Abstract
In response to demand from investors and other stakeholders, companies have increased voluntary disclosure of climate change-related policies and performance. Information intermediaries have correspondingly emerged to provide needed credibility and commensurability of climate disclosures. However, the provision of performance ratings and lax audit capabilities creates opportunities for firms to manipulate those ratings for impression management. This article explains how firms may attain an intermediary’s favorable assessment of climate performance using varied methods of strategic disclosure. Using data from a prominent climate intermediary (CDP, formerly the Carbon Disclosure Project), I find that strategic disclosure is widespread and effective in attaining higher ratings, particularly through disclosure of forward-looking factors, suggesting that ratings may not be an accurate indicator of a firm’s true underlying performance. This article contributes to commensuration and impression management literatures, offers guidance for mechanism design to improve the reliability of intermediated voluntary disclosure, and highlights important areas for future research.
Keywords
Investors increasingly view climate change and a firm’s associated risks and ongoing management as material to long-term financial performance (Flammer et al., 2021; Khan et al., 2016), while a broad range of stakeholder groups continue to press firms to address climate-related impacts (Durand et al., 2019; Wright & Nyberg, 2017). This heightened societal concern over corporate management of climate change risks has arguably outpaced the response of regulatory institutions in developing systematic mechanisms for reporting and verifying corporate claims (Hahn et al., 2015). Firms have markedly increased proprietary disclosure (e.g., standalone corporate sustainability reports) but lack of consistent standards and unlimited firm discretion in disclosure yields this information of little use in assessing actual quality of firm response to climate change (Boiral, 2013; Talbot & Boiral, 2018). Generalized disclosure standards (Etzion & Ferraro, 2010) and environmental, social, and governance (ESG) rating schemes (Berg et al., 2020; Cort & Esty, 2020) address broad sustainability concerns but often lack specific focus on the pressing issue of climate change. Coincident with heightened investor concern over corporate climate risk, a new class of climate disclosure intermediary has emerged that both solicits a fixed format disclosure from a broad population of global firms (like a sustainability disclosure standard) and processes these data into commensurable climate performance ratings—where climate performance comprises varied aspects corresponding to a firm’s overall policies and achievements in addressing the climate crisis—using a transparent methodology (like an ESG rating scheme). By dictating a fixed disclosure format, the climate intermediary limits firm discretion and heightens focus on issues of material concern (Grewal et al., 2021). Climate disclosure is complex, and the intermediary adds value through provision of summary performance ratings to help stakeholders make sense of underlying information.
However, the assignment of ratings and relatively lax audit and enforcement standards may expose the intermediary to strategic manipulation by firms making misleading disclosures (Callery & Perkins, 2021). Despite the relative lack of strong accountability mechanisms, research indicates that investors and other stakeholders tend to place value in intermediaries’ evaluative assessments (Busch & Hoffmann, 2011; Cho et al., 2012; Kim & Lyon, 2011; Lyon & Shimshack, 2015), such that firms may substitute such assessment for lack of capacity to act (Slager et al., 2021). I posit that firms seek to leverage the trust of stakeholders in the intermediary’s evaluative judgment to manage external perceptions of unobservable performance. Whereas external rating schemes consolidate and evaluate public information about a firm’s activities (Doh et al., 2010), the climate intermediary directly solicits firm disclosure under a standard format with a transparent rating methodology, potentially allowing firms to craft desired impressions through strategic disclosure. Research has documented firm efforts to relax negative perceptions through strategic disclosure (Fabrizio & Kim, 2019; Hahn & Lülfs, 2014), yet we understand relatively little about proactive firm efforts to strategically manipulate intermediaries to attain higher ratings.
This article applies and extends theory on commensuration and impression management to examine multiple distinct methods that firms may use to optimize intermediary ratings without substantively addressing underlying performance. In line with extant research on the lack of credibility of proprietary firm disclosure (Talbot & Boiral, 2018), the article considers firm efforts to foster a positive impression of its underlying quality—defined as externally unobservable performance over a broad range of factors contributing to reduced current and future climate impacts and associated risks. Firms seek to promote a positive impression using voluntary disclosure, but external audiences may doubt firms’ intentions (Lyon & Maxwell, 2011). To mitigate this information problem, firms may seek the favorable assessment of a credible third party to influence stakeholder impressions (Highhouse et al., 2009). Whereas stakeholders may view a firm’s direct disclosure with skepticism, a trusted intermediary’s assessment of disclosure quality (i.e., its performance rating) provides perceived credibility to the rigor of the firm’s disclosure and allows commensuration of perceived performance among multiple firms (Rindova et al., 2018). Investors and other stakeholders may attend to detailed disclosure data according to nonfinancial performance dimensions of primary interest, yet in some cases they may defer to more simplified representations of firm performance (Arjaliès & Bansal, 2018). A sophisticated firm’s understanding of the factors driving the intermediary’s evaluation methodology may allow it to strategically construct its disclosure to obtain a more favorable rating, furthering impression management.
Drawing on prior research across multiple disciplines, I identify and examine four distinct methods of strategic disclosure and a set of disclosure factors aligned with dimensions of stakeholder interest that support intermediary assessment and vary in temporal orientation, where forward-looking performance assessments are arguably more difficult to objectively verify and thus may be associated with higher susceptibility to misleading intent. I hypothesize a series of relationships between these factors and intermediary assessment and test these hypotheses using data drawn from several years of detailed firm disclosures to a prominent climate intermediary: CDP (formerly the Carbon Disclosure Project). Findings show that these methods of strategic disclosure are prevalent and moderately successful at improving intermediary assessment of a firm’s disclosure quality. Moreover, I find that some of the supporting factors intended as safeguards to enhance quality and accountability of disclosure may be ineffective at deterring strategic disclosure.
This article makes important contributions to theory, research, and practice. First, it contributes to literatures on commensuration and organizational impression management by articulating theory and providing supporting empirical evidence demonstrating the varied linkages between multiple methods of strategic disclosure employed by firms and other disclosure factors supporting intermediary assessment with corresponding ratings. Building on prior organizational impression management research that has largely focused on the use of individual tactics (Bolino et al., 2008) and on various approaches for legitimizing negative information through qualitative text in sustainability disclosure (Bansal & Clelland, 2004; Fabrizio & Kim, 2019; Hahn & Lülfs, 2014), I examine both qualitative and quantitative methods of strategic disclosure targeting multiple dimensions (i.e., verification of prior performance versus expected future performance) solicited by intermediaries for commensuration of corporate sustainability performance, an area of increasing importance to regulators, investors, and other stakeholders (Azar et al., 2021; Krueger et al., 2020). Specifically, findings indicate multiple forward-looking aspects of sustainability disclosure may be more susceptible to impression management over multiple methods of strategic disclosure, illuminating the variety of tactics employed by firms to influence ratings.
Second, this article contributes to empirical research on climate disclosure, and ESG reporting more broadly, by constructing a detailed coding of a prominent intermediated disclosure standard and its rating methodology over several years and developing a corresponding set of measures that capture the multi-faceted nature of strategic disclosure and the various factors that support intermediary assessment or ratings. While this set of measures covers a broad range of impression management tactics, it is not necessarily exhaustive and suggests opportunity for future research to continue to deconstruct disclosure strategies. Third, these findings have startling implications generalizable to broader sustainability disclosure through voluntary programs, ESG rating schemes, and emerging mandatory regimes—when ratings are corrupted, meaningful improvements to societal outcomes are diminished (Aragòn-Correa et al., 2020; Espeland & Sauder, 2007; Wright & Nyberg, 2017). The article thus contributes to policy and practice by highlighting the need to refocus intermediary assessments of sustainability disclosure on more measurable aspects of performance while deemphasizing those aspects that can be more easily manipulated. At the same time, findings suggest the need for disclosure institutions to look toward methods of rating manipulation more germane to financial reporting (e.g., fraud and earnings management; Harris & Bromiley, 2007; Healy & Wahlen, 1999; Wesley & Ndofor, 2013). Considering the growing convergence between data from corporate voluntary disclosure and metrics that drive ESG investing decisions (Berg et al., 2020; Cort & Esty, 2020; Khan et al., 2016), this article highlights flaws and offers solutions to improve the reliability of climate disclosure.
Theoretical Background
In response to demand from investors and other stakeholders for material details of individual firms’ sustainability performance, firms use voluntary disclosure to communicate sustainability-related information. As the practice of voluntary disclosure has grown, various reporting standards have emerged in an effort to improve the comparability and credibility of individual company reports. However, standards-based voluntary disclosure is yet complex: Proliferation of performance indicators along with qualitative aspects of performance belies easy interpretation (Boiral & Henri, 2017). Moreover, the lack of rigorous verification mechanisms allows firms broad discretion to craft disclosures in ways that mislead stakeholders as to the true level of performance, while stakeholders are aware of this and may not believe disclosures to be truthful (Diouf & Boiral, 2017). The ultimate consequence of this information problem is a market failure (Akerlof, 1970), conceivably leading to proliferation of misleading information (Lyon & Montgomery, 2015) while driving honest firms to withhold information for fear of excess scrutiny (Kim & Lyon, 2015; Lyon & Maxwell, 2011). Whereas standardization of information disclosure may be ineffective and potentially counterproductive, an expert, unbiased assessment of aggregate performance improves information utility (Bae et al., 2010). Stakeholders thus seek a source of credible information disclosure to enable commensuration while reducing proprietary costs in verifying and interpreting claims of individual firms, in terms of both prior and expected future performance. In other words, they demand two important information services: evaluation and verification of disclosures.
Sustainability rating schemes seek to address this demand for evaluation. Rating providers collect issue-specific information directly from firms, media, and other third-party reports, then analyze, aggregate, and disseminate consolidated outputs to the general public or subscribers. The rating enables commensuration of performance along dimensions relevant to user interest, while reducing information processing costs for those users (Lyon & Shimshack, 2015). Given the widespread utility of performance ratings for a broad range of stakeholder groups (Rindova et al., 2018), firms generally attend to those ratings through a variety of responses (Slager et al., 2021). Whereas ratings may motivate firms to demonstrably improve performance (Chatterji & Toffel, 2010), firms may likewise find motivation to address ratings through symbolic or misleading means (Fabrizio & Kim, 2019). Considering the recent proliferation of sustainability and ESG rating schemes, recent research has revealed systematic inconsistencies between similar raters (Berg et al., 2020; Chatterji et al., 2016), and this divergence is more pronounced for firms that have disclosed a greater quantity of information (Christensen et al., 2022), arguably supporting the notion that ratings may be influenced through disclosure of more subjective metrics. Moreover, higher visibility of ratings may also drive heightened media attention which can enhance the benefits to firms of obtaining a higher rating (Bermiss et al., 2014), such that firms may strategically promote the more favorable ratings to enhance impression management (Luca & Smith, 2015).
Specialized disclosure intermediaries occupy a somewhat unique niche in the world of sustainability ratings. Rather than collecting public information disclosed by and about firms, disclosure intermediaries directly solicit firm disclosure in a fixed format along specific issue domains (e.g., climate change), maintaining the benefits of standardization while limiting firm discretion in selective reporting. A successful intermediary achieves credibility by harnessing network effects generated through its normative and coercive influence (Delmas & Toffel, 2008; Reid & Toffel, 2009), distinguishing itself in the market for information (Boiral et al., 2020; Etzion & Ferraro, 2010), and engendering stakeholder trust through transparency of its own evaluation process (Schnackenberg & Tomlinson, 2016). Stakeholders, in turn, tend to place higher value on credibility of disclosure content than on mere conformance (Philippe & Durand, 2011), particularly when a formal evaluation of firm disclosure is provided. The intermediary’s formal assessment of disclosure quality thus becomes a key value proposition to stakeholders, and its favorable rating effectively serves as an explicit endorsement of firm performance. Achieving a higher rating thus allows the firm to differentiate from similar others that merely seek legitimation through proprietary disclosure (Cho et al., 2012). When stakeholders perceive the evaluative capabilities of the intermediary to be high, they associate a stronger third-party endorsement with higher perceived firm quality on relevant dimensions of performance (Stuart et al., 1999). Furthermore, higher credibility of authoritative intermediary assessments may influence stakeholders to rely on those evaluations (Tost, 2011) in place of their own proprietary analysis. The intermediary’s rating thus represents a thing of tangible value that firms can strategically target.
Notably, as voluntary mechanisms, intermediaries have no coercive authority to enforce accountability for accuracy in disclosure. Professional, third-party audits are the standard for ensuring accountability in regulated financial disclosure, though this mechanism has proven an imperfect mechanism over time (Healy & Palepu, 2003; Healy & Wahlen, 1999). Moreover, external audit is a costly practice, perhaps prohibitively so for unregulated, nonfinancial performance metrics. Meanwhile, stakeholders seek credible, third-party intermediaries that can independently verify accuracy of corporate claims (Dando & Swift, 2003). To meet this demand, intermediaries largely delegate verification to firm discretion using relevant third-party standards and external assurers, an approach previously developed to improve credibility of proprietary sustainability reports (Junior et al., 2014) but often found to be an insufficient mechanism of accountability for voluntary disclosure (Boiral & Heras-Saizarbitoria, 2020). A decentralized accountability mechanism may thus allow firms to attain higher ratings through misleading or symbolic means. Where the intermediary’s assessment of disclosure quality carries value, that assessment can be co-opted over time to promote a firm’s sustainability credentials (Slager et al., 2012), particularly when that assessment is highly visible (Bermiss et al., 2014). These two problematic aspects of intermediary services—credible assessment and perceived rigor in verification—may compromise effectiveness of the disclosure mechanism (Callery & Perkins, 2021).
Hypotheses
Firms may employ various strategies in effort to manipulate intermediary assessment and attain higher ratings. In this section, I introduce various potential modes of strategic disclosure and hypothesize how intermediary structure and methodology may unwittingly encourage their use. Following this, I highlight factors associated with firm disclosure along dimensions valued by stakeholders (e.g., verification and performance evaluation) that ultimately support a more favorable intermediary assessment. Finally, I hypothesize how strategic disclosure may partly explain the relationship between those valued factors and intermediary assessment depending on the temporal orientation (e.g., verifiable past performance versus expected future performance) of those factors.
Strategic Disclosure
As noted above, while a favorable intermediary assessment (i.e., a high rating) may convey positive impressions of firm performance, lack of audit rigor exposes its assessment to misleading disclosure. Sophisticated firms, armed with a detailed methodology for optimizing the intermediary’s rating, may adeptly find loopholes to manipulate disclosure requirements to improve impressions of performance and reduce probability of detection (Healy & Wahlen, 1999). Strategic disclosure represents one or more approaches used by a firm to influence the intermediary assessment, misleading or not, outside of substantive demonstration of performance factors assessed by the intermediary’s formal methodology. Prior research has explored several general methods of strategic disclosure firms may employ to influence ratings and external perceptions, both qualitative and quantitative. Here I describe four methods specifically associated with voluntary carbon disclosure in the literature: obfuscation, complexity, performance manipulation, and attentiveness to rating methodology.
Obfuscation
The use of obfuscation in corporate reporting is a well-studied phenomenon, particularly in accounting and finance literature (Loughran & McDonald, 2016). Obfuscation is generally used to obscure or distract attention from poor performance, particularly when managerial intention is to maintain information asymmetry (Bushee et al., 2018) or impression management (Asay et al., 2018). Obfuscating text is also prominent in corporate sustainability and other nonfinancial disclosures. Firms may obfuscate to exaggerate value provided to external stakeholder groups (Haller et al., 2018) or to cloud perceptions of negative impacts. Recent research has specifically found obfuscation to have a buffering effect on intermediary ratings (Fabrizio & Kim, 2019).
Complexity
Disclosure complexity is often associated with obfuscation, though research increasingly identifies these as a separate constructs (Loughran & McDonald, 2014). Increased complexity of a corporate report may limit capacity of stakeholders to isolate and interpret material disclosures. Routine corporate disclosures often contain value-relevant information, but increasing complexity of reports reduces the effectiveness of investor response (You & Zhang, 2009). Evidence from stock returns shows that negative, material information disclosures are often overlooked by sophisticated investors due to complexity of reports (Cohen et al., 2020). More complex disclosures may be more likely to convey completeness and competence, while often written primarily to enhance impression management (Talbot & Boiral, 2018).
Performance manipulation
In regulated financial reporting, material misstatements of performance occur not infrequently, despite substantial consequences for detection (Healy & Palepu, 2003). Research has attributed various factors to the direct manipulation of accounting measures, such as incentive-based compensation and underperformance (Harris & Bromiley, 2007). Falsified voluntary disclosure of nonfinancial performance arguably carries lesser consequences and may encourage heightened incidence, considering the difficulty in detecting such misstatements given the wide latitude and variation in measurement methods for ESG performance (Semenova & Hassel, 2015). As institutional investors increasingly turn to ESG performance metrics to guide investment strategies, research indicates that investors perceive certain metrics as material (Grewal et al., 2021). Moreover, research has indicated provision of false accounts in intermediated disclosure may be prevalent (Callery & Perkins, 2021), suggesting that firms see value at acceptable risk in manipulating disclosures to enhance impressions.
Attentiveness to rating methodology
Opportunistic disclosure represents a relatively recent phenomenon in management research (Callery & Perkins, 2021). Where motivation to improve ratings is high and the capacity for improvement is low, firms may substitute symbolic attainment of higher ratings for substantive efforts to address underlying performance (Slager et al., 2021). Armed with guidelines and criteria for intermediary evaluation methodology, firms may have the ability to craft disclosure in ways that improve rating. Attentiveness may then be revealed by how responsive a firm is to changes in the evaluation methodology that occur over time. When evaluation criteria change to reward a certain aspect of disclosure content, an attentive firm is likely to change its disclosure to match this new criteria in the interest of optimizing its rating. Attentiveness to such changes in the “rules” of assessment is further bolstered by the emergence of a cottage industry of disclosure consultants promising to improve firms’ external ratings.
The four methods of strategic disclosure highlighted above all represent tools available to firms to take advantage of intermediary mechanisms to optimize its disclosure rating, enhancing stakeholder perceptions of performance through potentially misleading means.
Factors Supporting Intermediary Assessment
As noted above, the climate intermediary seeks to satisfy stakeholder demand for credible information disclosures (i.e., demand for verification of disclosure and evaluation of both past and expected future performance). In this role, it solicits disclosure from firms in support of these goals. Disclosures comprise various factors representing a firm’s policies and performance relevant to the quality of its efforts to confront the climate crisis. These factors are measurable indicators of climate disclosure and performance that directly contribute to the intermediary’s assessment, yet are potentially subject to manipulative intent by firms wishing to promote a misleading impression of climate performance. As research has shown the link between carbon disclosure and carbon performance is somewhat mixed (Qian & Schaltegger, 2017), I consider factors more closely associated with credibility and performance in preference over strict measures of disclosure quantity or completeness. According to the aforementioned stakeholder demand for credible intermediary services, I broadly consider these factors as contributing to (a) verification of prior performance and (b) expected future performance.
Verification of prior performance
Verification is the informed, professional opinion of a qualified, external observer that a firm’s disclosure is accurate. Whereas intermediaries generally lack coercive power to ensure verification of corporate claims expressed through voluntary disclosure, they may reward aspects of disclosure that approach formal verification. Firms often obtain and report third-party assurance of sustainability claims, including climate performance measures, to address stakeholder demand for verification. However, prior research has shown assurance is subject to manipulation concerns given variability in applied standards, assurer expertise, and potential conflicts of interest (Ball et al., 2000; Boiral & Heras-Saizarbitoria, 2020). Moreover, firms have the option of obtaining assurance at different levels of rigor with corresponding difference in associated costs incurred. The level of assurance granted is generally classified as either “reasonable” or “limited.” Reasonable assurance expresses a relatively high level of confidence in compliance of the entity being assessed and thus dictates a more rigorous (and costly) investigation. Limited assurance expresses a more moderate level of confidence, is less costly, and is associated with less rigorous investigation. This dichotomy suggests that firms intending to mislead may choose limited assurance and that firms seeking verification of substantive disclosure may choose reasonable assurance to signal a greater level of disclosure credibility (Bagnoli & Watts, 2017).
Intermediaries may solicit other aspects of disclosure that implicitly support perceived verification. Some firms may disclose related information through other external mechanisms with more stringent accountability for claims. For example, information disclosed through regulated channels (e.g., comprehensive financial disclosures mandated by securities regulators) may carry higher stakeholder scrutiny and associated penalties for inaccurate claims, which in turn may support greater external perceptions of credibility (Reimsbach et al., 2018). Considering that sophisticated stakeholders may be more directly attuned to consistency in disclosures (Cohen et al., 2020), firms may seek to minimize differences in disclosure between channels to avoid scrutiny (Depoers et al., 2016). Moreover, recent research suggests that firms exercise greater restraint in disclosing sustainability information via these regulated channels relative to disclosure in voluntary reports (Grewal, 2019), suggesting firms may be less willing to report misleading information through intermediaries that may increase exposure to unwanted scrutiny (Carlos & Lewis, 2018).
Expected future performance
With respect to performance evaluation, stakeholders seek the intermediary’s assessment of the quality of a firm’s policies and performance, not only with respect to prior accomplishments but also future prospects. As an analogy to the climate intermediary’s assessment, consider the role of equity research analysts. An analyst not only assesses past performance, but also must interpret a broad range of factors and offer its assessment of expected future performance (Bowers, 2015). In the context of climate change, where global policy drivers are necessarily oriented toward commitments of future performance (Höhne et al., 2021), a firm’s expected future performance may carry significant weight in an overall evaluation of the quality of the firm’s approach to addressing the climate crisis. While disclosure of past performance provides the intermediary a set of commensurable metrics to judge a firm’s accomplishments to date, the disclosure and perceived relevance of adopted policies and associated practices inform the intermediary’s evaluation of future performance. The intermediary directly solicits a range of disclosure factors that allow it to assess a firm’s prospects for future performance improvements (i.e., superior ability to both hasten its adaptation to climate risks and reduce its climate-related impacts). Firms have several tools to manage future climate performance, for example, adoption of performance-based incentives for achieving climate-related goals, emissions reduction initiatives, and formal emissions reduction targets. Such tools have been found to be largely effective in improving performance (Blanco et al., 2017; Flammer et al., 2019; Ioannou et al., 2016). However, aspects of these disclosures also lend to potential for manipulation (Berrone & Gomez-Mejia, 2009; Callery & Kim, 2020; Dahlmann et al., 2019).
In summary, these factors ostensibly represent tangible efforts by a firm to communicate the quality of its approach to addressing climate change, and are arguably supportive to an intermediary’s assessment, and thus stakeholder perceptions of verification and expected future performance. To the extent these supporting factors are solicited and rewarded by an intermediary, they are systematically linked to its rating. The second hypothesis is thus largely structural:
Temporal Orientation of Supporting Factors
Whereas these relevant supporting factors may typically be rewarded with a more favorable intermediary assessment and corresponding rating, as described above, they may also be susceptible to manipulation by firms with misleading intentions. The extent to which these supporting factors credibly represent a firm’s underlying quality may vary according to the temporal orientation of their contribution to the intermediary’s assessment. Verification of disclosure implies an assessment of recent performance (Boiral & Heras-Saizarbitoria, 2020) and is relatively unambiguous in terms of its ability to convey aspects of firm performance, whereas evaluation of expected future performance is subject to greater variability in interpretation (Christensen et al., 2022) and thus may be more susceptible to bias or manipulation (Ikram et al., 2019). In terms of the methods of strategic disclosure outlined above, firms may find greater success in leveraging disclosure of more forward-looking factors to attain higher rating by, for example, using obfuscating text to confuse interpretation over claims of efforts to improve future performance, or overwhelming the intermediary’s assessment with excessively complex information. Firms with misleading intent may likewise find greater ease in outright manipulation of performance disclosures or tailor responses to established assessment criteria when those criteria are more forward-looking.
Importantly, strategic disclosure represents methods to influence intermediary assessment outside of direct demonstration of climate performance, and is not necessarily always misleading. Accordingly, supporting factors may mitigate or otherwise exacerbate the effects of strategic disclosure when used for misleading intent according to their temporal orientation. In other words, if strategic disclosure partly explains (mediates) the relationship between a supporting factor and intermediary rating, one may infer that strategic disclosure represents a successful effort to illicitly attain a higher rating. Whereas I expect all supporting factors to be positively associated with ratings (Hypothesis 2), factors associated with verification of past performance will be less susceptible to manipulation through strategic disclosure, while factors associated with expected future performance will be more susceptible. The presentation of a formal statement of assurance from a licensed professional is evidentiary (though the underlying assurance process is imperfect; see Boiral & Heras-Saizarbitoria, 2020), and firms may face stakeholder sanctions when perceived to act in a manner inconsistent with its disclosure (Carlos & Lewis, 2018).
Meanwhile, factors used to assess future firm performance are often more subjective in nature (Arjaliès & Bansal, 2018) and may require detailed knowledge of organizational practices to accurately evaluate. Forward-looking performance-related factors may thus exacerbate the misleading effects of strategic disclosure, as the potential effectiveness of these factors in enhancing climate-related performance relies on subjective assessment of efforts that have not yet occurred. For example, subjectivity in performance-based executive compensation contracts is associated with management capture (Ittner et al., 2003) and lower attainment of ESG-related performance (Maas, 2018), while emissions reduction targets may be structured and managed in ways that defer accountability over future performance (Callery & Kim, 2020). I thus predict that strategic disclosure will partly explain the relationship between forward-looking supporting factors and rating, while strategic disclosure is less likely to explain the relationship between verification-related supporting factors and rating. Figure 1 presents a conceptual model of these hypothesized relationships.
Model of strategic disclosure and supporting factors on rating.
Method
Research Setting
To test these hypotheses, I analyze several years of corporate disclosures to CDP. CDP is a not-for-profit organization founded in 2002 with backing from global institutional investors that seeks to improve corporate transparency and performance related to climate change. CDP is perhaps the preeminent intermediary in this domain and has attained substantial growth since its inception; Figure 2 displays the number of publicly participating firms and institutional investor signatories over time.
CDP participating firms and institutional investor signatories.
CDP annually solicits voluntary disclosure from the largest publicly traded global firms using a fixed format questionnaire. In 2015, this questionnaire comprised more than 250 individual items utilizing a variety of response formats: numerical entry, discrete choice (i.e., drop-down menu selection), free text, and document upload (e.g., for certificates of assurance). Each year, CDP collects both solicited and unsolicited responses, analyzes the data, computes ratings, and publishes a series of regional and sectoral reports based on the content of disclosures made by firms. From 2010 to 2015, CDP compiled and promoted two separate quantitative ratings for each firm based on (a) completeness of disclosure and (b) overall management of climate performance. CDP’s disclosure score is a numerical rating on a 0 to 100 scale that ranks firms on overall transparency. CDP’s performance score is a categorical rank score that places firms in one of six performance bands (A, A-, B, C, D, E) based on its assessment of the firm’s overall climate change management. Various questions are assigned a point value on either or both rating scales, and respondent firms may attain those points by providing sufficient acceptable answers. Thus, whereas CDP’s disclosure score represents its assessment of the level of firm transparency in responding to its questionnaire (i.e., the quantity or completeness of information disclosed), its performance score represents its assessment of a firm’s efforts to confront climate change: a measure of its underlying quality. Importantly, in the interest of its own organizational transparency, CDP makes public its scoring methodology, effectively providing firms with a “recipe” to improve scores through disclosure choices.
Data
The sample is drawn from CDP ratings and detailed firm disclosures from years 2011–2015, which when combined with supporting data sets (Thomson-Reuters Worldscope and Trucost, as described below) contain 5,638 unique firm-year observations from 1,672 individual firms. To aid interpretation of the following variable definitions, Table 1 provides a summary description of the key constructs and individual measures.
Variable Descriptions.
Rating
The dependent variable measure is an industry-relative rating based on CDP’s performance score. I first assign the CDP letter grade performance bands to an integer scale, where “A” equals 6 and “E” equals 1 (“A-” is scored 5). The variable Rating is then calculated as each firm’s performance score minus its industry mean performance score, using the Global Industry Classification Standard (GICS) industry group (four-digit code).
Strategic disclosure
In the above hypotheses, I conceptualize multiple methods of strategic disclosure featuring both qualitative and quantitative constructs. Separate independent variables correspond to each of the four methods of strategic disclosure, with more qualitative constructs (i.e., obfuscation and complexity) corresponding to presentation of disclosure content in text to influence external perceptions, and more quantitative constructs (i.e., performance manipulation and attentiveness to rating methodology) addressing the numerical structure of intermediary assessment. For the qualitative measures, the Gunning fog index is commonly used as a measure of textual and verbal obfuscation in both accounting (Bushee et al., 2018; Loughran & McDonald, 2016) and management (Fabrizio & Kim, 2019) literatures. I aggregate the raw text for each free text questionnaire item with nonzero performance score value (i.e., only those items that CDP assesses to compute the performance score). The variable Obfuscation uses the Gunning fog index of this aggregate text, calculated as
To measure the more quantitative constructs performance manipulation and attentiveness to rating methodology, I follow recent work (Callery & Perkins, 2021) to exploit the firm-level panel structure of recurring CDP responses. The binary variable Manipulation uses the measure of false accounts developed by Callery and Perkins (2021); it takes the value 1 for observations in which a firm claims attainment of carbon emissions reductions (a claim rewarded by CDP’s rating methodology), but instead exhibits an actual increase based on comparison of current year disclosed emissions with prior year. The variable thus exposes inconsistency in simple disclosure claims that is accretive to a firm’s rating. Meanwhile, the binary variable Attentiveness uses the measure of opportunism developed by Callery and Perkins (2021). This measure exploits minor changes in the CDP’s rating methodology from year to year to identify firm-year observations in which a firm’s response to a particular questionnaire item has changed from the prior year in concert with a corresponding change in the rating points awarded by CDP for that item; such a change is indicative of the respondent firm’s attentiveness to the assessed value of a particular item response. Attentiveness takes the value 1 for observations in which the firm exhibits such a response.
Supporting factors
Factors supporting assessment are taken from CDP disclosure data. To analyze a broader range of factors rewarded by CDP’s rating methodology, and to distinguish between verification of prior performance and expected future performance, I specify multiple supporting factors: three measures corresponding to verification and three measures corresponding to expected future performance. For verification, note that CDP requests firms to indicate whether assurance was obtained for their Scope 1 and 2 emissions inventory disclosures, and separately, the level of assurance obtained. The binary variable Reasonable Assurance indicates firms that declared and provided evidence of reasonable assurance, and the corresponding variable Limited Assurance in kind. Next, as described above, firms disclosing similar climate-related information through other regulated channels may be more likely held accountable for disclosure consistency. CDP requests firms to indicate whether they disclose through such channels (e.g., via annual financial reports and regulated securities filings); the binary indicator Regulated Disclosure identifies these firms.
For future performance-related factors, note that firms may indicate to CDP whether executives are paid performance-based incentives for achievement of climate-related goals, a factor that explicitly relates to expected future climate-related performance improvement, yet is relatively ambiguous in terms of the materiality of incentivized performance metrics (Flammer et al., 2019; Ikram et al., 2019); the binary indicator Management Incentives identifies such firms. Firms may also report to CDP the presence of emissions reduction initiatives, a forward-looking indicator of future performance; the indicator Emissions Initiatives identifies firms that so report. Finally, formal emissions reduction targets are likewise explicit indicators of expected future performance; however, the association of targets with future emissions reductions is highly variable (Dahlmann et al., 2019; Ioannou et al., 2016); the indicator Emissions Target identifies such firms.
Control variables
We add a set of controls that are commonly associated with voluntary disclosure of environmental and other nonfinancial performance metrics. The binary indicator Environmentally Sensitive identifies firms operating in emissions-intensive industries, defined as GICS industry groups Materials, Energy, Transportation, and Utilities. The natural log of firm total Assets represents size and a corresponding proxy for firm visibility. Return on Assets controls for firm profitability, a factor often associated with environmental performance. These measures are obtained from Thomson-Reuters Worldscope financial database. Finally, firm Carbon Emissions intensity is measured as the natural log of combined Scope 1 and Scope 2 emissions per unit revenue (in U.S. dollars), as reported by Trucost. Importantly, including emissions data provides a control for relative performance between firms in the context of climate change. These three performance-related controls are lagged 1 year to signify firm performance prior to CDP disclosure.
Descriptive statistics
Table 2 displays a summary of descriptive statistics, with variables grouped by analytical categories for clarity. Obfuscation is skewed right from normal and is winsorized at the 98th percentile to reduce influence of outliers. Complexity is natural log transformed to normalize the raw exponential distribution. The binary indicators Manipulation and Attentiveness are found in roughly 13% and 30% of observations, respectively. Sample means (proportion) for the six supporting factors are likewise noted in Table 2.
Summary Statistics.
Table 3 contains covariate correlations; most are significant at p < .05 (coefficients where p > .05 are italicized). While Complexity is structurally correlated with the dependent variable, it is not significantly correlated with other independent variables. An examination of variance inflation factors (VIF) for baseline regression models indicates multicollinearity is not a concern (mean VIF = 1.45 and all individual VIF < 2, safely below commonly accepted levels).
Correlation Matrix.
Note. Coefficients where p > .05 are italicized.
Analytical Approach
Hypotheses 1 and 2 are tested using panel ordinary least squares with firm-level fixed effects. All models include year fixed effects, with robust standard errors clustered by firm. Tests of Hypotheses 3a and 3b are complicated by the inclusion of multiple measures of the theoretical constructs (strategic disclosure and supporting factors). To account for the potential correlation of error terms between individual models relating each measure of strategic disclosure to supporting factors (see Figure 1), I estimate a multiple mediation analysis (Preacher & Hayes, 2008) using seemingly unrelated regression estimation (SURE; Zellner, 1962). I use a panel variable within-transformation to preserve firm-level fixed effects and test significance of indirect (mediating) effects using bootstrapped standard errors and bias-corrected confidence intervals (Preacher & Hayes, 2008). SURE is indicated generally to avoid estimation bias in cases with systems of multiple linear models due to potential correlation of individual regression residuals (Carlson, 1978; Knott, 2001) and specifically in cases of multiple mediation (Zhu et al., 2018).
Results
Results of the first analysis—identification of key factors influencing ratings—are displayed in Table 4. Model 1 includes only supporting factors and control variables and establishes the positive and significant effect of each supporting factor on rating, as predicted by Hypothesis 2; all six factors are jointly significantly different from zero (p < .001). As each factor is represented by a binary variable, interpretation of effect sizes is similar; for example, firms providing reasonable assurance, all else equal, increase industry-relative score by roughly one quarter point on the letter grade performance band. Models 2 to 5 include each strategic disclosure variable individually, with Model 6 including all measures simultaneously. To better interpret effect sizes, for example, note that a 10% increase in Complexity, all else equal, is associated with roughly 0.043 points increase in industry-relative score, while Manipulation is associated with 0.074 points increase, all else equal. Interestingly, while three of the four measures are positively associated with rating, obfuscation is negatively associated with rating, indicating only partial support for Hypothesis 1.
Effects of Strategic Disclosure and Supporting Factors on Rating.
Note. Robust standard errors clustered by firm in parentheses. All models are panel ordinary least squares with firm-level fixed effects and year indicators.
p < .1. ** p < .05. *** p < .01.
To conduct formal tests of Hypotheses 3a and 3b, I perform multiple mediation analysis (Preacher & Hayes, 2008) using SURE (Zellner, 1962). SURE analyzes the complete system of multiple mediation equations relating all supporting factors to each measure of strategic disclosure (four models), and all measures simultaneously with rating (one model), while accounting for potential correlation between error terms of the multiple models. Table 5 displays the baseline SURE results; a Breusch–Pagan test rejects the null hypothesis of independent residuals from the individual mediator models (p < .001), indicating SURE is advised (Knott, 2001; Zhu et al., 2018). Models 1 to 4 show results of individual regressions of each strategic disclosure measure on all supporting factors and control variables. The presence of at least one significant coefficient in each model indicates a significant relationship between supporting factors and each method of strategic disclosure, indicating a potential mediating relationship. Model 5 displays the results of the full model, confirming significant relationships for all strategic disclosure measures and supporting factors with rating. Model 5 coefficients on the six supporting factors represent the respective direct effects on rating. To test for mediating effects (Hypotheses 3a and 3b), one must next calculate the indirect (i.e., mediated) effects using partial nonlinear combinations of the various model coefficients.
Direct and Indirect Mediating Effects.
Note. Results of seemingly unrelated regression estimation on full multiple mediation system of equations including four methods of strategic disclosure (Models 1–4) and one measure of rating (Model 5) using within transformation. Standard errors in parentheses. Model 5 corresponds to the direct effect of all factors on rating. Indirect effects of each method of strategic disclosure (i.e., multiple mediation) are computed and aggregated separately (see Table 6).
p < .1. ** p < .05. *** p < .01.
Table 6 displays estimated indirect effects (i.e., the portion of the total effect of supporting factors on rating that is mediated by strategic disclosure). Individual and total indirect effects are computed from nonlinear combinations of the SURE coefficients (Table 5), using bootstrapped standard errors (1,000 replications) to account for nonnormally distributed residuals. I further compute bias-corrected 95% confidence intervals (in square brackets) for the calculated indirect effects instead of traditional significance tests given the potential for skewed distributions (Preacher & Hayes, 2008). The proportion of the total effect that is mediated is shown in curly braces (for total indirect effect only; right-hand column).
Multiple Mediation Indirect Effects.
Note. Coefficients are indirect mediation effects computed by nonlinear combination using results from seemingly unrelated regression estimation. Bootstrapped standard errors (1000 replications) in parentheses. Bias-corrected 95% confidence intervals in square brackets. Proportion of total effect that is mediated listed in curly braces (right-hand column only).
Proportion of Reasonable Assurance total effect on Rating mediated by strategic disclosure has a point estimate of 5% but neither direct nor indirect effect is significantly different from zero at 95% confidence level.
Indicates estimated effects with strictly positive bias-corrected 95% confidence interval.
Results show that the total indirect (i.e., mediated by strategic disclosure) effect of Reasonable Assurance on rating is not significantly different from zero (i.e., bias-corrected 95% confidence intervals overlap zero), indicating a null mediating effect of strategic disclosure and providing partial support for Hypothesis 3a. A plausible explanation is that this verification-related supporting factor may dissuade misleading through strategic disclosure. Meanwhile, the effects of Limited Assurance and Regulated Disclosure are partly mediated by strategic disclosure, suggesting the lower stringency associated with obtaining limited assurance, or insufficient accountability for regulatory disclosure of climate-related factors, may make these factors more susceptible to strategic disclosure. However, note that only 8% of the total effect (direct plus indirect) of Limited Assurance is mediated, while one third of the total effect of Regulated Disclosure is mediated; in both cases, the mediating effect is largely driven by Complexity. Results also show significant indirect effects of all three forward-looking supporting factors on rating, supporting Hypothesis 3b. Moreover, all four methods of strategic disclosure contribute (variably across different supporting factors) to the mediating effect, with proportion of total effect mediated slightly larger. For example, the total indirect effect of Management Incentives is significantly larger than that of Limited Assurance (difference = 0.050, p < .001).
I performed several robustness checks and additional analyses. Using different dependent variable measures of intermediary assessment, for example, the raw (without industry-relative adjustment) performance score (using ordered logit regression) and a binary indicator corresponding to performance scores of “A” (signifying inclusion on CDP’s publicly promoted “A-list”) yielded results to be qualitatively similar to baseline. I also tested quadratic terms on Obfuscation and Complexity, as well as an interaction term between these two variables, again finding no notable change in results. Recognizing that the four methods of strategic disclosure and six supporting factors analyzed here may constitute multiple realizations of two broader underlying constructs (as per Figure 1), a more parsimonious mediation model may be obtained using aggregate measures of the hypothesized constructs. An exploratory factor analysis of all 10 independent variables using varimax rotation finds Obfuscation and Complexity load on one factor while Manipulation and Attentiveness load on a second. Meanwhile, all three future performance-related supporting factors load on a third factor, while the two assurance variables load on a fourth, with the final factor (Regulated Disclosure) loading together with Obfuscation and Complexity on the first factor. The presence of multiple underlying factors for each of the two hypothesized constructs (methods of strategic disclosure and supporting factors) indicate the need for a multiple mediation analysis (Preacher & Hayes, 2008) regardless of aggregation, supporting the baseline results as described in Table 6.
Considering the significant mediating effect of strategic disclosure on the relationship between Regulated Disclosure and Rating, and the off construct loading of Regulated Disclosure with qualitative methods of strategic disclosure (see above), I further analyzed the nature of regulated disclosure by considering the phenomenon of mandatory climate disclosure. Whereas the Regulated Disclosure measure describes voluntary disclosure of climate-related information within a regulated financial disclosure filing, mandatory disclosure represents a specific government regulation dictating the disclosure of climate-related information. To measure this, I identified the year in which various countries initiated mandatory disclosure of climate-related information (data from OECD, 2015) and defined an indicator variable identifying which firm-year observations were subject to such mandated disclosure according to their primary country of domicile. Whereas this mandatory disclosure indicator has no significant effect on Rating as a control variable, the indicator has a positive moderating effect on the relationship between two methods of strategic disclosure (Obfuscation and Complexity) and Rating (i.e., as per Table 4 analyses), but no direct effect on any of the methods of strategic disclosure (i.e., as per Table 5 analyses). This result suggests that while a mandatory climate disclosure obligation appears to have no direct effect on a firm’s rating or propensity for strategic disclosure, it may strengthen the relationship between the more qualitative methods of strategic disclosure and rating.
Finally, I consider the possibility that firms may exhibit learning effects, developing greater propensity for strategic disclosure as they gain experience. I alternately impose a time trend variable (i.e., disclosure year) and a measure of firm experience (i.e., number of years the firm has been present in the sample) and include in regressions of Rating on all factors (as per Table 4 analyses) and of each method of strategic disclosure on supporting factors (as per Table 5 analyses). All else equal, both time measures are negatively associated with Rating, indicating a general downward trend in CDP performance score for firms in the sample over time, all else equal. However, both measures positively moderate the effects of Complexity and Attentiveness on Rating, suggesting the potential presence of a learning effect for firms employing these methods of strategic disclosure. Moreover, both time measures are negatively associated with Obfuscation, indicating a decreased prevalence in this method over time, but are positively associated with the other three methods of strategic disclosure, indicating that these methods become more commonly used over time. Implications of these findings are addressed in the Discussion section.
Discussion
Motivated by the apparent disconnect between theory and practice in the perceived credibility of unaudited, intermediated voluntary disclosure, this article set out to examine the prevalence and effectiveness of several potential firm strategies for strategic disclosure to CDP, a prominent carbon intermediary. Findings show that efforts to achieve higher ratings through strategic disclosure are varied (multiple methods employed), widespread (binary measures observed in substantial portions of sample, see Table 2), and relatively effective (positively associated with rating). According to priors on the temporal orientation of disclosure factors supporting intermediary assessment, forward-looking factors are more susceptible to strategic disclosure. Findings show strategic disclosure explains a significant portion of the overall effect of three different expected future performance-related supporting factors on rating, while strategic disclosure does not significantly mediate the relationship between two of three verification-related supporting factors and rating. Indicators of expected future climate performance (management incentives, emissions initiatives, and emissions targets) are arguably easier to misrepresent. For example, specific details of climate-related performance incentives are difficult to fully interpret, even with public disclosure of incentive contracts available (Flammer et al., 2019), and may be prone to management capture (Keddie, 2020). Although this article does not explicitly evaluate environmental outcomes, results highlight a potential tension in performance incentives that warrants further scholarly attention. In addition, certain aspects of emissions reduction targets are associated with symbolic intent (Dahlmann et al., 2019) and may serve to defer accountability for climate action well into the indeterminate future (Callery & Kim, 2020). When the effect of these factors is partly explained by both qualitative (i.e., obfuscation, complexity) and quantitative (i.e., manipulation, attentiveness) modes of strategic disclosure, it suggests misleading intent. Meanwhile, the lack of mediating effect of strategic disclosure on some verification-related factors (reasonable assurance and, marginally, regulated disclosure) suggests these factors may deter firms from misleading disclosure. Finally, the relatively lax rigor associated with limited assurance engagements (Bagnoli & Watts, 2017; Callery & Perkins, 2021) may allow firms more leeway to engage in misleading disclosure.
With respect to direct effects of strategic disclosure on intermediary ratings, findings show that efforts to manipulate quantitative evaluative criteria (performance manipulation and attentiveness to rating methodology) are associated with incremental improvement in rating relative to industry peers. Regarding qualitative measures, obfuscation is weakly significant and associated with lower ratings, partly contradicting Hypothesis 1. While this result appears to contradict recent research (Fabrizio & Kim, 2019), note that Fabrizio and Kim (2019) examined heterogeneous effects across different discrete levels of performance score. Moreover, the multiple mediation analysis indicates that obfuscation yet partly explains the relationships of both regulated disclosure and emissions initiatives on ratings, which can be attributed to the negative association of these variables with obfuscation. That is, regulated disclosure may deter firms from obfuscation, with lower obfuscation then associated with slightly higher rating (taken together, the indirect effect). Meanwhile, findings show disclosure complexity to be the primary method of strategic disclosure driving the significant mediating effects. However, inferring strategic intent from a longer text response is not directly supported by the analysis and is a key limitation. Considering the unique context of the research setting, I reviewed selected passages and subjectively observed that many high-scoring text responses may be longer simply to convey a clear view of climate leadership rather than distract from a relative lack of substantive performance. These concerns highlight a need to better distinguish completeness of response from purposeful complexity. Future research may illuminate this puzzle by evaluating more precise measures of complexity or varying linguistic tone in text responses, perhaps by disaggregating latent constructs (Bushee et al., 2018; Crilly et al., 2016).
Meanwhile, supplementary analysis of mandatory disclosure regimes in specific countries suggests that firms subject to mandatory climate disclosure are associated with greater strategic disclosure over multiple methods, and may be more effective at using qualitative methods of strategic disclosure to influence intermediary ratings. A possible explanation for this result is that mandatory climate disclosure mechanisms often combine quantitative with qualitative mechanisms (Leong & Hazelton, 2019), potentially exposing those mechanisms to the well-documented legitimation practices of firms (Bansal & Clelland, 2004; Hahn & Lülfs, 2014), and perhaps rendering them less effective as mechanisms for improved transparency (Cong et al., 2020; Peters & Romi, 2013). Such an explanation would arguably be consistent with the possibility of strategic disclosure learning effects, as suggested by results of the supplementary time trend analysis, which indicates that two methods of strategic disclosure (complexity and attentiveness) grew more effective at improving ratings over time. Overall, given the prominent role of intermediaries as well as ESG raters on quantifying performance indicators based on text disclosures, the effect of misleading intent on third-party ratings of qualitative disclosure factors warrants further attention from researchers. Whereas prior research has suggested that CDP and other intermediaries do not lead firms to a significantly higher level of performance on dimensions of concern (Cho et al., 2012; Matisoff, 2013), these findings suggest that intermediated disclosure does not necessarily provide a true picture of corporate transparency.
These findings have troubling implications. Whereas academics and stakeholder groups have long understood that dubious firm intentions lead to untrustworthy voluntary disclosure (Laufer, 2003; Talbot & Boiral, 2018), intermediaries such as CDP are arguably developed to improve reliability of disclosure, absent any formal mandate to govern disclosure accuracy. The general market success of these mechanisms (see Figure 2) and growing financial market embeddedness may partly contribute to the problem (Clark & Newell, 2013). According to the authors’ interviews with multiple institutional investors, CDP data are widely integrated into other corporate ESG information and data services, and are increasingly being used as investment evaluation criteria as more capital is being directed into ESG-integrated investments. Whereas prior research has shown that similar intermediaries may wield substantial power in driving firms to improve sustainability performance (Slager & Chapple, 2016), these findings suggest that firms may have the capacity to turn this power on its head through strategic disclosure. CDP has leveraged its prominent role in governing voluntary climate disclosure to establish its framework as a model for ongoing efforts to develop regulatory disclosure standards (IFRS, 2022), heightening the importance of constraining the ability of firms to manipulate the standard.
Institutional theory has suggested and provided support for Campbell’s Law: As the use of a rating system becomes more widespread, corruptive pressures begin to dissipate the substantive meaning that originally made the ratings credible (Espeland & Sauder, 2007); these results are consistent with this notion. Sophisticated firms recognize that unaudited and unsubstantiated voluntary disclosure of nonfinancial performance is subject to stakeholder scrutiny and direct effort to mislead carries risk of backfire (Marquis et al., 2016). With a clear roadmap of how disclosed performance will be rated (e.g., the CDP scoring methodology), opportunistic firms are better able to use strategic disclosure to craft a desired impression with lower risk of detection. Firms may thus attain a more favorable intermediary assessment to enhance the credibility of their claims under stakeholder scrutiny; the extent to which such scrutiny is relieved through attainment of higher ratings represents a potential avenue for future research. That a concerning number of firms manipulate certain performance metrics (roughly 13% of the sample, see Table 2) is an interesting question itself, and critical for intermediaries to understand to maintain credibility over the long run; some characteristics of substantive disclosure are addressed in this article, with significant opportunity for future research to explore this in greater detail.
This article presents important implications for theory. The primary contribution to literature on commensuration and organizational impression management extends knowledge of ways that firms may use strategic disclosure to illicitly attain higher intermediary ratings, specifically highlighting the relative importance of temporal orientation of disclosure factors (i.e., verification of prior performance versus assessment of expected future performance) in determining the susceptibility of performance ratings to strategic disclosure. Notably, the results contradict prior research suggesting that intermediary ratings accurately portray firm quality (Giannarakis et al., 2017; Luo & Tang, 2014; Uyar et al., 2020). It also extends recent work on divergence of ESG ratings (Berg et al., 2020; Christensen et al., 2022) to encompass impression management theory and the practical context of intermediated disclosure. A key limitation of the article is that firms’ true underlying performance is not directly established; in fact, this is a largely unobservable trait given the nature of subjective firm disclosures assessed by intermediaries and other raters (Christensen et al., 2022). Importantly, the extent to which firms mobilize ratings for benefit in this context is not clear: Which firm types tend to actively publicize higher ratings and what are the outcomes? Recent research (Carlos & Lewis, 2018; Heras-Saizarbitoria et al., 2020) finds evidence that firms in some contexts attain costly certifications but do not actively promote. Other research offers conflicting findings on value of certain intermediary assessments (Dow Jones Sustainability Index [DJSI], as per Cho et al., 2012; Hawn et al., 2018; Robinson et al., 2011; and FTSE4Good, as per Slager & Chapple, 2016; Slager et al., 2012). Future research might explore these questions in the current context; for example, how do key stakeholders (investors, securities regulators, nongovernmental organizations) respond to CDP ratings and to what extent might CDP ratings (as an aggregate measure of performance) distract stakeholder scrutiny from direct assessments of underlying firm quality?
These findings also have important implications for the practice of intermediated voluntary disclosure, answering recent calls for increased attention to intermediary audit and verification mechanisms (Cort & Esty, 2020). Intermediaries promise to enhance transparency of corporate nonfinancial performance, particularly in terms of social and environmental impacts, by providing a level of governance lacking in proprietary firm disclosures around key issues. For example, carbon emissions and climate risk management are of growing importance to investors, regulators, and the public (Azar et al., 2021; Krueger et al., 2020). However, all disclosure institutions—including regulated, audited financial disclosures—are imperfect to varying extents (Healy & Wahlen, 1999). One problematic trend focal to this article is the propensity for intermediaries toward implementing simplified rating schemes. Corporate information disclosure is complex, and users typically require information processing services to make sense of the data (Bae et al., 2010). Importantly, voluntary disclosure intermediaries add value for stakeholders as trusted experts that distill complex corporate information into a small number of commensurable metrics. A result is that users are more likely to gravitate toward the metric and ignore the underlying data (Lewis & Carlos, 2019; Lyon & Shimshack, 2015), thus placing the burden of verification on the intermediary. When an intermediary’s rating is valuable, and the firm has access to the detailed rating methodology, the intermediary may cease to be pure vehicle of disclosure and instead become a rating scheme that firms can manipulate (Healy & Palepu, 2003; White, 2010). A major question yet to be answered by research is, “Why would investors and other stakeholders overlook these weaknesses in audit and accountability?” The authors’ interviews with institutional investors suggest that “better” data are not yet available, heightening the importance of establishing more reliable standards with more stringent accountability.
Whether intermediaries can strike an optimal balance between the information demands of users and credibility of data inputs is not addressed in this article; however, disclosure institutions must evolve to deal with these threats to credibility. The increasing recognition that institutionalization of formalized systems for assessing corporate social and environmental performance may stifle progress on substantive reforms (Wright & Nyberg, 2017) puts greater onus on those systems to add mechanisms that deter misleading disclosure. To its credit, in 2016 CDP overhauled its performance rating system, ostensibly raising the bar to attain recognition on its “A-List,” and arguably to mitigate the risk of manipulation such as highlighted by this article. In fact, as noted above, time trend analysis indicates certain methods of strategic disclosure grew more effective at influencing ratings over time; future research may examine whether such learning effects were neutralized by major changes in assessment methodology. Integrating other corroborating evaluations may further strengthen the value of its data and assessments; CDP already engages with other organizations (e.g., Global Reporting Initiative, RepRisk AG, Science-Based Targets, Task Force on Climate-Related Financial Disclosures, Climate Disclosure Standards Board) to work toward integrated and interdependent performance assessments, and is an active participant in driving efforts to produce climate disclosure standards for adoption by securities regulators worldwide (IFRS, 2022). Moreover, incorporating a validation model that rewards consistency in reporting over time, and with other channels of voluntary and regulated disclosure, can add credibility to assessments. Future research may further examine characteristics of firms making inconsistent disclosure across different channels (Depoers et al., 2016) and determine whether such structural changes are able to overcome the lack of formal audit and accountability mechanisms. Finally, as various regulatory jurisdictions move toward establishing mandatory sustainability reporting (Giamporcaro et al., 2020; Grewal et al., 2019), intermediaries can leverage external audit and accountability mechanisms to increase perceived penalties for misleading disclosure.
Conclusion
This article set out to examine the prevalence and effectiveness of various strategic disclosure methods in the rapidly growing practice of intermediated voluntary disclosure. Analyzing detailed firm disclosures to a prominent climate intermediary (CDP), results show that many firms employ strategic disclosure to successfully attain higher ratings. These findings have significant implications for the credibility of CDP and nonfinancial intermediaries in general; they highlight the tension faced by intermediaries to achieve their missions of improving corporate transparency and performance on important societal issues while also tightening control and accountability for their roles in providing credible assessment of firm performance. In light of these findings, climate disclosure mechanisms may improve effectiveness through focusing assessment more closely on measurable aspects of performance and away from forward-looking policy declarations and enhancing the rigor of external verification of disclosures. As rating schemes continue to proliferate and gain influence among investors and other stakeholder groups, accurately measuring and reflecting true company performance along nonfinancial measures of interest will continue to escalate in importance for both the credibility of institutions and the stakeholder groups they seek to support.
Footnotes
Acknowledgements
I thank Desirée Pacheco, Jonatan Pinkse, and participants at the 22nd GRONEN Reading Group meeting for their valuable feedback and suggestions. I also thank Associate Editor Frederik Dahlmann for his constructive feedback and helpful guidance, as well as two anonymous reviewers.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
