efforts to standardize sustainability reporting to ensure a comprehensive and faithful representation of companies’ ESG performance, skepticism and dissatisfaction remain widespread (Boiral, 2013; Michelon et al., 2015; Pizzi et al., 2024). Accordingly, many scholars have brought to light the concept of decoupling, documenting misinformation in sustainability reports arising from a misalignment between external sustainability efforts (reporting/“talk”) and internal sustainability actions (performance/“walk”) (Walker and Wan, 2012; Sauerwald and Su, 2019; Velte, 2023).

Empirical studies have investigated the concept of decoupling from different perspectives. One strand of research has analyzed corporate sustainability disclosures in depth to discern how companies deal with negative events and poor performance (e.g. Moneva et al., 2006; Boiral, 2013; Talbot and Boiral, 2013, 2015, 2018; Hahn and Lülfs, 2014; Cho et al., 2015; Roszkowska-Menkes et al., 2024). An overall tendency to avoid disclosing adverse events and significant sustainability challenges has emerged, aiming to project a narcissistic, self-laudatory image disconnected from reality (Moneva et al., 2006; Boiral, 2013). In particular, high emitters tend to withhold sensitive data on energy efficiency and financial relationships with government authorities to avoid legitimacy and reputational threats (Talbot and Boiral, 2013). They also use various neutralization techniques to rationalize their environmental impacts (Talbot and Boiral, 2015). Likewise, companies in the energy sector often conceal information on climate performance measurement, carbon balances and emissions, while downplaying their environmental externalities (Talbot and Boiral, 2018). Moreover, Hahn and Lülfs (2014) observed that US and German listed companies use symbolic legitimation strategies in their sustainability reports to alter stakeholders’ perceptions of adverse events, while Cho et al. (2015) evidenced that contradictory societal and institutional pressures prompt US oil and gas companies to engage in hypocrisy and develop façades. Such evidence is echoed by Roszkowska-Menkes et al. (2024), who noted that MSCI-covered companies tend to use various forms of selective disclosure – vagueness, avoidance and hypocrisy – to camouflage negative performance and project a biased impression of sustainability commitment.

Another strand of research has examined decoupling by analyzing the alignment (or misalignment) between different dimensions of corporate sustainability performance and the extent and quality of disclosure (e.g. Cho and Patten, 2007; Cho et al., 2010; Herbohn et al., 2014; Luo, 2019; Papoutsi and Sodhi, 2020; Jiang et al., 2023; Nicolò et al., 2025b; Schwoy et al., 2025). From this standpoint, the evidence remains mixed. Some studies found that the level and quality of disclosure move in step with performance (e.g. Herbohn et al., 2014; Hummel and Schlick, 2016; Papoutsi and Sodhi, 2020; Nicolò et al., 2025b), suggesting that companies achieving strong results have greater incentives to avoid decoupling to enhance their reputation and differentiate themselves from weaker peers. By contrast, other studies reported a negative relationship between environmental performance and disclosure, indicating legitimacy-seeking behaviors based on decoupling (e.g. Cho and Patten, 2007; Luo, 2019; Jiang et al., 2023). In addition, Cho et al. (2015) found that US companies with poorer environmental performance tend to use convoluted and less certain narratives to obscure internal attributions for unfavorable outcomes. Similarly, Schwoy et al. (2025) observed that firms in the pharmaceutical and textile industries often conceal or misreport their ESG controversies.

The last strand of research has analyzed the possible drivers of sustainability decoupling practices, with primary attention devoted to the corporate governance mechanism. Specifically, higher board gender diversity reduces sustainability decoupling practices (Eliwa et al., 2023; Gull et al., 2023a), particularly for firms domiciled in countries with low levels of religiosity (Eliwa et al., 2023) and those with weaker governance quality (Gull et al., 2023a). Moreover, the presence of a CSR committee on the corporate board curbs CSR decoupling, especially when committees are larger and benefit from more independent, longer-tenured members (Gull et al., 2024). By contrast, sustainability decoupling is fostered by boards of directors that permit greater managerial entrenchment – i.e. more discretion in strategic decision-making (García-Sánchez et al., 2020) – and by more powerful (Shahab et al., 2022; Gull et al., 2023b) and overconfident CEOs (Sauerwald and Su, 2019). In addition, Liu et al. (2023) found that firms with distracted mutual fund investors tend to be more inclined to engage in ESG decoupling practices.

Few empirical studies have explored the external drivers of sustainability decoupling. Among these, Marquis and Qian (2014) found that Chinese firms located in regions with more developed government institutions face stronger scrutiny and monitoring pressures, and are therefore less likely to engage in CSR decoupling. Similarly, Tashman et al. (2019) showed that home-country institutional voids increase the CSR decoupling of emerging-market multinationals, whereas the degree of internationalization exerts the opposite effect. In contrast, Zhang (2022) reported that analyst coverage decreases CSR decoupling among listed Chinese firms, particularly non-state-owned enterprises.

Some scholars have focused on assurance practices. Specifically, Sauerwald and Su (2019) found that external assurance reduces CSR decoupling, while Braam et al. (2025) reported that CSR decoupling is significantly reduced when firms obtain reasonable assurance combined with either a broad assurance scope or an assurance provider belonging to the auditing profession. García-Sánchez et al. (2022a) further showed that the application of GRI guidelines and certain assurance-provider characteristics – such as being an accounting firm, having greater experience and possessing sectoral specialization – help curb CSR decoupling.

The academic debate surrounding external assurance and the quality of sustainability reporting has intensified in recent years, particularly following the introduction of the CSRD. The CSRD requires in-scope firms to obtain limited assurance on sustainability reporting, with the initial ambition of moving toward reasonable assurance in due course (Krasodomska et al., 2025; Sabbatucci et al., 2026; Delgado Sánchez et al., 2026). This represents an important step toward bringing sustainability and financial reporting closer together, thereby strengthening stakeholders’ confidence in the credibility, accuracy and reliability of sustainability information (Krasodomska et al., 2024; Jharni and Leoni, 2026). Mandatory assurance has also been framed as a key mechanism to mitigate managerial capture, that is, managers’ strategic control and discretion over both the information disclosed and the quality of the assurance process itself, including the selection of the assurance provider, as well as the scope and level of assurance (Braam et al., 2025; Krasodomska et al., 2025). In particular, the shift from narrower limited assurance (typically expressed in negative form) to broader-scope reasonable assurance (typically expressed in positive form) is pivotal in assuring stakeholders that disclosed sustainability information is free from material misstatements and consistent with companies’ actual performance (Braam et al., 2025; Krasodomska et al., 2025). Nonetheless, following the approval of the first Omnibus package, the envisaged transition toward reasonable assurance was removed and limited assurance was retained, with EU-level limited assurance standards expected to be issued in late 2026 (Krasodomska, 2025; Nicolò et al., 2025a). During this transitional period, EU Member States retain some latitude in allowing the use of national standards and authorizing different categories of practitioners to perform sustainability assurance engagements, ranging from accounting firms to other specialized providers (e.g. engineering firms and consultancy or boutique firms), including the possibility of relying on the incumbent statutory auditor (Krasodomska et al., 2025; Sabbatucci et al., 2026). This institutional flexibility is likely to perpetuate a fragmented and discretionary assurance market, with potential adverse implications for the quality and comparability of sustainability reporting across countries (Krasodomska et al., 2024; Sabbatucci et al., 2026; Delgado Sánchez et al., 2026). As a consequence, in late 2024, the Committee of European Auditing Oversight Bodies (CEAOB) issued Guidelines on limited assurance for sustainability reporting under the CSRD, with explicit reference to the ESRS (CEAOB, 2024). The document provides non-binding, high-level recommendations intended to promote greater consistency in limited assurance practices across EU countries during the interim period before the adoption of EU-level assurance standards.

Against this backdrop, Sabbatucci et al. (2026) assessed the alternative scenarios emerging from the adoption of the CSRD, raising concerns about the potential monopoly of incumbent statutory auditors, which could foster a “financialized” approach to sustainability reporting. At the same time, they highlighted possible shortcomings in terms of comparability and standardization under a competitive scenario involving both statutory auditors and independent assurance service providers. According to (Delgado Sánchez et al., 2026), firms that appoint the same auditor for both financial and ESG assurance may achieve more integrated and credible reporting. Such dual engagement represents a pragmatic approach to mitigating timing misalignments in the assurance process without compromising the quality of reporting outcomes. Krasodomska (2025) examined the advantages and disadvantages of permitting non-accounting firms to provide assurance services in Central and Eastern European (CEE) countries. She highlighted that Big Four accounting firms may enjoy a competitive advantage owing to their extensive operational infrastructures and international networks, which enable them to respond effectively to emerging challenges by leveraging their global connections. Krasodomska et al. (2025) further observed that accountants who perceive themselves as having a stronger understanding of sustainability reporting regulations, and who recognize greater benefits associated with mandatory sustainability reporting, are more likely to support the introduction of mandatory assurance for sustainability reports.

Therefore, the literature review highlights that the academic debates on decoupling and assurance have developed in parallel, assuming even greater relevance in light of the recent regulatory changes introduced by the CSRD. However, scholars have primarily focused on analyzing the alignment (or misalignment) between different dimensions of corporate sustainability performance and the extent and quality of disclosure, or the role of internal factors linked to corporate governance mechanisms in shaping decoupling practices. Consequently, limited attention has been devoted to the impact of assurance quality on firms’ sustainability decoupling practices.

In this regard, scholars have recognized that assurance quality is a multidimensional construct encompassing several interrelated dimensions, including the type of assurance provider as well as the level and scope of assurance (García-Sánchez et al., 2022a; Krasodomska et al., 2024; Braam et al., 2025). Building on legitimacy theory, this study addresses this research gap by examining whether high-quality assurance – in terms of scope, content, level, methodology and applied standards –operates as a substantive mechanism that enhances transparency and credibility in sustainability reporting, or rather as a symbolic device aimed at managing stakeholders’ perceptions.

This study draws upon legitimacy theory to explain the relationship between assurance quality and corporate ESG decoupling (Dowling and Pfeffer, 1975; Ashforth and Gibbs, 1990; Suchman, 1995; Michelon et al., 2015). Legitimacy theory asserts that business organizations, like all social institutions, do not hold an intrinsic right to operate and secure strategic resources from the surrounding environment – such as human capital, financial means, raw materials, infrastructure and customers (Shocker and Sethi, 1973; Yongvanich and Guthrie, 2007; Grossi et al., 2016). This right is conditional upon obtaining legitimacy, understood as a conferred status or condition of acceptance granted by the social community in which the company operates (Dowling and Pfeffer, 1975; Chen and Roberts, 2010). Accordingly, companies are perceived as legitimate when they pursue socially accepted objectives – i.e. when their actions and operations comply with the broader system of norms, expectations and values encapsulated in a social contract virtually signed with society (Chen and Roberts, 2010; L’Abate et al., 2023; Nicolò et al., 2025a). Any disparity – actual or potential – between business conduct and the social contract constitutes a threat to corporate legitimacy, which can hinder a firm’s ability to operate and access strategic resources, thereby jeopardizing its long-term survival (Deegan, 2002; Michelon et al., 2015; Nicolò and Cervilla‐Bellido, 2025).

As Suchman (1995, p. 574) stated, “legitimacy is possessed objectively, yet created subjectively”. Achieving legitimacy, therefore, depends upon legitimation, defined as “the process whereby an organization justifies to a peer or superordinate system its right to exist” (Maurer, 1971, p. 361). In this view, many studies highlight that ESG disclosure serves as a pivotal legitimation instrument through which companies demonstrate how their strategies and actions align with the tenets of the social contract and stakeholder expectations (Deegan and Rankin, 1996; L’Abate et al., 2023; Nicolò et al., 2025a). However, according to Ashforth and Gibbs (1990), the process of legitimation is not uniform and can be interpreted through two main behavioral approaches: substantive and symbolic. The substantive perspective suggests that companies pursue legitimacy by undertaking real, material changes in their strategic objectives, processes and structures, or in socially institutionalized practices (Ashforth and Gibbs, 1990). From this perspective, ESG disclosure testifies how organizations have concretely addressed social and environmental issues across their operations (Ashforth and Gibbs, 1990; García-Sánchez et al., 2022a; Nicolò et al., 2024), thereby reflecting “the outcome of a sense of accountability to stakeholders, driven by a genuine interest in enhancing transparency” (Michelon et al., 2015, p. 62). External assurance provided by independent professional accountants, auditors or private consultants can reinforce firms’ substantive pursuit of legitimacy, as it serves as a pivotal instrument to enhance the credibility and reliability of ESG information disclosed in sustainability reports – analogous to the role of external auditing in financial reporting (Moroney et al., 2012; Ballou et al., 2018). It can be viewed as complementing a company’s strategic approach to sustainability, offering clearer and more credible signals of its actual ESG performance and practices (Krasodomska et al., 2024). Hence, third-party verification is assumed to exert a disciplinary effect on companies, encouraging improved sustainability performance and disclosure practices, thereby reducing uncertainty and skepticism among investors and other stakeholders (Moroney et al., 2012; Boiral et al., 2019).

From this perspective, scholars have provided empirical evidence showing that external assurance enhances the quality of sustainability reports (e.g. Moroney et al., 2012; Sethi et al., 2017; Ballou et al., 2018; Manes-Rossi et al., 2021). Companies aiming to build trust through sustainability-related performance are more prone to obtain higher quality and larger in scope external assurance for their sustainability reports, primarily in larger companies operating in environmentally sensitive sectors (Krasodomska et al., 2024). It has also been observed that assurance helps companies signal superior sustainability performance (Karaman et al., 2021) and improve their internal information and reporting systems (Alazzani and Wan-Hussin, 2013). Assuring ESG information further enhances professional investors’ perceptions of sustainability performance and increases their willingness to invest (Reimsbach et al., 2018). Moreover, specific characteristics of the assurance provider, such as experience and specialization, have been found to reduce CSR decoupling practices (García‐Sánchez et al., 2022b). For these reasons, and in line with the substantive perspective of legitimacy theory, it is expected that higher assurance quality – i.e. greater depth of the assurance process – reduces ESG decoupling. Accordingly, the following hypothesis is posited:

Conversely, the symbolic approach to legitimacy seeks to build and preserve organizational legitimacy by portraying the organization as socially responsible and aligned with prevailing societal norms and expectations (Nicolò et al., 2024; Awuah et al., 2026). Within this perspective, companies may attempt to shape and influence stakeholders’ perceptions by constructing a “new face to the outside world while protecting the inner workings of the organization from external view” (Hopwood, 2009, p. 437). Rather than implementing substantive changes to internal processes and activities, organizations may instead rely on communicative tools that allow them to project a favorable image, even in the presence of weak or negative performance outcomes (Michelon et al., 2015; Lodhia et al., 2023; Nicolò et al., 2024).

From this perspective, scholars have questioned the substantive effectiveness of sustainability assurance. Cho et al. (2014) found that CSR assurance does not enhance firms’ market value but is associated with greater inclusion in sustainability rankings and a stronger “green” reputation. This evidence suggests that assurance may primarily strengthen corporate image rather than generate tangible financial benefits, supporting its interpretation as a symbolic practice. Similarly, Michelon et al. (2015) reported no significant relationships between external assurance and the quality of sustainability disclosure, interpreting this finding as evidence that assurance may be used mainly as a symbolic legitimation tool rather than as a mechanism for strengthening accountability. Other scholars have also raised concerns regarding the effectiveness and credibility of sustainability assurance practices. Boiral (2013) highlighted limited transparency and recurring non-conformities in GRI-based reports issued by firms operating in the energy and mining sectors. These concerns were reinforced by Talbot and Boiral (2013, 2018), who described sustainability assurance as resembling a “rational myth” and emphasized its limited reliability given the widespread inconsistencies identified in sustainability reports within the energy sectors. In addition, Manetti and Toccafondi (2012) showed that the reliability and quality of the external assurance may be compromised by strong managerial influence over the assurance process and by limited stakeholder involvement (Manetti and Toccafondi, 2012).

For these reasons, and in line with the symbolic perspective of legitimacy theory, it is expected that assurance quality – i.e. greater depth of the assurance process – does not materially influence ESG decoupling. Accordingly, the following hypothesis is posited:

Consistent with prior literature, the primary source of ESG data for this study is the LSEG Data and Analytics database (formerly Refinitiv/Thomson Reuters Asset4). Covering more than 70% of global market capitalization, LSEG has become a standard academic data source due to its reliability and the standardized nature of its metrics, which ensure comparability across European jurisdictions (Kim and Park, 2023).

Nevertheless, the measurement validity of ESG ratings remains a subject of ongoing academic debate. Recent studies highlight a lack of convergence among major ESG rating providers – such as LSEG, MSCI and Sustainalytics – with scores often exhibiting relatively low correlations (Chatterji et al., 2016; Dimson et al., 2020; Larcker et al., 2021; Berg et al., 2022; Capizzi et al., 2021). These discrepancies are commonly attributed to differences in scope, weighting schemes and the “black box” nature of certain rating methodologies. Despite these limitations, LSEG remains widely used in empirical research due to its relatively transparent methodology and its reliance on objective and standardized data points. Moreover, prior evidence suggests that results based on LSEG data are consistent with economic fundamentals and remain robust when compared with alternative measures such as MSCI (Demers et al., 2021). The database has also been extensively used in recent studies examining sustainability decoupling (Gull et al., 2023a) and sustainability assurance (García-Sánchez et al., 2025a). Therefore, while the ESG scores are interpreted with appropriate caution, LSEG represents a reliable and widely accepted proxy for the purposes of this study.

However, standard databases often lack detailed information regarding the specific characteristics of sustainability assurance engagements. To ensure a more accurate measurement of assurance quality, the automated data were complemented with a manual content analysis. In particular, information regarding the scope, level, methodology and standards of assurance was manually extracted from the assurance reports available on the websites of the selected companies. This procedure allows for a more precise characterization of assurance practices than what is typically available in commercial data sets.

The initial target population was drawn from companies subject to the European NFRD and CSRD. These regulations require companies located in Europe to disclose sustainability information if they meet at least two of the following three criteria over two consecutive fiscal years:

1. net turnover exceeding €50 million;

2. total assets exceeding €25 million; and

3. a workforce of more than 250 employees.

These thresholds are consistent with the definition of “large undertakings” effective from 1 January 2024 under Commission Delegated Directive (EU) 2023 / 2775, issued October 17, 2023. Applying these criteria resulted in an initial sample of 1,731 eligible companies. Among these, 1,280 firms reported that their sustainability information had been externally verified at least once during the period 2014–2023.

Subsequently, two additional data filters were applied:

1. the availability of information for all variables included in the empirical model (Equation 1), and

2. the availability of data for at least five consecutive years.

The latter requirement is essential for panel data analyses to properly control for unobserved heterogeneity. After applying these criteria, the final sample consists of 717 European companies, forming an unbalanced panel of 6,925 firm-year observations. Figure 1 illustrates the geographical (Panel A) and sectoral (Panel B) distribution of the sample.

To empirically test the proposed hypotheses, a regression model was specified to examine the relationship between assurance quality and ESG decoupling. The estimation of equation (1) allows for the assessment of whether assurance quality reduces or fails to reduce ESG decoupling, thereby providing evidence in support of either the substantive or the symbolic approach to legitimacy:

where i refers to the European firm (ranging from 1 to 717) and t denotes the fiscal year for the period 2014–2023. The error term includes the firm-specific effect (η_i), capturing unobservable heterogeneity and the idiosyncratic error term (ε_it).

Given the censored nature of the dependen7t variable, equation (1) was estimated using a Tobit regression model for panel data, with a one-year lag applied to the explanatory variables to mitigate concerns about endogeneity. The vector of control variables includes Sector, Country and Year, which account for sectoral, geographic and temporal effects, respectively.

Following García-Sánchez et al. (2022a), the dependent variable used to measure ESG decoupling – i.e. the gap between a company’s “walk” and “talk” – is a variant of the approach proposed by Hawn and Ioannou (2016). This measure excludes the initial items related to the explanatory variables that define the empirical model specified in equation (1).

The ESG Decoupling score ranges from −1 to + 1 and is calculated as the normalized difference between ESG disclosure (20 external items) and ESG performance (18 internal items). Values approaching + 1 indicate ESG decoupling consistent with greenwashing, where companies communicate more than they actually do – that is, the sum of ESG disclosure items exceeds that of ESG performance items. Conversely, values approaching −1 reflect greenhushing, where companies disclose less than they actually do, underreporting positive performance.

The validity of using LSEG data to capture the “walk-talk” gap relies on the detailed distinction within the database between management-oriented and outcome-oriented indicators. Following the theoretical framework of Hawn and Ioannou (2016), LSEG data points are classified into two distinct categories: symbolic disclosure and substantive performance. As shown in  Appendix 1, disclosure (“talk”) is proxied by reporting indicators, which capture the information published in corporate reports, whereas performance (“walk”) is proxied by operational metrics and quantitative outcomes reflecting actual implementation. This distinction enables the derivation of a decoupling measure, consistent with prior high-impact studies (Gull et al., 2023; Sánchez et al., 2026). The construction of this metric involves three steps. First, aggregation, whereby the mean score is computed across all constituent items for the disclosure and performance dimensions, respectively. Second, normalization, whereby each dimension is rescaled to ensure comparability. Third, calculation, whereby the final gap metric is obtained by subtracting the lagged performance score from the current disclosure score.

Figure 2 illustrates the evolution of this gap over time for the companies in the sample. On average, European firms do not exhibit significant ESG decoupling, as the mean values are negative and close to zero, ranging from −0.07 to −0.12.

Following the approach proposed by Hummel et al. (2019) and García-Sánchez (2020), the AssuQuality variable represents a score ranging from 0 to 9, obtained by summing five items that capture the scope, method and level of detail of the sustainability assurance process. Table 1 presents the components of the score, along with their descriptive statistics. A team of six researchers conducted data collection and coding. To ensure consistency, a standardized coding framework was developed and implemented through a structured Excel interface. Following a pilot calibration phase involving all researchers, a double-coding protocol was used, whereby two researchers independently analyzed each assurance report using content analysis. Any discrepancies in the initial assessment were resolved through consensus discussions. Finally, the coded items were automatically aggregated using spreadsheet functions to generate the final data set.

Figure 3 presents the temporal evolution of the AssuQuality score, showing a slightly positive trend over the analyzed period.

To avoid biased results, and following prior studies (e.g. Clarkson et al., 2019; Gull et al., 2023a, 2023b; García-Sánchez et al., 2025a), the control vector includes 14 variables related to corporate capabilities and resources, board effectiveness and institutional differences based on country of origin and time period. Table 2 presents their conceptualization and descriptive statistics.

Corporate resources and capabilities are relevant because they enable firms to develop strategies, identify and manage stakeholder interests and implement transparent and responsible disclosure policies. These factors allow companies to differentiate themselves from competitors and create long-term value (Bothello et al., 2023; Xu et al., 2025). With respect to board effectiveness, the board not only oversees corporate strategy and safeguards the interests of all stakeholders but also makes key strategic decisions and acts as a guarantor of transparency and regulatory compliance (Raimo et al., 2025b). At the European level, institutional pressures on sustainability have been central to the development of regulatory frameworks that mandate the disclosure and external assurance of ESG information (García-Sánchez et al., 2022b). A notable example is the adoption of the CSRD, which establishes detailed reporting standards and introduces mandatory assurance requirements.

Table 2 reports the mean and standard deviation of the numerical variables, as well as the relative frequencies of the dichotomous ones. On average, the ESG Decoupling score is −0.09, a value close to zero, suggesting that reporting levels are generally aligned with firms’ actual ESG performance. However, the average level of external assurance quality is 5.65 out of a maximum of nine points, indicating a moderate-to-high degree of assurance quality across the sample.

Table 3 presents the correlation matrix and indicates that no multicollinearity issues are detected that could affect the validity of the regression model used to test the research hypotheses. In this context, the reported coefficients for binary-continuous variable pairs correspond to point-biserial correlations, which are mathematically equivalent to Pearson correlations and widely accepted in the econometric literature for diagnosing linear dependence (Bonett, 2020). In addition, the relationships involving dummy variables were further examined using t-tests for binary-continuous pairs and chi-square tests for binary-binary pairs. The results are consistent with the direction and statistical significance indicated by the correlation matrix. Detailed results are reported in  Appendix 2.

Prior to estimating the regression model, the censoring of the ESG Decoupling variable was examined. Specifically, 8% of observations are left-censored (at −1) and 3% are right-censored (at + 1). Given that a non-negligible proportion of observations is censored at both bounds, the use of a Tobit specification is appropriate to account for the limited nature of the dependent variable and to avoid bias associated with OLS estimation (Amemiya, 1984).

Column 1 of Table 4 presents the results obtained from estimating equation (1) using a panel Tobit regression with random effects. The findings show that, at the 99% confidence level, AssuQuality has a negative and statistically significant effect on ESG Decoupling. Specifically, β₁ = −0.00731 < 0, supporting the acceptance of H1. These results indicate that higher quality assurance – namely, an assurance process characterized by a broader scope, more comprehensive content, higher assurance level, a multi-method approach and the use of recognized standards – reduces the decoupling of ESG information reported by companies. Although the coefficient (−0.00731) may appear modest in absolute terms, its economic significance is substantial when considered relative to the mean of the dependent variable (−0.09). A one-unit increase in AssuQuality implies a reduction in decoupling equivalent to approximately 8.1% of its mean value. This suggests that assurance quality operates as an effective governance lever, whereby incremental improvements in verification translate into meaningful reductions in ESG decoupling. Furthermore, the use of panel data estimation controls for unobserved heterogeneity. In this context, coefficients tend to be more conservative than in cross-sectional analyses, as they capture within-firm variation. Therefore, the statistical significance of the coefficient at the 1% level further supports the robustness of the relationship.

From a theoretical standpoint, this finding aligns with the substantive perspective of legitimacy theory, which posits that organizations undertake concrete and material changes in their processes and practices to comply with societal norms, values and expectations, thereby safeguarding their license to operate. Within this framework, external assurance represents a key mechanism through which firms enhance the credibility and reliability of ESG information disclosed in sustainability reports. However, unlike financial reporting – characterized by mandatory disclosure and well-established auditing standards – sustainability reporting has historically been voluntary and fragmented, with limited standardization in both reporting and assurance practices. In addition, sustainability reporting combines quantitative indicators with qualitative information, including stakeholder engagement, materiality assessments and ESG issue mapping, which are inherently difficult to standardize. Moreover, both accounting and non-accounting firms may provide assurance services, contributing to a fragmented market and limiting comparability. This complexity is further reinforced by the coexistence of multiple assurance standards (e.g. ISAE 3000 and AA1000), each covering different aspects of sustainability reporting. Consequently, firms seeking external assurance must make several strategic choices regarding the assurance provider, as well as the scope, content, level and methodology of the engagement – key dimensions that define overall assurance quality (Krasodomska et al., 2024, 2025).

In this context, assurance quality reflects a deliberate strategic choice aimed at strengthening corporate credibility and enhancing stakeholder confidence in ESG disclosure. High-quality assurance, therefore, operates as a substantive legitimacy mechanism that effectively bridges the gap between what firms disclose and what they actually do, thereby reducing ESG decoupling.

This evidence complements prior findings. Sauerwald and Su (2019) showed that external assurance reduces CSR decoupling, whereas García-Sánchez et al. (2022a) found that merely adopting assurance does not mitigate CSR decoupling, highlighting the importance of assurance-provider characteristics. Similarly, Braam et al. (2025) demonstrated that CSR decoupling decreases when firms obtain reasonable assurance combined with either a broad scope or an assurance provider from the auditing profession. However, the measure of assurance scope adopted in prior studies is often limited to a subset of assurance characteristics. By contrast, the assurance quality construct used in this study incorporates multiple dimensions – scope, content, level, methodology and applied standards – thereby providing a more comprehensive representation of the assurance process and offering deeper insights into how external assurance enhances corporate transparency.

These findings stand in contrast to the symbolic legitimacy perspective, which views external assurance as a window-dressing mechanism aimed primarily at shaping stakeholders’ perceptions rather than improving the reliability of ESG disclosures. In this respect, the results challenge prior evidence (e.g. Boiral, 2013; Michelon et al., 2015; Talbot and Boiral, 2018), which emphasized the limitations and potential “dark side” of the assurance process.

To ensure the robustness of these results, Columns 2, 3 and 4 of Table 4 present estimations of equation (1) using alternative proxies for decoupling practice, following García-Sánchez et al. (2022a) and Gull et al. (2023a). Column 2 reports results based on the absolute value of ESG Decoupling, treating greenwashing and greenhushing symmetrically. The coefficient of AssuQuality (β1 = −0.007279 < 0) remains negative and significant at the 95% confidence level, confirming that assurance quality reduces both forms of decoupling. Columns 3 and 4, which isolate underperformance (greenwashing) and underdisclosure (greenhushing), respectively, yield consistent results.

A comparison of these coefficients reveals an asymmetry in the effect of assurance quality. The coefficient for Underdisclosure (−0.00546) is more than double that of Underperformance (−0.00208) and exhibits stronger statistical significance. This finding provides important insights into the mechanisms of assurance, suggesting that high-quality assurance is more effective in enforcing disclosure completeness (reducing greenhushing) than in correcting discrepancies between actual and reported performance (reducing greenwashing). This asymmetry can be explained in terms of auditability. Omissions of material information represent objective and verifiable issues that can be more easily identified and corrected by assurance providers. In contrast, underperformance often involves narrative framing and qualitative disclosure, which are inherently more difficult to challenge. As a result, assurance quality primarily enhances transparency rather than directly correcting underlying performance-related discrepancies.

Regarding control variables, the findings indicate that larger firms – characterized by greater resources, capabilities and visibility – exhibit lower levels of ESG decoupling. Similar patterns emerge for firms with boards featuring a higher proportion of independent directors, female members and dedicated sustainability committees. These results are consistent with previous studies (Gull et al., 2023a, 2024; Pisano et al., 2025). Finally, the results show that the introduction of the CSRD, together with post-2020 disruptions related to the pandemic and geopolitical tensions, has contributed to improving ESG disclosure practices and reducing decoupling. These institutional and contextual factors appear to mitigate information asymmetries and discourage decoupling or greenwashing, ultimately helping to alleviate the financial impact of such shocks on firms (García-Sánchez et al., 2025b).

Given the censored nature of the dependent variable, a Tobit model for panel data is used. While fixed effects estimation is typically preferred, non-linear models such as Tobit suffer from the incidental parameters problem, leading to inconsistent estimates when T is relatively small (Neyman and Scott, 1948; Lancaster, 2000).

To address this issue, the Correlated Random Effects (CRE) approach proposed by Mundlak (1978) is adopted. This method incorporates the time averages of time-varying regressors into the model, thereby relaxing the orthogonality assumption of standard random effects.

A Wald test on the joint significance of the Mundlak terms (⁠$χ2 =129.77, p < 0.01$⁠) rejects the null hypothesis of no correlation between individual effects and regressors. Consequently, the Mundlak terms are retained, allowing the coefficients to approximate within-firm effects. The results, reported in Table 5, confirm the robustness of the main findings.

In addition, to further assess the robustness of the findings and rule out alternative explanations, Table 6 presents a set of robustness checks estimated using fixed effects (Column 1), random effects (Column 2) and the generalized method of moments (GMM) (Column 3). These approaches serve complementary purposes: fixed effects control for time-invariant heterogeneity, while GMM addresses potential endogeneity and reverse causality. The estimates across all three specifications are consistent with the baseline results in terms of sign and statistical significance. This consistency confirms that the negative effect of assurance quality on ESG decoupling is not driven by model specification or endogeneity concerns, but instead reflects a robust structural relationship whereby higher assurance quality effectively mitigates decoupling practices.

Figure 4 illustrates the inverse relationship between assurance quality and ESG decoupling. Panel A shows that, as assurance quality increases, ESG Decoupling converges toward 0. Panel B confirms that the absolute decoupling gap decreases as assurance quality improves.

Following García-Sánchez et al. (2025a), endogeneity represents a critical concern in assurance research, as the decision to obtain high-quality assurance is not random but reflects an endogenous managerial choice. In particular, according to signaling theory, firms with superior ESG performance – and consequently lower levels of decoupling – are incentivized to signal their quality to the market. As a result, they are more likely to voluntarily invest in high-quality assurance to validate their disclosures. Consequently, if assurance is treated as an exogenous variable, the coefficient associated with AssuQuality may be biased upward, as it captures both the effect of assurance and the underlying (unobserved) commitment to sustainability that drives the selection decision.

To address this concern and strengthen the robustness of the results, an identification strategy based on an instrumental variable (IV) approach is adopted. Specifically, an external instrument (Z) is used, which is correlated with the likelihood of selecting high-quality assurance (X), but does not directly affect ESG decoupling (Y), except through assurance.

Following prior literature in accounting and corporate finance (e.g. Cao et al., 2012; Liu et al., 2023; Seo, 2021), PeerPressure is used as an instrument for the assurance decision. This variable is defined as the average assurance level of peer firms within the same industry and geographic region in year t, excluding the focal firm. The validity of this instrument rests on two key assumptions. First, from a theoretical perspective, drawing on institutional theory (DiMaggio and Powell, 1983), firms are subject to normative and mimetic pressures that encourage the adoption of legitimation practices observed among competitors. Accordingly, when a large proportion of peer firms adopts high-quality assurance, the focal firm is incentivized to imitate this behavior to avoid reputational disadvantages, implying a strong correlation between the instrument and the endogenous variable. Second, from a methodological perspective, the instrument satisfies the exclusion restriction. While the assurance choices of peer firms may influence the focal firm’s decision to adopt assurance, there is no plausible direct channel through which peers’ assurance decisions affect the focal firm’s ESG decoupling, beyond common industry-level shocks, which are controlled for through industry and year fixed effects.

The IV approach is implemented using a two-stage least squares estimator. In the first stage, the endogenous variable AssuQuality is regressed on the instrument (PeerPressure) and the full set of control variables, to isolate the exogenous component of assurance quality. In the second stage, ESG Decoupling is regressed on the predicted values of AssuQuality obtained from the first stage.

Table 7 reports the results of the IV estimation. Column 1 shows that PeerPressure has a positive and statistically significant effect on AssuQuality. The strength of the instrument is confirmed by a Kleibergen–Paap rk F-statistic of 19.397, which exceeds the Stock–Yogo critical value of 16.38, indicating that the instrument is not weak. Column 2 shows that the coefficient associated with the instrumented AssuQuality remains negative and statistically significant, thereby confirming the robustness of the main findings and mitigating concerns related to endogeneity.

This study aimed to investigate whether and how assurance quality mitigates the misalignment between corporate sustainability disclosure and performance, commonly referred to as ESG decoupling. Drawing on the dual perspective of substantive and symbolic legitimacy theory, the analysis examined whether assurance quality reduces the gap between corporate ESG “walk” and “talk.” The main findings of this study revealed that assurance quality has a negative and statistically significant effect on ESG decoupling, confirming that a higher-quality assurance process – characterized by broader scope, more comprehensive content, a higher level of assurance, a multi-method approach and the use of recognized standards – reduces the gap between what companies disclose and what they actually do. This evidence supports the substantive perspective of legitimacy theory, indicating that assurance quality effectively reduces ESG decoupling. The results remained robust across alternative model specifications and decoupling measures.

These findings generate several theoretical and practical implications. From a theoretical standpoint, this study contributes to the growing body of literature on sustainability decoupling, which has predominantly focused on internal governance mechanisms while paying comparatively limited attention to external determinants that may mitigate such practices. Mainstream accounting research has extensively examined external assurance as a mechanism to enhance stakeholder confidence in the credibility and reliability of ESG information. However, empirical evidence remains inconclusive. One stream of research suggests that external assurance improves the quality of sustainability reporting, thereby reinforcing stakeholder trust. Conversely, another stream adopts a more critical perspective, finding no consistent positive association between assurance and reporting quality and raising concerns about managerial capture and structural limitations of the assurance process (e.g. Talbot and Boiral, 2013, 2018; Michelon et al., 2015). In this vein, sustainability assurance has been described as a “rational myth,” primarily serving symbolic purposes rather than strengthening accountability (Talbot and Boiral, 2013).

Against this backdrop, by examining assurance quality as an external accountability mechanism, this study addresses a relevant gap in the literature and advances understanding of how independent verification processes shape the alignment between corporate discourse and actual practices. In particular, it focuses on dimensions of assurance quality – such as scope, content, level, methodology and applied standards – that have received limited attention in prior research, thereby enriching the debate on how assurance practices influence sustainability credibility. Moreover, by adopting a legitimacy theory lens, this study extends its application to the domain of sustainability assurance, demonstrating that high-quality assurance can operate as a substantive mechanism for enhancing transparency and mitigating ESG decoupling. The findings suggest that adopting higher levels of assurance quality is associated with greater ESG transparency, reflected in a reduced gap between corporate “talk” and “walk,” thereby supporting a substantive rather than symbolic pathway to legitimacy.

From a practical perspective, the findings provide relevant insights for policymakers, regulators, assurance providers and companies. In the context of the CSRD, although the initial ambition was to introduce limited assurance as a step toward reasonable assurance, recent regulatory developments have postponed this transition, maintaining limited assurance while delaying the adoption of common EU-level standards. During this interim phase, the sustainability assurance market remains fragmented, as EU Member States retain discretion in applying national standards and authorizing different categories of practitioners. In this context, the findings highlight the importance of establishing clearer and more robust quality requirements for sustainability assurance. Regulatory guidance should ensure that assurance engagements comprehensively cover sustainability reports, include all material ESG indicators, clearly define methodological approaches and specify the level of assurance to be achieved. Without such safeguards, limited assurance risks becoming a formal compliance exercise rather than an effective accountability mechanism. Accordingly, forthcoming EU-level standards should preserve a high degree of rigor, consistency and comparability to prevent symbolic adoption of assurance practices and effectively mitigate ESG decoupling. These implications are further reinforced by recent developments such as the introduction of the International Standard on Sustainability Assurance (ISSA) 5000 by the International Auditing and Assurance Standards Board (IAASB). Pending the finalization of EU-level standards, such frameworks – together with guidance issued by oversight bodies – may serve as interim benchmarks to enhance consistency and credibility in assurance practices. Policymakers may also consider targeted incentives to promote high-quality assurance, particularly in high-impact sectors, thereby strengthening ESG transparency in a still-evolving regulatory landscape. For assurance providers, the findings underscore the importance of methodological rigor, professional independence and sector-specific expertise. The adoption of multiple assurance techniques – such as documentation checks, interviews, data analyses, on-site visits and media searches – can enhance the depth and credibility of the assurance process, thereby strengthening stakeholder confidence. For companies, investing in high-quality assurance should be viewed not merely as a compliance requirement but as a strategic governance tool that enhances transparency, reduces ESG decoupling and reinforces organizational legitimacy.

Despite these contributions, some limitations should be acknowledged. The sample is limited to European-listed companies, which may constrain the generalizability of the findings to other institutional contexts. Future research could extend this analysis by adopting cross-country comparative designs to examine how differences in regulatory frameworks and institutional environments influence the effectiveness of assurance quality in mitigating ESG decoupling. In addition, further studies could use quasi-experimental approaches – such as difference-in-differences or regression discontinuity designs – leveraging regulatory changes to provide stronger causal evidence on the impact of assurance quality on ESG decoupling. The use of natural experiments or exogenous shocks affecting the assurance market may also help isolate causal mechanisms more effectively. Experimental designs could offer complementary insights by testing how different assurance attributes (e.g. scope, level, provider type or standards adopted) influence stakeholders’ perceptions of credibility and trust in ESG disclosure. Also, future research could investigate the joint role of external assurance and internal governance mechanisms – such as board composition, sustainability committees and managerial incentives – by using configurational approaches (e.g. fsQCA) or interaction models, thereby providing a more comprehensive understanding of how different governance mechanisms interact in shaping ESG decoupling practices. Finally, this study relies on the LSEG Data and Analytics database (formerly Refinitiv/Thomson Reuters Asset4) to collect ESG data. Although this database is widely adopted in prior research and provides comprehensive and standardised ESG metrics, alternative ESG data providers exist, which may differ in terms of methodological approaches, coverage and rating construction.

Accordingly, future research could extend this analysis by triangulating the findings using ESG data sourced from other major providers, such as Bloomberg or MSCI, to assess the robustness and comparability of results across different databases.