Corporate Social Performance,Firm Size,and Organizational Visibility: Distinct and Joint Effects on Voluntary Sustainability Reporting

Abstract

This study investigates the distinct and joint effects of corporate social performance (CSP), firm size, and visibility on a company’s decision to disclose sustainability-related information through sustainability reports. It seeks to provide more nuanced explanations for why certain companies tend to extensively report on their sustainability performance. First, while prior studies have predominantly focused on environmental reporting, the current analysis considers comprehensive sustainability reports that include both environmental and social issues. Second, the article argues that the effects of two important antecedents of legitimacy pressure—firm size and organizational visibility—should be analyzed separately rather than restricting the analysis on the effects of legitimacy pressure per se. Third, it argues that the hypothesized effects are nonlinear because the marginal costs and benefits of sustainability reporting vary with a company’s CSP level, its size, and its visibility in the public. Finally, although there is a strong link between CSP and sustainability reporting, the strength of this link depends on its size and visibility. The study of 280 companies in environmentally and socially sensitive industries provides considerable support for these hypotheses, including evidence that size and visibility independently affect sustainability reporting and that the shape of the CSP/sustainability reporting link is contingent upon firm size and visibility.

Keywords

corporate social responsibility sustainability reports stakeholder communication sustainability accounting and reporting voluntary disclosure

Many firms intending to generate additional monetary and nonmonetary benefits pay attention to corporate social responsibility (CSR) and sustainability (Schaltegger & Wagner, 2006). Whether these firms can reap the expected benefits of such engagement often depends on whether their stakeholders have sufficient and credible information about the firms’ actual corporate social performance (CSP).¹ Firms may, therefore, choose to voluntarily account for their sustainability performance (Brown & Fraser, 2006; Schaltegger & Burritt, 2006) and disclose that information to influence favorably the decisions of external stakeholders (Montiel, Husted, & Christmann, 2012; Narayanan, Pinches, Kelm, & Lander, 2000).

As a consequence, sustainability reporting² has developed into a mainstream practice during the recent decades (Burritt & Schaltegger, 2010; Herzig & Schaltegger, 2006; Schaltegger, Bennett, & Burritt, 2006). The number of sustainability reports has been continuously growing (Dhaliwal, Li, Tsang, & Yang, 2011), and the reports are becoming increasingly professionalized in terms of the use of international reporting standards and external audits (Kolk, 1999; Kolk & Perego, 2010; Simnett, Vanstraelen, & Wai Fong, 2009). Various public rankings regularly assess the quality of such reports, suggesting that this development is being closely followed by interested stakeholder groups (CorporateRegister.com, 2011; Lyon & Shimshack, 2015). These general trends notwithstanding, sustainability reports significantly differ from each other. In the absence of formal mandatory rules comparable with those of financial reporting, managers can and, to a great extent, do exert discretion as to the amount and quality of sustainability-related information they disclose (Brammer & Pavelin, 2008).³

This raises a question regarding the types of companies and their circumstances that tend to voluntarily disclose sustainability-related information to stakeholders: Is it primarily the noble social performers that disclose, or is the decision to issue sustainability reports independent of actual social performance? Why would companies with a nonsatisfactory social performance voluntarily disclose information regarding their inferior performance? Various stakeholders, such as investors, employees, or customers, base their investment, job, or consumer decisions upon their perceptions of the respective company’s social performance (Auger, Devinney, Louviere, & Burke, 2008; Bhattacharya & Sen, 2004; Hummels & Timmer, 2004; Turban & Cable, 2003). As such, perceptions are significantly influenced by firms’ sustainability reports; thus, a thorough understanding of these reports’ underlying motives is of great importance.

Prior research has identified two major drivers of voluntary sustainability reporting: CSP and legitimacy pressure.⁴ One argument that has been brought forward in the context of environmental reporting is that especially those companies with a relatively high environmental performance can report on their superior performance to reduce information asymmetry to interested stakeholders (Clarkson, Yue, & Richardson, 2004; Richardson & Welker, 2001). In this case, the primary function of environmental reports is to signal superior environmental performance and to demonstrate a competitive advantage through industry differentiation. Another argument is that some companies—irrespective of their CSP—simply disclose due to increased legitimacy pressure. That is, because they are expected to publish sustainability reports as a legitimate management tool (Aerts & Cormier, 2009; Hooghiemstra, 2000; Patten, 2002).

While such studies have improved our understanding of two important drivers of voluntary environmental disclosures, several important questions remain unsolved. These will form the point of departure for this study.

First, there is a trend toward publishing comprehensive sustainability reports that go beyond merely environmental disclosures. Besides environmental reporting, there has been a long tradition of social and ethical accounting, especially in European countries (Belal, 2002; Christophe & Bebbington, 1992; Gray, 2001, 2007; Owen, Swift, & Hunt, 2001; van der Laan, Adhikari, & Tondkar, 2005; von Wysocki, 1993). Not least due to the increasingly accepted multidimensional concept of sustainability, companies now typically include information on both, social and environmental information in their nonfinancial reports (Owen & O’Dwyer, 2008). Accommodating for this trend, the present study extends the analysis to sustainability reports including both environmental and social issues.

Second, prior studies generally assume linear relationships between social performance and legitimacy pressure on the one side, and sustainability reporting on the other. As recent research on CSR and sustainability in general, and sustainability reporting in particular suggests, these relationships may be more complex. For example, the relationship between legitimacy pressure (as a function of size and visibility) and sustainability engagement has been argued to be nonlinear (Udayasankar, 2008).

Third, prior studies have not yet considered the interplay of CSP and legitimacy pressure in explaining voluntary sustainability reporting. As many of the prior studies have focused on either CSP or legitimacy pressure as the most important determinant of sustainability reporting (cf. Clarkson, Li, Richardson, & Vasvari, 2008, p. 308), they were not designed to explore how both drivers might actually be valid, albeit for different companies in different contexts.

Fourth, legitimacy pressure has generally been specified and operationalized as a function of visibility and firm size. Research on CSR and sustainability, however, has shown that these two factors are not only conceptually different but may also have distinct effects on a company’s CSP (Brammer & Millington, 2006; Gallo & Christensen, 2011). Based on these arguments, this study will offer a differentiated analysis of the distinct effects of size and visibility on sustainability reporting.

Building upon prior research, this study takes these open questions as a point of departure and seeks to provide more nuanced explanations for why certain companies tend to extensively report on their sustainability performance, while others do not. This approach rests on the notion that, depending on specific company and environmental characteristics, different companies choose different disclosure strategies. Taking these issues into consideration, this study posits that some companies report on their CSP to signal superior performance to interested stakeholders, notably shareholders and potential investors. Other companies use sustainability reports as a response to external pressures, mostly a function of organizational visibility. Finally, large companies can simply afford sustainability reports more easily and have higher benefits to reap from such reports. In addition, it will be argued that company size and visibility do not only affect the level of sustainability reporting but also the strength and form of the CSP/sustainability reporting link.

To test these hypotheses, this study analyzes sustainability reports of 280 publicly traded companies from industries with a high social or environmental impact. To assess a given company’s sustainability disclosure level, the authors developed a content analysis index based on the Global Reporting Initiative guidelines (Global Reporting Initiative, 2006) and coded sustainability reports.

The remainder of the article is structured as follows. The next section develops the hypotheses on the distinct and joint effects of CSP, firm size, and organizational visibility on sustainability reporting. The “Method” section explains the data used and the methods applied. The results are presented in the “Results” section and their implications will be discussed in the final section.

Conceptual Background and Hypotheses Development

Drawing on Wood’s (1991, p. 693) model of CSP, sustainability reporting is defined as the systematic disclosure of information concerning a business organization’s principles of social responsibility, processes of social responsiveness, and outcomes as they relate to the firm’s societal relationships.⁵ Such disclosure can take place in the form of specific sections in annual reports, corporate websites, or stand-alone sustainability reports (KPMG Global Sustainability Services, 2008). The constant rise in the number of such reports and the significant costs involved in collecting, analyzing, and reporting data on social performance raise the question of why firms voluntarily disclose such information.

An economic perspective will frame the discussion of alternative suggestions brought forward to address this question. That is, the section’s arguments are based on the theoretical assumption that companies have economic reasons to engage in sustainability (Hart & Ahuja, 1996). Depending on their firm- and context-specific costs and benefits, companies choose the optimal level of sustainability performance (McWilliams & Siegel, 2001). This logic also applies to the level of voluntary sustainability reporting, which this study considers to be a function of economic calculus: Companies disclose proprietary information on their social performance to the extent to which they expect a benefit from such disclosures (Dawkins & Fraas, 2013).

The CSP/Disclosure-Link

A large body of conceptual and empirical research suggests that voluntary sustainability reporting is a function of a firm’s social performance (Dawkins & Fraas, 2013). One argument that has been put forward mainly from an accounting theory perspective is that sustainability reports are a means of reducing informational asymmetry by providing a signal to external observers (Clarkson et al., 2008). As Godfrey, Merrill, and Hansen (2009, p. 428) argue, any socially responsible activity can become effective only if it is public knowledge through either self-disclosure or reports of others. Because stakeholders, like investors or customers, are interested in identifying “high CSP” companies (Barnett & Salomon, 2006; Dhaliwal, Radhakrishnan, Tsang, & Yang, 2012; Sen & Bhattacharya, 2001), these companies have incentives to effectively communicate their superior social performance. In this case, sustainability reporting is a means for high CSP companies to differentiate themselves from their low CSP counterparts. Such differentiation strategies can come along with competitive advantages. For instance, the release of sustainability reports has been shown to decrease firms’ cost of equity (Dhaliwal et al., 2011).⁶

Note that this argument rests on two assumptions. First, it assumes that sustainability reports are important to external stakeholders because they are observable at low cost, whereas actual performance itself is not. Thus, observers who cannot directly assess a given company’s social performance may depend on the disclosure of proprietary information from that company (TriplePundit, 2010). Second, it assumes that sustainability reporting “cannot be easily mimicked” (Clarkson et al., 2008, p. 308) by “low CSP” companies so that more extensive sustainability reporting would not only increase transparency about inferior social performance but rational observers would interpret nondisclosures as “bad news” (Verrecchia, 2001).

Analogously to earlier studies on environmental reporting (Aerts & Cormier, 2009; Clarkson et al., 2008; Dawkins & Fraas, 2011a, 2013; Dhaliwal et al., 2011), this study expects higher levels of CSP to come along with higher levels of voluntary sustainability reporting. However, prior research will be qualified and extended by analyzing the CSP/sustainability reporting link contingent upon a firm’s CSP level: It will be argued that the positive relationship between CSP and sustainability reporting only persists to a certain degree. When companies reach higher CSP levels, they will find it increasingly difficult to achieve additional differentiation through their sustainability reports, implying a nonlinear relationship between CSP and sustainability reporting.

The benefits of additional voluntary disclosures should increase for low CSP companies because they meet the demand for information by interested stakeholder groups such as investors and professional lobbying groups who are interested in identifying superior social performers. However, increases in benefits will flatten because just as the benefits of increases in CSP become lower at high levels of CSP (H. Wang, Choi, & Li, 2008), high CSP companies have less opportunities for further differentiation within an industry. Stakeholders are often interested in a company’s relative social performance compared with peer companies within a given industry, which is why professional rating agencies typically apply so-called best-in-class approaches (White, 2005). With increasing performance differences relative to other industry members, there will be less scope (and less need) for further differentiation, and the demand for additional information by investors and other stakeholders will decrease. As information on improvements on low CSP levels will be more interesting for investors and other stakeholders than improvements on very high CSP levels, a company’s “cost-benefit trade-off assessments” (Cormier, Magnan, & Van Velthoven, 2005, p. 8) of voluntary disclosure will depend on its CSP level. Eventually, companies with very high CSP levels may even report less extensively on their sustainability performance than those with high CSP levels. In consequence, the marginal benefits of sustainability reporting will decrease as the CSP level increases and eventually turn negative, leading to the following hypothesis:

Hypothesis 1 (H1): There is a concave relationship between a firm’s social performance and its sustainability reporting level.

The Differential Roles of Firm Size and Organizational Visibility

If the only purpose of sustainability reports were to signal superior CSP, we would not observe the detailed sustainability reports voluntarily issued by “low CSP” companies. There are, however, additional reasons to disclose. Taking the perspective of legitimacy theory, some earlier studies were inspired by the argument that firms may voluntarily disclose information on their sustainability initiatives to hedge reputational risks and to prevent or to react to attacks from powerful stakeholder groups, such as nongovernmental organizations (NGOs), customer pressure groups, and the media (Bansal & Clelland, 2004; Chatterji & Toffel, 2010). In that case, sustainability reporting is simply a response to increased legitimacy pressure and can benefit the firm by serving as an impression management tactic that is shaping the perception of stakeholders (Bansal & Roth, 2000; Cho & Patten, 2007).

While many corresponding empirical studies have operationalized legitimacy pressure unidimensionally (either firm size or firm visibility), the current study posits that legitimacy-related motivations for sustainability reporting are more complex and that their effects should be separately accounted for when explaining sustainability reporting. Specifically, prior studies show how company size and organizational visibility distinctly affect sustainability performance in general, and reporting in particular (Brammer & Millington, 2006; Udayasankar, 2008; Ullman, 1985). Both factors are positively correlated and jointly affect sustainability reporting, albeit through different mechanisms. The next section begins by hypothesizing on the effects of firm size on sustainability reporting before it goes on with an analysis of the distinct effects of organizational visibility.

A number of arguments in the literature suggest that firm size can affect sustainability reporting independently of visibility (i.e., holding visibility constant). This notion is partly based on the argument that larger firms can afford sustainability expenditures more easily. Large firms enjoy higher levels of resource availability (Brammer & Millington, 2006) and greater resource-slack (Udayasankar, 2008). These advantages make it more likely for large firms to adopt socially desired practices such as sustainability reporting (Gallo & Christensen, 2011).

A second, though related, argument is based on the observation that the relative costs of sustainability reports are lower for larger firms than they are for smaller ones. Larger firms can benefit from more economies of scale in their sustainability performance (Brammer & Millington, 2006). In a similar vein, Baumann-Pauly, Wickert, Spence, and Scherer (2013) argue that, due to functional differentiation, specialization, and decentralization (Damanpour, 1987; Moch, 1976) larger firms have more specialized staff, more evolved administrative processes, and have more sophisticated internal systems to deal with business issues, resulting in lower relative (sustainability) reporting cost and higher likelihood to adopt new behaviors for large firms (Damanpour, 1996).

With increasing firm size, however, the marginal utility of additional sustainability reporting diminishes. First, very large companies are usually more powerful (Meznar & Nigh, 1995), can more easily resist external pressures, and are therefore less socially responsive (Brammer & Millington, 2006). Second, very large firms have already access to the resources they need (Udayasankar, 2008). The largest firms will thus feel less need to report more extensively on their CSP. The following hypothesis reflects these arguments:

Hypothesis 2a (H2a): There is a concave relationship between firm size and the sustainability reporting level.

Independent of a firm’s size, its visibility in the public has been identified as another important driver of sustainability performance. As has been argued from a legitimacy-based perspective, highly visible firms are increasingly exposed to pressure from various stakeholders to gain or secure their legitimacy by engaging in sustainability (Dawkins & Fraas, 2011b; Hooghiemstra, 2000; Udayasankar, 2008). The reason is that highly visible companies are more exposed to public scrutiny, which reflects the public attention the firm receives by external stakeholders, including the press, independent NGOs, and social movement organizations (Campbell, 2007). For example, Brammer and Millington (2006) have shown in their study on determinants of charitable donations that visibility is associated with higher levels of scrutiny from various stakeholders and thus drives the willingness to engage in sustainability. As the authors conclude, “visibility may therefore generate a general propensity for organizations to be more highly sensitive to social and political stakeholders” (p. 8). One way to achieve this is to follow the increasing demand for transparency and disclose more information on the company’s social and environmental performance (Aerts & Cormier, 2009; Reid & Toffel, 2009). Higher firm visibility will thus come along with higher levels of sustainability reporting.

To assume this relationship to be linear would, however, mean to neglect the special case of less visible firms. Firms with very low visibility may extensively report on an improved sustainability performance to attract media attention and increase their visibility (Udayasankar, 2008). As expressed in the following hypothesis, there should be a U-shaped form relationship between organizational visibility and sustainability reporting:

Hypothesis 2b (H2b): There is a convex relationship between organizational visibility and sustainability reporting levels.

In summary, this study hypothesizes two positively correlated functions (because of the correlation between visibility and size) with different shapes, though. This distinction allows us to differentiate the various effects outlined above, as summarized and visualized in Figure 1.

Figure 1.

Hypothesized relationships between firm size/visibility and sustainability reporting.

The Interplay Between CSP, Firm Size, and Organizational Visibility

After the discussion of the hypothesized distinct effects of CSP, firm size, and visibility on sustainability reporting, let us now turn to their joint effects. The hypothesis on the positive and concave relationship between CSP and sustainability reporting was based on the assumption that the benefits of disclosing information on its social performance increases with better social performance: The better a firm performs with respect to sustainability, the more it wants to talk about it (with decreasing marginal benefits).

Concerning firm size, it was argued that larger firms can more easily afford sustainability reporting because of greater resource availability, lower relative costs, and more specialized administrative structures and processes (Brammer & Millington, 2006; Donaldson, 2001; Udayasankar, 2008). These relative advantages imply that larger firms’ sustainability reporting efforts are potentially more sensitive to expected benefits induced by revealing CSP improvements to external stakeholders. Conversely, managers of smaller firms face much more limited resources and higher opportunity costs, that is, they can invest in sustainability reporting only at the expense of other potentially more viable activities. Generally, managers of small firms are more risk averse because misallocation of sparse resources can endanger the survival of the firm, whereas large firms benefit from a buffering effect due to a large stock of resources (Audia & Greve, 2006). To reduce uncertainty, small firms might therefore demand higher expected benefits from the same level of investments in sustainability reporting compared with a larger firm. Hence, the sensitivity of sustainability reporting efforts to CSP improvements is extenuated: CSP improvements and hence the expected benefits from voluntarily disclosing CSP (i.e., the “project success”) have to be much larger for smaller firms to show the same level of sustainability reporting as compared with larger firms that can afford risk taking behavior as they have more leeway to bear the losses of a less successful project (Damanpour, 1996).

As some authors have argued from a legitimacy theory perspective, sustainability reporting can be seen as a function of external pressure, which companies try to counteract by issuing sustainability reports (Cho & Patten, 2007; Cormier et al., 2005). This argument may especially apply to low-performing companies and their need to respond to increased legitimacy pressure because associated reputational risks, such as alienating buyers and investors, are amplified. Low-CSP firms may want to “separate [potentially] threatening revelations from larger assessments of the organization as a whole” (Suchman, 1995, p. 598). This perspective has served as an explanation for a potential negative relation between sustainability reporting and actual sustainability performance (Patten, 2002). Given the analysis to this point, the argument can be used to explain how CSP and visibility interact in driving sustainability reporting. If, as argued above, legitimacy pressure is a function of visibility, the incentives to report more extensively on superior CSP will be highest for very visible firms, and less visible firms experience less benefits to report on performance improvements.

The pressure to report superior CSP to the public is higher for more visible firms as these are most interested in shaping the public’s impression on their CSP and to gain in legitimacy (Cormier et al., 2005). By expressing commitment to the environment and society at large by issuing sustainability-related information about good sustainability performance may repel a few skeptical stakeholders and reinforce the opinion of others (Bansal & Clelland, 2004). Consequently, more visible firms benefit much more from disclosing improved CSP than less visible firms. This argument implies that high visible firms’ investments in sustainability reporting are more sensitive to CSP improvements.

The following two hypotheses capture the arguments that both company size and visibility will moderate the CSP/sustainability reporting link:

Hypothesis 3a (H3a): If companies are large (small), the relationship between CSP and the level of sustainability reporting will be stronger (weaker).

Hypothesis 3b (H3b): For firms with high (low) organizational visibility, the relationship between CSP and the level of sustainability reporting will be stronger (weaker).

Method

Models and Sample Selection

To test the hypotheses, the empirical model regresses a sustainability reporting variable on a set of explanatory variables capturing CSP, firm size, organizational visibility, and their interactions. It also controls for additional variables that are known to affect sustainability reporting. The following equation captures the model:

\begin{array}{l} S U S T R E P = β_{0} + β_{1} C S P + β_{2} C S P^{2} + β_{3} S I Z E + β_{4} S I Z E^{2} + β_{5} V I S + β_{6} V I S^{2} + β_{7} C S P \times S I Z E \\ + β_{8} C S P^{2} \times S I Z E + β_{9} C S P \times V I S + β_{10} C S P^{2} \times V I S + x β + u \end{array}

where SUSTREP is a measure of the company’s sustainability reporting level; CSP is a measure of the company’s social performance level; SIZE is a measure of firm size; VIS is a measure of organizational visibility; x is a vector of control variables; and u is the error term.

To construct the data set, the authors collected data from multiple sources. Investors belong to the most important addressees of sustainability reports (Hummels & Timmer, 2004), so the sample should only include firms for which shareholders play an important role. The analysis was thus restricted to publicly traded companies. To maintain comparability in terms of the relevance of sustainability, it was further restricted to industries with a high social and environmental impact. Earlier studies with a focus on environmental reporting restricted their analysis to polluting industries (Cho & Patten, 2007; Clarkson et al., 2008). As this study is interested in the broader concept of CSP (i.e., social and environmental issues), it includes all industries that oekom research, an international sustainability rating agency classified as either environmentally or socially sensitive industries (a detailed explanation will follow below). As listed in Table 1, the final sample contains cross-industry data on 280 publicly traded companies for which data on CSP and all other relevant (control) variables were available for the year 2009.

Table 1.

Sample Distribution Across Industries.

Distribution of observations across industries
Industry	Frequency	%	Cumulative
Automobile	30	10.7	10.7
Chemicals	22	7.9	18.6
Construction	30	10.7	29.3
Food & beverages	18	6.4	35.7
Household & personal products	19	6.8	42.5
Machinery	35	12.5	55.0
Metals & mining	26	9.3	64.3
Oil & gas	21	7.5	71.8
Paper & forest products	10	3.6	75.4
Pharmaceuticals & biotechnology	20	7.1	82.5
Transport & logistics	21	7.5	90.0
Utilities	28	10.0	100.0
Total	280	100

Dependent Variable: Sustainability Reporting Level

To measure a company’s sustainability reporting level (as opposed to its CSP level) the study employs a quantitative content analysis of sustainability reports covering the year 2009. The year 2009 was the latest year for which all potential reports were available when the coding process began in early 2012.⁷ To qualify as a “sustainability report,” information had to be disclosed in a specifically devoted company document. This could be via stand-alone reports, a specific website, or a particular section in the annual report. Only those reports were included that had a specified reporting period covering the year 2009. Thus, websites that were continually updated were not included in the sample, as they did not allow for an evaluation of the company’s reporting practice for the year of 2009. To identify sustainability reports, the authors searched each company’s website. If no sustainability report was found, they additionally contacted the company via email to ask whether a report was available.

This study’s research question required measurement of the extent of information disclosed in a firm’s sustainability report. Therefore, simply observing whether or not a company publishes a sustainability report at all (Dhaliwal et al., 2011), was not enough for this study’s purposes. It thus followed other approaches (Aerts & Cormier, 2009; Branco & Rodrigues, 2008; Clarkson et al., 2008) that analyzed whether or not a company’s sustainability report includes information on certain sustainability issues. To obtain a catalog of such issues, it followed Clarkson et al. (2008) and used the GRI reporting guidelines to construct a quantitative disclosure index. However, in line with the goal to measure sustainability comprehensively, the measurement instrument was extended beyond environmental issues and also included social aspects.

GRI is a multistakeholder initiative that develops sustainability reporting guidelines. The guideline’s version G3 was published in 2006 and provides detailed principles and indicators for companies to report on their CSP (Global Reporting Initiative, 2006). As GRI is the most commonly used sustainability reporting guideline, it provides an adequate benchmark to assess a given company’s sustainability reporting level.

GRI G3 defines reporting principles and a list of reporting items defined as standard disclosures. In line with this study’s aim to assess quantitative differences in sustainability reporting, the index is entirely based on the standard disclosure items. These are divided into the following three categories: strategy and profile, management, and performance indicators (Global Reporting Initiative, 2006, p. 20). Categories contain “aspects” that cover “core” and “additional” items. In the performance indicators category, one core reporting item was taken for each aspect, resulting in 20 environmental and social performance indicators and six economic performance indicators. Economic performance indicators were also included as economic responsibility is commonly included in CSP definitions (Carroll, 1979). In the strategy and profile and the management categories, GRI does not provide core items. In the absence of comparable studies published in academic journals, the authors followed the approach of SustainAbility, UNEP, and Standard & Poor’s (2006) to construct a list of items in these two categories. Table 2 provides an overview of the items included in the index.

Table 2.

Items Used to Assess the Extent of Sustainability Reporting.

Strategy and profile (0/1)

Top management statement

Sustainable development vision and/or strategy

Core values, principles, or policies to sustainable development

Sustainable development implementation challenges

The business case: Rationale for pursuing sustainable development

Company and industry profile

Issue identification/prioritization

Stakeholder identification/list of main stakeholder

Description of stakeholder engagement process/involvement

Management approach (0/1)

Governance structure and responsibilities/CSR committee

Influence on/support of industry standards (for sustainable development)

Customer influence (regarding sustainable development issues)

Sustainability in the value chain

General sustainability management systems

Environmental management systems (e.g., ISO 14001)

Performance indicators (0/1/2)

Economic performance indicators

Economic goals/targets

Economic value generated and distributed

Employee compensation

Investments in research and development (R&D)

Environmental spending

Social investments/contributions

Environmental performance indicators

Environmental goals/targets

Materials use/inputs

Energy use or energy efficiency

Water use or water efficiency

Nature conservancy, land use and/or biodiversity

Gas emissions

Waste generation or management

Environmental impact of products and services

Fines and penalties related to environmental issues

Transportation

Social performance indicators

Social goals/targets

Workforce profile

People management

Diversity/nondiscrimination

Training and education

Health and safety

Human rights

Community engagement

Corruption

Public policy (lobbying)

Note. This table lists the items included in the GRI G3-based scoring tool used in this study to assess the extent of sustainability reporting. An item in the strategy & profile and the management categories was assigned a value of 1 if the reports contained information on the items and 0 if they did not. The same procedure was used for the performance items with the modification that a value of 2 was assigned to any item for which the report contained quantitative information (i.e., at least three numbers). CSR = corporate social responsibility; GRI = global reporting initiative.

An item in the strategy and profile and the management categories was assigned a value of 1 if the reports contained information on the items and 0 if they did not. The same procedure was used for the performance items with the modification that a value of 2 was assigned to any item for which the report contained quantitative information.⁸ This methodology allowed for a reasonable differentiation of reporting levels (in an effort to maximize validity) while reducing method bias to a minimum (to maximize reliability). Following Clarkson et al. (2008, p. 316) again, a total index score of 0 was assigned to firms that did not have a sustainability report.

A group of six student research assistants rated the reports over a period of 3 months. Prior to participating in this exercise, all raters had to partake in a training session and were instructed to discuss questions openly in the group to ensure that everybody applied the same methodology. After the training session and before the rating began, two raters rated one industry (Oil & Gas) independently to verify that the scoring instrument was reliable. Two common measures of intercoder reliability, Cohen’s kappa (κ = .9140) and Spearman’s rank correlation coefficient (R = .9917) suggested a sufficiently high reliability of the instrument (Lombard, Snyder-Duch, & Bracken, 2002).

Independent Variables in Hypotheses

To measure a company’s social performance level (as opposed to its sustainability reporting level), CSP data were obtained from oekom research AG, which uses a comprehensive indicator-based approach to evaluate the social and environmental performance of publicly traded companies. Compared with other databases, oekom research’s data include certain advantages, primarily because their ratings are quasimetric and industry-specific and thus do not require the specific construction of indices (such as aggregating and weighting strengths and weaknesses; cf. Schreck, 2011). The final rating scale ranges from A+ (excellent record) to D– (poor record); this range corresponds to a 12-step numerical scale. The information used in the evaluation process is derived from both the companies themselves (mainly questionnaires) and various independent external sources. oekom research’s database includes more than 1,000 companies from various countries and all major industries; the stocks analyzed cover international indices, such as Dow Jones STOXX 600 and the MSCI World Index (by Morgan Stanley Capital International). To warrant comparability with respect to CSP, the analysis was restricted to companies from industries classified as high social or environmental impact industries by oekom research.

Firm size was measured by (the log of) the number of employees in 2009. Organizational visibility was approximated by (the log of) media coverage in 2009. In line with this study’s goal to obtain a measure that was independent of CSP, the authors followed the common approach (Cormier et al., 2005; Dawkins & Fraas, 2011a, 2011b) to measure media coverage by (the log of) the number of news articles written about a specific firm in a given year (in this case, 2009). All data were obtained from Worldscope, except media coverage data, which were retrieved from the LexisNexis database.

Control Variables

To avoid misspecification problems, the empirical model controls for various variables at the company and the country level that have been argued to affect sustainability reporting levels in the literature on the determinants of environmental reporting. It includes Tobin’s Q and stock price volatility (Clarkson et al., 2008; Dhaliwal et al., 2011) as measures of information asymmetry based on the argument that companies voluntarily disclose sustainability-related information to reduce information asymmetries. Stock price volatility is measured as the maximum deviation from average stock price in 2009.

Tobin’s Q is defined as the ratio of the market value of assets to their replacement value at the end of fiscal year 2009. The model uses the well-established approach developed by Chung and Pruitt (1994) and approximates the market value of assets by the book values of assets (TA) minus the book value of equity (CE) minus deferred taxes (DefTax) plus the market value of common stocks (MV). The replacement value of assets is approximated by the book value of assets (TA).

Q = \frac{T A - C E - D e f T a x + M V}{T A}

The model further controls for a firm’s financial performance based on the argument that financially successful companies will be more likely to voluntarily disclose proprietary information regarding their CSP (Clarkson et al., 2008; Cormier et al., 2005). Specifically, it includes the variables return on equity and leverage. Return on equity is measured as net income in fiscal year 2009 divided by the common equity at the end of fiscal year 2008; leverage as the debt-to-equity ratio in fiscal year 2009 (Cormier & Gordon, 2001).

The cost of equity capital has also been shown to be a relevant driver of voluntary sustainability reporting (Dhaliwal et al., 2011), so the regression model includes that variable. The procedure follows Sharfman & Fernando (2008) who, based on the capital asset pricing model (Lintner, 1965; Sharpe, 1964), estimate a company’s cost of equity capital as $r_{E} = r_{F} + β_{E} (r_{M} - r_{F})$ , where r_F = the risk-free rate; r_M = the return on the market portfolio; and β_E = the firm’s systematic risk.

The r_F is calculated as the average risk-free rate from 1960 to 2009 and r_M-r_F as the average market premium on the market portfolio in the same period; both values were retrieved from the French data library. Beta is calculated by regressing (dividend adjusted) company stock returns on market returns (MSCI World Index) on a daily basis between January 1, 2009, and December 31, 2009.

The model also controls for the percentage of concentrated ownership as an approximation of the share of strategic investors because their specific interest in sustainability-related information might affect its supply (Cormier et al., 2005). To control for industry fixed effects on sustainability reporting, the regression analyses include industry dummy variables. Moreover, all independent variables are centered on industry means to allow for cross-industry comparisons of variance in the independent variables. Finally, to control for country differences in sustainability legislation, this study follows the procedure suggested by Dhaliwal et al. (2012). In their empirical study on sustainability reporting, the authors distinguish three groups of countries according to national sustainability reporting legislation (for a list of countries and legislation differences, compare Dhaliwal et al., 2012, Appendix A). All financial and market data were obtained from Worldscope and Thomson Reuters Datastream.

Results

Hypothesis Tests

Table 3 provides descriptive summary statistics and Pearson correlations for all variables included in the model. As the maximum score of the sustainability reporting variable indicates, at least one sustainability report achieved the highest rating. In contrast, the maximum value of the CSP variable is 9 showing that oekom research rated no company better than that grade (i.e., the equivalent of “B+”). Correlation coefficients indicate significant associations between voluntary sustainability reporting and CSP (.452, p < .001), firm size (.358, p < .001), and firm visibility (.194, p < .01). Furthermore, the correlation coefficient between firm size and visibility is significantly positive but its comparably low magnitude shows that both variables do not measure the same concept (.529, p < .001).

Table 3.

Summary Statistics and Pairwise Correlations.

Model variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
1. Sustainability reporting	1.000
2. CSP	.452	1.000
3. Firm size	.358	.289	1.000
4. Firm visibility	.194	.338	.529	1.000
5. Leverage	−.057	−.014	.068	.046	1.000
6. Return on equity	.059	.025	.132	.087	.099	1.000
7. Cost of equity	.026	.125	.227	.371	.046	−.084	1.000
8. % of strategic investors	−.039	−.115	−.300	−.160	.031	−.088	.058	1.000
9. Tobin’s Q	−.003	.120	−.049	.028	−.065	.440	−.059	.016	1.000
10. Share price volatility	−.042	−.154	−.147	−.067	−.105	−.124	.304	.081	−.129	1.000
M	36.33	5.43	10.37	5.77	1.01	8.51	0.11	23.74	1.45	28.01
SD	20.38	1.56	0.64	1.63	1.62	18.91	0.03	24.29	0.66	8.65
Minimum	0.00	1.00	9.60	0.00	−17.02	−98.65	0.05	0.00	0.76	13.66
Maximum	64.00	9.00	12.62	9.21	7.97	105.68	0.19	99.00	5.42	59.42

Note. This table provides descriptive summary statistics and Pearson correlations for all variables included in the model. Bold numbers mark correlations that are significant at the 10% level or below. CSP = corporate social performance.

Table 4 presents estimates of the various models tested. The true regression coefficients in the population are estimated by ordinary least squares (OLS) while using heteroscedasticity–robust standard errors (White, 1980). To control for industry effects, all independent variables are standardized by centering values on industry means, and industry fixed effects are added to each equation. Multicollinearity does not present a problem to the analyses, as the variance inflation factor (VIF) scores are below 10 (Hair, Black, Babin, Anderson, & Tatham, 2006, p. 230).

Table 4.

Regression Results.

Dependent variable: Sustainability reporting	Hypothesis (expected sign)	Model
Dependent variable: Sustainability reporting		(1)	(2)	(3)	(4)	(5)	(6)	(7)
Independent variables		Controls only	Linear terms of focal covariates	(2) + quadratic terms of covariates	(3) + interaction terms	(1) + (2)	(1) + (3)	(1) + (4)
CSP			5.280*** (0.845)	4.599*** (0.836)	4.585*** (0.945)	5.474*** (0.825)	4.822*** (0.849)	4.767*** (0.958)
CSP ²	H1 (−)			−0.956** (0.342)	−1.391** (0.439)		−0.759* (0.372)	−1.187* (0.476)
Firm size			3.862*** (1.044)	2.048^† (1.051)	2.807* (1.174)	4.526*** (1.077)	2.709** (1.039)	3.542** (1.166)
Firm size ²	H2a (−)			−1.144** (0.409)	−1.293** (0.410)		−1.203** (0.416)	−1.337** (0.415)
Firm visibility			−0.853 (0.865)	0.129 (0.833)	−0.555 (0.981)	−0.345 (0.895)	0.644 (0.911)	−0.136 (1.058)
Firm visibility ²	H2b (+)			0.496^† (0.263)	0.457 (0.280)		0.535* (0.248)	0.505^† (0.275)
CSP × Firm size	H3a (+)				0.517 (0.671)			0.175 (0.663)
CSP² × Firm size					−0.566* (0.239)			−0.668** (0.255)
CSP × Firm visibility	H3b (+)				−0.109 (0.646)			−0.047 (0.655)
CSP² × Firm visibility					0.512^† (0.269)			0.573* (0.279)
Leverage		−1.018 (0.662)				−0.929 (0.671)	−0.895 (0.623)	−0.860 (0.661)
Return on equity		0.029 (0.081)				0.032 (0.081)	0.019 (0.074)	0.026 (0.074)
Cost of equity		64.233 (46.119)				−65.808 (45.945)	−62.795 (44.866)	−58.423 (45.521)
% of strategic investors		0.059 (0.057)				0.128* (0.052)	0.124* (0.054)	0.125* (0.054)
Tobin’s Q		−2.396 (2.022)				−2.243 (1.793)	−1.718 (1.737)	−2.472 (1.823)
Share price volatility		−0.178 (0.210)				0.225 (0.191)	0.190 (0.187)	0.179 (0.189)
Industry dummies		Yes	Yes	Yes	Yes	Yes	Yes	Yes
Country group dummies		Yes	Yes	Yes	Yes	Yes	Yes	Yes
N		280	280	280	280	280	280	280
F		3.750***	12.700***	17.333***	21.616***	11.830***	14.850***	19.984***
R ²		.175	.352	.388	.395	.383	.414	.422
Adjusted R²		.098	.299	.330	.328	.317	.343	.341
Maximum VIF		1.860	1.850	2.140	6.350	1.860	2.300	6.780

Note. This table presents OLS estimates of different models using a sample of 280 companies in environmentally and socially sensitive industries. Model 1 is the baseline model with controls only. Model 2 tests the main effects of CSP, firm size, and firm visibility on voluntary sustainability reporting. Models 3 to 7 test our studies’ hypotheses. All independent variables are centered on industry means. Heteroscedasticity–robust standard errors are reported in parentheses. CSP = corporate social performance; VIF = variance inflation factor; OLS = ordinary least squares.

Individual coefficients are statistically significant at the ^†10%, *5%, **1%, or ***0.1% level (two-tailed t test).

Different models are presented to test the hypotheses and document the robustness of the results. Model 1 includes control variables only and has a low model fit (adjusted R² = .098). The three focal covariates, CSP, firm size, and firm visibility, explain much more variance in sustainability reporting (see Model 2, Table 4: adjusted R² = .299). Models 3 to 7 combine control variables, focal covariates with linear, quadratic, and interaction terms. The fully specified Model 7 shows a comparably good fit (adjusted R² = .341). The following paragraphs present the study’s results in detail.

Nonlinear relationship between CSP and sustainability reporting levels

In line with prior studies, this study finds a positive relationship between CSP and the levels of sustainability reporting. The coefficient of the linear term of CSP is significantly larger than zero (5.280; p < .001; see Model 2 in Table 4). H1 argues however that due to a diminishing marginal utility the level of sustainability reporting grows slower as the CSP level increases. To test H1, the quadratic term CSP² was added, yielding a significant negative coefficient (−.956; p < .01; see Model 3 in Table 4). This effect is robust to various model specifications (cf. Models 4, 6, and 7 in Table 4). The first graph in Figure 2 visualizes this finding by plotting the predicted sustainability reporting values stemming from a simple model with the linear and quadratic terms of CSP as predictors (F = 47.960; p < .001; adjusted R² = .252; coefficient of quadratic term is −1.554 with p < .001). This graph shows that the positive slope measuring the strength of association between CSP and sustainability reporting decreases with increasing levels of CSP. In sum, the findings lend empirical support for H1.

Figure 2.

Estimated relationships between CSP, firm size/visibility, and sustainability reporting.

Nonlinear relationship between firm size/visibility and sustainability reporting levels

Comparable with the CSP/sustainability reporting relationship, the analysis finds a positive relationship between firm size and visibility on one hand, and sustainability reporting on the other hand. However, both effects show differences in magnitude and significance. While the correlation coefficients of firm size (.358, p < .001) and firm visibility (.194, p < .01) are positive and significant at a 1% significance level, the correlation coefficient of firm visibility is substantially lower. The coefficient of the linear term of firm size is stable across models at a 10% significance level (see Table 4), whereas the linear term of firm visibility does not have a significant coefficient (p > .10).

H2a assumes a nonlinear relationship between firm size and sustainability reporting. In line with this hypothesis, the quadratic term of firm size is significantly smaller than zero (−1.144, p < .01; see Model 3 in Table 4). This effect is stable across models (cf. Models 4, 6, and 7 in Table 4). The second graph in Figure 2 visualizes this finding (predicted values stem from a model with linear and quadratic terms of firm size; F = 25.630, p < .001; adjusted R² = .150; coefficient of quadratic term is −1.370 with p < .001). The strength of association between firm size and sustainability reporting decreases with increasing levels of firm size. This empirical evidence is in line with the theoretical reasoning outlined above (cf. with Figure 1). In sum, the findings lend empirical support for H2a.

H2b, which assumes a nonlinear relationship between firm visibility and sustainability reporting, finds only partial empirical support. The second graph in Figure 2 shows that the sample produces a nonlinear relationship (U-shape) between firm visibility and sustainability reporting comparable with the predictions that arise from the earlier theoretical reasoning (see Figure 1). The coefficient of the quadratic term is positive and significantly different from zero at a 10% significance level in the majority of the models (see Models 3, 6, and 7 in Table 4). Hence, there is (weak) empirical support for H2b.

Interplay between CSP and firm size/visibility

H3a and H3b assume that the relationship between CSP and sustainability reporting becomes stronger if either firm size is large or firm visibility is high. There is no empirical support for both hypotheses because the coefficients of interaction terms CSP × Firm size and CSP × Firm visibility are not significantly different from zero (p > .10; see Models 4 and 7 in Table 4).

However, there is empirical support for a negative interaction term considering the quadratic term of CSP and the linear term of firm size (−.668 with p < .01; see Model 7 in Table 4). Furthermore, there is a positive interaction term considering the quadratic term of CSP and the linear term of firm visibility (.573 with p < .05; see Model 7 in Table 4). Plotting the predicted values from Model 7 in Table 4, Figure 3 visualizes both interaction terms. The relationship between CSP and sustainability reporting is almost linear for small firms (holding organizational visibility constant), while the relationship between CSP and sustainability reporting is curvilinear for large firms (first graph in Figure 3). Concerning firm visibility, effects are inverted. For low visible firms (holding firm size constant), the relationship between CSP and sustainability reporting is curvilinear, while for high visible firms the relationship between CSP and sustainability reporting is flatter (second graph in Figure 3). Although there is no empirical support for H3a and H3b, results indicate a three-way interaction between CSP and firm size/visibility. The relationship of CSP and sustainability reporting does not only depend on firm size (and firm visibility) but simultaneously depends on the firm’s CSP level.

Figure 3.

Interactions of CSP and firm size/visibility.

Supplementary Analyses

This study was motivated with the observation that external stakeholders use sustainability reports as a clue to assess the company’s actual CSP and that it is thus important to understand the drivers of sustainability reporting. The analysis revealed a positive correlation between CSP and sustainability reporting, suggesting that a high level of sustainability reporting is generally a signal of high CSP. However, earlier studies on the determinants of environmental reporting have found a negative relationship between environmental performance and reporting (Hughes, Anderson, & Golden, 2001; Patten, 2002), suggesting that, sometimes, it is exactly the low CSP companies that find it useful to publish extensive sustainability reports. This study lends no support for a generally negative CSP/sustainability reporting link but, as the imperfect positive correlation indicates, the sample also includes “low CSP” companies that extensively communicate on their CSP (even though the model controls for firm size and visibility effects).

This prompts the question why companies with an inferior CSP would disclose detailed information about their CSP at all. One potential answer is that such companies engage in “window dressing” and try to divert attention from their unfavorable performance, “rendering private SER a predominantly cosmetic, theatrical and empty exercise” (Solomon, Solomon, Joseph, & Norton, 2013, p. 195). An alternative answer would be that low CSP companies who are serious about change use - as a first step toward improved CSP - sustainability reports to increase transparency toward external stakeholders to “build or repair reputation” (Dawkins & Fraas, 2013, p. 245).

Which of these arguments is empirically more valid eventually requires an analysis of the quality of low CSP companies’ sustainability reports. In addition to the analysis of the main hypotheses, this part of the analysis thus explores two factors that may be indicative of the sustainability reports’ credibility: report assurance and CSP improvements. First, if low CSP companies use sustainability reports to signal strategy change, one way to add credibility to their reports would be to have them assured by third parties (O’Dwyer, Owen, & Unerman, 2011; Owen & O’Dwyer, 2008; Simnett et al., 2009). In this case, one would expect that low CSP companies with high levels of sustainability reporting have their reports assured as or even more often than their “high CSP” counterparts. Second, if sustainability reports really anticipate strategy change, one would expect CSP improvements over time for low CSP companies with high levels of sustainability reporting.

The authors examined these issues empirically by first identifying companies with a low CSP (bottom 50%) and a high sustainability reporting score (top 50%). They then analyzed how “low performance/high reporting” companies compared with their “high CSP counterparts” with respect to assurance levels. The share of assured reports with “high CSP/high reporting” firms is more than twice as high than with “low CSP/high reporting” companies (28% vs. 12%; t = −2.48, p < .01).

Furthermore, “low CSP/high reporting” companies’ CSP development over time was compared with that of “low CSP/low reporting” firms. Data show that reporting (low CSP) companies on average did not improve their CSP from 2009 to 2010 (difference of −.11 points; this change is not significantly different from 0, t = .8104; p > .10). In contrast, “low CSP/low reporting” companies did improve by 0.27 points on average during the same period (t = −1.4475, p < .10). Based on these results, the authors tentatively conclude that extensive reporting by low CSP companies cannot be taken as a signal for strategy change. In combination with the lower level of third-party assurance, “low CSP/high reporting” companies partially engage in “window dressing.”

Discussion and Conclusion

External stakeholders who wish to assess a firm’s CSP level may use the information voluntarily published by that firm as a low cost signal for its underlying, actual CSP (Clarkson et al., 2004; Richardson & Welker, 2001). Although a firm’s sustainability report may help stakeholders to identify a firm’s beliefs and attitudes toward sustainability, the long-term image of the firm and thus the stakeholders’ willingness to provide the firm with resources depends on whether this information corresponds to actual firm behavior (Donaldson & Preston, 1995; Jones, 1995). The credibility and value of self-reported sustainability-related information for external agents depends on how accurately this information measures the underlying CSP. By addressing several ambiguities and intricacies concerning theoretical arguments and empirical findings, this study contributes to the literature on sustainability that has acknowledged the importance of sustainability reports (Bansal & Clelland, 2004; T. Wang & Bansal, 2012).

High CSP firms have an incentive to reduce information asymmetry by disclosing more sustainability-related information, thereby gaining some competitive advantage over low CSP firms, which in turn may find it difficult to disclose sustainability-related information without revealing their poor performance.⁹ This study provides a more fine-grained view on this generally accepted view by providing evidence for a nonlinear relationship between CSP and sustainability reporting. A potential reason is that with increasing CSP, firms find it more and more difficult to differentiate themselves through higher sustainability reporting levels because sustainability reporting is a means of differentiation especially for those companies that have most opportunities to improve their CSP. The better a company’s social performance within an industry is, the smaller the benefits of additional reporting on improved CSP become. External stakeholders may therefore use very low levels of sustainability reporting as a reliable signal for low CSP firms, but they may find it increasingly difficult to identify best-in-class CSP firms by merely looking at voluntarily disclosed sustainability-related information.

Prior studies report legitimacy pressure to be another important driver of sustainability reporting levels (Aerts & Cormier, 2009; Hooghiemstra, 2000; Patten, 2002). While engagement in sustainability in response to stakeholder demands can enhance access to or limit loss of resources, underlying mechanisms and empirical evidence suffer from some ambiguity caused by the operationalization of legitimacy pressure. Frequently, legitimacy pressure is measured by firm size or firm visibility although both variables do not measure exactly the same concept. By distinguishing firm size and visibility, this study augments Udayasankar (2008), who theorizes about differential and nonlinear effects of various firm attributes (visibility, resources access, and scale of operations), with first empirical evidence in the context of sustainability reporting. Doing so, this study also extends Brammer and Millington’s (2006) study on corporate philanthropy, which is restricted to linear effects of size and visibility and does not consider differently shaped, nonlinear functions. The analysis also broadens the study of Gallo and Christensen (2011) who find a positive relationship between firm size and sustainability reporting but who do not consider firm visibility and nonlinear effects of firm size.

For both variables, firm size (strong empirical support) and firm visibility (weaker empirical support), results indicate a nonlinear relationship with sustainability reporting. The relationship between firm size and sustainability reporting describes an inverted U-shaped function, whereas for firm visibility this relationship follows a U-shape (see second graph in Figure 2). This finding implies that with increasing firm size, the relative cost-advantage may stimulate firms to engage more in sustainability (hoping for access to additional resources), while very large and already resource-rich firms (holding firm visibility constant) have less incentive to further increase sustainability reporting (Brammer & Millington, 2006; Udayasankar, 2008). Concerning firm visibility (holding firm size constant), both low visible firms, looking for additional visibility through sustainability reporting, and high visible firms, finding themselves under pressure from external stakeholders, show higher levels of sustainability reporting. This empirical result is consistent with theorizing in extant literature (Udayasankar, 2008).

Furthermore, the analysis finds two interesting interaction effects between firm size and organizational visibility on one hand, and the quadratic term of CSP on the other hand. For large firms, sustainability reporting is more sensitive to CSP improvements if the CSP level is below average, whereas this sensitivity decreases if CSP level is above average (see first graph in Figure 3). This finding is in line with the slack-resources and differentiation arguments (Brammer & Millington, 2006; Gallo & Christensen, 2011; Udayasankar, 2008). Large firms have a relative cost-advantage and might invest more likely in sustainability reporting if they believe to benefit from revealing their CSP improvements to external stakeholders. However, if CSP and the sustainability reporting levels are high, they will find it increasingly difficult to achieve additional benefits through their sustainability reporting. For small firms, the expected cost–benefit ratio might be different because of relative cost-disadvantages. Benefits from sustainability reporting have to be larger before smaller firms decide to invest in sustainability reporting, that is, additional investments in sustainability reporting appear to be less sensitive to CSP improvements.

The interaction effect for firm visibility with the quadratic term of CSP seems to be less straightforward. For low visible firms (holding firm size constant), sustainability reporting levels are more sensitive to CSP improvements if the CSP level is below average, whereas this sensitivity decreases if CSP level is above average (see second graph in Figure 3). The sensitivity of sustainability reporting to CSP improvements is almost constant for high visible firms. Less visible firms might seek visibility through sustainability initiatives to enhance access to resources (Udayasankar, 2008). Therefore, small CSP improvements might trigger more sustainability reporting but if CSP and sustainability reporting levels are already high, low visible firms might find it increasingly difficult to achieve additional differentiation and visibility through their sustainability reports. On the contrary, highly visible firms do not seek visibility trough sustainability reporting but, due to their high visibility, these firms face more legitimacy pressure by external stakeholders (Brammer & Millington, 2006). Assuming that this stakeholder demand is fairly constant, highly visible firms do more likely respond to this demand if CSP is improving (i.e., positive slope) while this sensitivity of reporting to CSP improvements is relatively independent from the actual CSP level (i.e., constant slope).

Results suggest that some firms may use voluntary sustainability reports as a form of symbolic legitimacy management and “window dressing.” This finding is consistent with an agency theory perspective, which suggests that self-interested managers sometimes have incentives to act against stakeholders’ interests (Jensen & Meckling, 1976). If external stakeholders, such as investors or the general public, do not adequately monitor managers, self-interested managers tend to conceal negative information because they want to protect their personal interests and career outlooks. This is an additional argument for the positive relationship between CSP and sustainability reporting. Even though corporate leaders are under pressure to reveal information, they might even be tempted to demonstrate conformity with prevailing ideologies by acting in a relatively symbolic rather than substantive matter (Westphal & Graebner, 2010). This view is also in line with the finding that “low CSP/high reporting” companies are less likely to substantiate their sustainability reports by third-party assurance. Thus, stakeholders are advised to be skeptical about the reliability of information that is disclosed without being externally verified.

This study argued and provides empirical evidence that companies have different motivations to disclose information on their CSP. This finding implies that sustainability reports are only an imperfect indicator about firms’ attitudes toward sustainability and their actual CSP. This result may emphasize the importance of independent and credible third-party evaluators, which provide potentially more reliable signals that truly reduce information asymmetry (Montiel et al., 2012). Recent research confirms that independent third-party ratings may even substantially change the behavior of firms (Chatterji & Toffel, 2010). While this implies that the publication of information to monitoring agents, such as investors and customers, may be sufficient for controlling a firm’s behavior, government agencies should put more emphasis on mandatory sustainability reporting or auditing programs to deter corporate leaders from opportunistic information disclosure practices.

Finally, this study offers avenues for further research. While its findings suggest that CSP level, firm size, and firm visibility are associated with sustainability reporting, the authors encourage replications of this study to substantiate the results presented. Furthermore, more factors could be explored that might be associated with sustainability reporting. Also, additional interaction effects might indicate a gap in the literature concerning theories about underlying moderation and mediation mechanisms that explain sustainability initiatives in general and sustainability reporting in particular. Hopefully, this study will inspire future research to investigate those mechanisms and contingency factors.

Footnotes

Acknowledgements

The authors are grateful for valuable comments on earlier versions of this manuscript from Sebastian Becker, Graeme Dean, Hans-Ulrich Küpper, participants of the IPC research seminar 2012 at Ludwig-Maximilians-University, the vhb/IAAER conference 2013, the Discipline of Accounting Research Seminar 2013 at the University of Sydney, and participants of the Academy of Management Annual Meeting 2013. They also thank Matthias Bönning of oekom research AG for his ongoing and generous support of this research.

The article was accepted during the editorship of Duane Windsor.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Philipp Schreck (PhD, Ludwig-Maximilians-University in Munich, Germany) holds the Friede-Springer-Endowed Chair of Business Ethics and Management Accounting at the Martin-Luther-University Halle-Wittenberg. His research interests are on corporate responsibility, sustainability reporting, and ethical decision making. His articles have appeared in such journals as (in alphabetical order) Business Ethics Quarterly, Critical Perspectives on Accounting, and Journal of Business Ethics.

Sascha Raithel (PhD, Ludwig-Maximilians-University in Munich, Germany) is a professor of marketing at Freie Universität Berlin, Germany, School of Business & Economics. His research interests focus on corporate social responsibility, corporate reputation, and brand equity. His articles have appeared in such journals as (in alphabetical order) Business & Society, Journal of the Academy of Marketing Science, Journal of Business Research, Journal of Marketing Research, and Strategic Management Journal.

References

Aerts

Cormier

(2009). Media legitimacy and corporate environmental communication. Accounting, Organizations and Society, 34, 1-27.

Audia

P. G.

Greve

H. R.

(2006). Less likely to fail: Low performance, firm size, and factory expansion in the shipbuilding industry. Management Science, 52, 83-94.

Auger

Devinney

T. M.

Louviere

J. J.

Burke

P. F.

(2008). Do social product features have value to consumers? International Journal of Research in Marketing, 25, 183-191.

Bansal

Clelland

(2004). Talking trash: Legitimacy, impression management, and unsystematic risk in the context of the natural environment. Academy of Management Journal, 47, 93-103.

Bansal

Roth

(2000). Why companies go green: A model of ecological responsiveness. Academy of Management Journal, 43, 717-736.

Barnett

M. L.

Salomon

R. M.

(2006). Beyond dichotomy: The curvilinear relationship between social responsibility and financial performance. Strategic Management Journal, 27, 1101-1122.

Barnett

M. L.

Salomon

R. M.

(2012). Does it pay to be really good? Addressing the shape of the relationship between social and financial performance. Strategic Management Journal, 33, 1304-1320.

Baumann-Pauly

Wickert

Spence

L. J.

Scherer

A. G.

(2013). Organizing corporate social responsibility in small and large firms: Size matters. Journal of Business Ethics, 115, 693-705.

Belal

A. R.

(2002). Stakeholder accountability or stakeholder management: A review of UK firms’ social and ethical accounting, auditing and reporting (SEAAR) practices. Corporate Social Responsibility and Environmental Management, 9, 8-25.

10.

Bhattacharya

C. B.

Sen

(2004). Doing better at doing good: When, why, and how consumers respond to corporate social responsibility. California Management Review, 47(1), 9-24.

11.

Brammer

Millington

(2006). Firm size, organizational visibility and corporate philanthropy: An empirical analysis. Business Ethics: A European Review, 15, 6-18.

12.

Brammer

Pavelin

(2008). Factors influencing the quality of corporate environmental disclosure. Business Strategy & the Environment, 17, 120-136.

13.

Branco

Rodrigues

L. L.

(2008). Factors influencing social responsibility disclosure by Portuguese companies. Journal of Business Ethics, 83, 685-701.

14.

Brown

Fraser

(2006). Approaches and perspectives in social and environmental accounting: An overview of the conceptual landscape. Business Strategy and the Environment, 15, 103-117.

15.

Burritt

R. L.

Schaltegger

(2010). Sustainability accounting and reporting: Fad or trend? Accounting, Auditing & Accountability Journal, 23, 829-846.

16.

Campbell

J. L.

(2007). Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility. Academy of Management Review, 32, 946-967.

17.

Carroll

A. B.

(1979). A three dimensional conceptual model of corporate performance. Academy of Management Review, 4, 497-505.

18.

Carroll

A. B.

(1991). Corporate social performance measurement: A commentary on methods for evaluating an elusive construct. In Post

J. E.

(Ed.), Research in corporate social performance and policy (Vol. 12, pp. 385-402). London, England: JAI.

19.

Chatterji

A. K.

Toffel

M. W.

(2010). How firms respond to being rated. Strategic Management Journal, 31, 917-945.

20.

Cho

C. H.

Patten

D. M.

(2007). The role of environmental disclosures as tools of legitimacy: A research note. Accounting, Organizations and Society, 32, 639-647.

21.

Christophe

Bebbington

(1992). The French Bilan social—A pragmatic model for the development of accounting for the environment? A research note. The British Accounting Review, 24, 281-290.

22.

Chung

K. H.

Pruitt

S. W.

(1994). A simple approximation of Tobin’s Q. Financial Management, 23(3), 70-74.

23.

Clarkson

P. M.

Richardson

G. D.

Vasvari

F. P.

(2008). Revisiting the relation between environmental performance and environmental disclosure: An empirical analysis. Accounting, Organizations and Society, 33, 303-327.

24.

Clarkson

P. M.

Yue

Richardson

G. D.

(2004). The market valuation of environmental capital expenditures by pulp and paper companies. Accounting Review, 79, 329-353.

25.

Cormier

Gordon

I. M.

(2001). An examination of social and environmental reporting strategies. Accounting, Auditing & Accountability Journal, 14, 587-616.

26.

Cormier

Magnan

Van Velthoven

(2005). Environmental disclosure quality in large German companies: Economic incentives, public pressures or institutional conditions? European Accounting Review, 14, 3-39.

27.

CorporateRegister.com. (2011). CRReporting awards ’11 (Experienced Stakeholders Judge the World’s Leading CR Reports). Available from CorporateRegister.com

28.

Damanpour

(1987). The adoption of technological, administrative, and ancillary innovations: Impact of organizational factors. Journal of Management, 13, 675-688.

29.

Damanpour

(1996). Organizational complexity and innovation: Developing and testing multiple contingency models. Management Science, 42, 693-716.

30.

Dawkins

Fraas

J. W.

(2011a). Coming clean: The impact of environmental performance and visibility on corporate climate change disclosure. Journal of Business Ethics, 100, 303-322.

31.

Dawkins

Fraas

J. W.

(2011b). Erratum to: Beyond acclamations and excuses: Environmental performance, voluntary environmental disclosure and the role of visibility. Journal of Business Ethics, 99, 383-397.

32.

Dawkins

Fraas

J. W.

(2013). An exploratory analysis of corporate social responsibility and disclosure. Business & Society, 52, 245-281.

33.

Dhaliwal

D. S.

O. Z.

Tsang

Yang

Y. G.

(2011). Voluntary nonfinancial disclosure and the cost of equity capital: The initiation of corporate social responsibility reporting. The Accounting Review, 86, 59-100.

34.

Dhaliwal

D. S.

Radhakrishnan

Tsang

Yang

Y. G.

(2012). Nonfinancial disclosure and analyst forecast accuracy: International evidence on corporate social responsibility disclosure. The Accounting Review, 87, 723-759.

35.

Donaldson

(2001). The contingency theory of organizations. London, England: Sage.

36.

Donaldson

Preston

L. L.

(1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20, 65-91.

37.

Dye

R. A.

(1985). Disclosure of nonproprietary information. Journal of Accounting Research, 23, 123-145.

38.

Gallo

P. J.

Christensen

L. J.

(2011). Firm size matters: An empirical investigation of organizational size and ownership on sustainability-related behaviors. Business & Society, 50, 315-349.

39.

Global Reporting Initiative. (2006). Sustainability reporting guidelines G3. Amsterdam, The Netherlands: GRI.

40.

Godfrey

P. C.

Merrill

C. B.

Hansen

J. M.

(2009). The relationship between corporate social responsibility and shareholder value: An empirical test of the risk management hypothesis. Strategic Management Journal, 30, 425-445.

41.

Gray

(2001). Thirty years of social accounting, reporting and auditing: What (if anything) have we learnt? Business Ethics: A European Review, 10, 9-15.

42.

Gray

(2007). Taking a long view on what we now know about social and environmental accountability and reporting. Issues in Social and Environmental Accounting, 1, 169-198.

43.

Hair

J. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

Tatham

R. L.

(2006). Multivariate data analysis (6th ed.). New Jersey: Prentice Hall.

44.

Hart

S. L.

Ahuja

(1996). Does it pay to be green? An empirical examination of the relationship between emission reduction and firm performance. Business Strategy and the Environment, 5, 30-37.

45.

Herzig

Schaltegger

(2006). Corporate sustainability reporting. An overview. In Schaltegger

Bennett

Burritt

(Eds.), Sustainability accounting and reporting (pp. 301-324). Dordrecht, The Netherlands: Springer.

46.

Hooghiemstra

(2000). Corporate communication and impression management—New perspectives why companies engage in corporate social reporting. Journal of Business Ethics, 27, 55-68.

47.

Hughes

S. B.

Anderson

Golden

(2001). Corporate environmental disclosures: Are they useful in determining environmental performance? Journal of Accounting and Public Policy, 20, 217-240.

48.

Hummels

Timmer

(2004). Investors in need of social, ethical, and environmental information. Journal of Business Ethics, 52, 73-84.

49.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305-360.

50.

Jones

T. M.

(1995). Instrumental stakeholder theory: A synthesis of ethics and economics. Academy of Management Review, 20, 404-437.

51.

Kolk

(1999). Evaluating corporate environmental reporting. Business Strategy and the Environment, 8, 225-237.

52.

Kolk

Perego

(2010). Determinants of the adoption of sustainability assurance statements: An international investigation. Business Strategy and the Environment, 19, 182-198.

53.

KPMG Global Sustainability Services. (2008). KPMG international survey of corporate responsibility reporting 2008. Amsterdam, The Netherlands: KPMG International.

54.

Lintner

(1965). The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. Review of Economics and Statistics, 47, 13-37.

55.

Lombard

Snyder-Duch

Bracken

C. C.

(2002). Content analysis in mass communication. assessment and reporting of intercoder reliability. Human Communication Research, 28, 587-604.

56.

Lyon

T. P.

Shimshack

J. P.

(2015). Environmental disclosure: Evidence from Newsweek’s green companies rankings. Business & Society, 54(5), 632-675.

57.

Margolis

J. D.

Walsh

J. P.

(2003). Misery loves companies: Rethinking social initiatives by business. Administrative Science Quarterly, 48, 268-305.

58.

McWilliams

Siegel

(2001). Corporate social responsibility: A theory of the firm perspective. Academy of Management Review, 26, 117-127.

59.

Meznar

M. B.

Nigh

(1995). Buffer or bridge? Environmental and organizational determinants of public affairs activities in American firms. Academy of Management Journal, 38, 975-996.

60.

Moch

M. K.

(1976). Structure and organizational resource allocation. Administrative Science Quarterly, 21, 661-674.

61.

Montiel

Husted

B. W.

Christmann

(2012). Using private management standard certification to reduce information asymmetries in corrupt environments. Strategic Management Journal, 33, 1103-1113.

62.

Narayanan

V. K.

Pinches

G. E.

Kelm

K. M.

Lander

D. M.

(2000). The influence of voluntarily disclosed qualitative information. Strategic Management Journal, 21, 707-722.

63.

O’Dwyer

Owen

Unerman

(2011). Seeking legitimacy for new assurance forms: The case of assurance on sustainability reporting. Accounting, Organizations and Society, 36, 31-52.

64.

Orlitzky

(2011). Institutional logics in the study of organizations: The social construction of the relationship between corporate social and financial performance. Business Ethics Quarterly, 21, 409-444.

65.

Owen

D. L.

O’Dwyer

(2008). Corporate social responsibility: The reporting and assurance dimension. In Crane

McWilliams

Matten

Moon

Siegel

D. S.

(Eds.), The Oxford handbook of corporate social responsibility (pp. 384-409). Oxford, UK: Oxford University Press.

66.

Owen

D. L.

Swift

Hunt

(2001). Questioning the role of stakeholder engagement in social and ethical accounting, auditing and reporting. Accounting Forum, 25, 264-282.

67.

Patten

D. M.

(2002). The relation between environmental performance and environmental disclosure: A research note. Accounting, Organizations and Society, 27, 763-773.

68.

Porter

M. E.

Kramer

M. R.

(2006, December). Strategy & society: The link between competitive advantage and corporate social responsibility. Harvard Business Review, 84(12), 78-92.

69.

Reid

E. M.

Toffel

M. W.

(2009). Responding to public and private politics: Corporate disclosure of climate change strategies. Strategic Management Journal, 30, 1157-1178.

70.

Richardson

A. J.

Welker

(2001). Social disclosure, financial disclosure and the cost of equity capital. Accounting, Organizations and Society, 26, 597-616.

71.

Schaltegger

Bennett

Burritt

(Eds.). (2006). Sustainability accounting and reporting. Dordrecht, Netherlands: Springer.

72.

Schaltegger

Burritt

R. L.

(2006). Corporate sustainability accounting: A nightmare or a dream coming true? Business Strategy and the Environment, 15, 293-295.

73.

Schaltegger

Wagner

(Eds.). (2006). Managing the business case for sustainability. Sheffield, UK: Greenleaf.

74.

Schreck

(2011). Reviewing the business case for corporate social responsibility: New evidence and analysis. Journal of Business Ethics, 103, 167-188.

75.

Schuler

D. A.

Cording

(2006). A corporate social performance-corporate financial performance behavioral model for consumers. Academy of Management Review, 31, 540-558.

76.

Sen

Bhattacharya

C. B.

(2001). Does doing good always lead to doing better? Consumer reactions to corporate social responsibility. Journal of Marketing Research, 38, 225-243.

77.

Sharfman

M. P.

Fernando

C. S.

(2008). Environmental risk management and the cost of capital. Strategic Management Journal, 29, 569-592.

78.

Sharpe

W. F.

(1964). Capital asset prices: A theory of market equilibrium under conditions of risk. Journal of Finance, 19, 425-442.

79.

Simnett

Vanstraelen

Wai Fong

(2009). Assurance on sustainability reports: An international comparison. The Accounting Review, 84, 937-967.

80.

Solomon

J. F.

Solomon

Joseph

N. L.

Norton

S. D.

(2013). Impression management, myth creation and fabrication in private social and environmental reporting: Insights from Erving Goffman. Accounting, Organizations and Society, 38, 195-213.

81.

Suchman

M. C.

(1995). Managing legitimacy: Strategic and institutional approaches. Academy of Management Review, 20, 571-610.

82.

SustainAbility, United Nations Environment Programme (UNEP), & Standard & Poor’s. (2006). Tomorrow’s value: The Global Reporters 2006 survey of corporate sustainability reporting. London, England: SustainAbility.

83.

TriplePundit. (2010, April). CSR reporting: What investors want. Retrieved from http://www.triplepundit.com/2010/04/csr-reporting-what-investors-want/

84.

Turban

D. B.

Cable

D. M.

(2003). Firm reputation and applicant pool characteristics. Journal of Organizational Behavior, 24, 733-751.

85.

Udayasankar

(2008). Corporate social responsibility and firm size. Journal of Business Ethics, 83, 167-175.

86.

Ullman

A. A.

(1985). Data in search of a theory: A critical examination of the relationship among social performance, social disclosure and economic performance of US firms. Academy of Management Review, 10, 540-557.

87.

van der Laan Smith

Adhikari

Tondkar

R. H.

(2005). Exploring differences in social disclosures internationally: A stakeholder perspective. Journal of Accounting & Public Policy, 24, 123-151.

88.

Verrecchia

R. E.

(1983). Discretionary disclosure. Journal of Accounting and Economics, 5, 179-194.

89.

Verrecchia

R. E.

(2001). Essays on disclosure. Journal of Accounting & Economics, 32(1-3), 97-180.

90.

Vogel

D. J.

(2005). Is there a market for virtue? The business case for corporate social responsibility. California Management Review, 47(4), 19-45.

91.

von Wysocki

(1993). Sozialbilanzen [Social Accounting]. In Wittmann

Kern

Köhler

Küpper

H.-U.

von Wysocki

(Eds.), Handwörterbuch der Betriebswirtschaft (5th ed., pp. 3904-3916). Stuttgart, Germany. Schäffer-Poeschel.

92.

Wang

Choi

(2008). Too little or too much? Untangling the relationship between corporate philanthropy and firm financial performance. Organization Science, 19, 143-159.

93.

Wang

Bansal

(2012). Social responsibility in new ventures: Profiting from a long-term orientation. Strategic Management Journal, 33, 1135-1153.

94.

Westphal

J. D.

Graebner

M. E.

(2010). A matter of appearances: How corporate leaders manage the impressions of financial analysts about the conduct of their boards. Academy of Management Journal, 53, 15-44.

95.

White

C. F.

(2005). SRI best practices: Learning from the Europeans. Journal of Investing, 14(3), 88-93.

96.

White

(1980). A heteroscedasticity consistent covariance matrix estimator and a direct test of heteroscedasticity. Econometrica, 48, 817-818.

97.

Wood

D. J.

(1991). Corporate social performance revisited. Academy of Management Review, 16, 691-718.