In the Eyes of the Beholder: Experimental Evidence on the Contested Nature of Materiality in Sustainability Reporting

Abstract

Materiality is an ambiguous concept in sustainability reporting where potential users are more heterogeneous than in financial reporting. Building on decision usefulness and dual process theory, we scrutinize in an experimental setting how two important stakeholder groups react to manipulations of the topic of nonfinancial information while controlling for differences in nonfinancial performance. We find that potential employees consider nonfinancial information as generally more material compared with capital market participants. Furthermore, for capital market participants, the nonfinancial topic of energy is more material than the topic of biodiversity, while for potential employees, both topics are equally material. Capital market participants seem to follow a more analytic decision-making process than potential employees. Capital market participants also adjust their decisions in combination with quantitative performance differences, and their risk assessment of a company mediates their decisions. Our findings show that materiality of nonfinancial information lies in the eye of the beholder.

Keywords

materiality nonfinancial disclosure sustainability reporting stakeholder experiment

Introduction

Sustainability reporting¹ covers a wide array of heterogeneous topics, for example, energy usage, biodiversity, human rights issues, and governance aspects, which are of interest to many different stakeholder groups. Consequently, the information needs of the heterogeneous audience of sustainability reports can be ambiguous, and companies struggle to determine what nonfinancial information is material and, thus, should be disclosed. Against this background, research in various disciplines is beginning to specifically discuss issues related to nonfinancial materiality (Bansal & Knox-Hayes, 2013; Edgley, Jones, & Atkins, 2015; Fasan & Mio, 2017; Khan, Serafeim, & Yoon, 2016; Moroney & Trotman, 2016).

Sustainability reporting guidelines from the Sustainability Accounting Standards Board (SASB) and the Global Reporting Initiative (GRI) present definitions of what characterizes “material” nonfinancial information. However, despite identical terminology (i.e., “materiality”), these guidelines have different perspectives. While the SASB focuses on investors, the GRI aims more broadly at stakeholders in general. Regardless of these conceptual differences, definitions of materiality in sustainability reporting rely explicitly on a user perspective. The materiality of information is judged by its potential impact on the decision of the information user. When making materiality judgments, companies have to anticipate potential user reactions.

The SASB and GRI acknowledge that materiality assumptions must be validated with an information user focus (GRI, 2016; SASB, 2017b). This user focus, however, is the crux of ambiguity in the materiality concept when dealing with nonfinancial information because stakeholder groups may differ in their expectations with regard to the information published in sustainability reports. Investors, for example, increasingly include nonfinancial information in their decision-making processes (Global Sustainable Investment Alliance, 2015). They do this not least to consider risk aspects, while potential employees increasingly value morally responsible business conduct (Burbano, 2016; Jones, Willness, & Madey, 2014; Turker, 2009). Consequently, the materiality of any information probably “[. . .], like beauty, is in the eye of the beholder” (Hicks, 1964, p. 159). Against this background, we aim to increase our understanding of nonfinancial materiality from the perspective of different “beholders” (i.e., stakeholder groups) and of the underlying mechanisms driving materiality evaluations. Incorporating the user-oriented perspective on materiality, we ask the following research question: How and why does the materiality of different nonfinancial topics vary among stakeholder groups?

To answer this research question, we applied an experimental approach, which provides the necessary focus on the decision-making perspective and is better able to capture cause-and-effect relationships than, for example, archival research which is often prone to endogeneity concerns (Elliott, Jackson, Peecher, & White, 2014). We conducted an experiment with two different stakeholder groups (capital market participants and potential employees) and analyzed whether and how the topic of the reported nonfinancial information influenced the decision-making of each group. Specifically, we manipulated the content of the nonfinancial information by providing two distinctly different nonfinancial topics (energy and biodiversity) while controlling for differences in nonfinancial performance. In line with decision usefulness theory, we find that potential employees perceived nonfinancial information per se as more material than capital market participants. In addition, we build upon dual-process theory of reasoning and find evidence that, for capital market participants, the topic of energy is more material than the topic of biodiversity. For potential employees, who utilize a more intuitive decision-making process, both topics are equally material. Capital market participants, however, seem to follow a more analytic decision-making process. They adjust their decisions in combination with quantitative performance differences and the risk assessment of a company mediates their decisions.

In answering our research question, we contribute to a deeper understanding of the materiality concept in the realm of nonfinancial reporting. Our focus on the users of company-supplied information adds to the recent foci on reporting companies (Fasan & Mio, 2017) and auditors (Edgley et al., 2015; Moroney & Trotman, 2016). By exposing differences in the decision-making processes (more analytical versus more intuitive systems) and outcomes of different stakeholder groups, our results empirically underline that materiality is an ambiguous concept. We show that materiality cannot be considered an inherent characteristic of any nonfinancial information but rather lies in the eyes of the beholder. On one hand, these results reveal severe challenges for companies engaged in sustainability reporting because they inevitably have to consider what to report and how to approach materiality for their disclosures. On the other hand, our results also help to explain why competing institutional logics have coevolved in the field of reporting nonfinancial information and why they are likely to remain.

Research Background and Hypothesis Development

The Ambiguous Nature of Materiality

The materiality principle plays a central role in corporate disclosure, including sustainability reporting. Of the various initiatives providing sustainability reporting guidelines, two globally recognized organizations stand out: the SASB and the GRI. These two organizations differ considerably in their characterization of the materiality of nonfinancial information. The SASB² aims to integrate its standards into the reporting firms’ filings with the Securities and Exchange Commission (SEC). Thus, the SASB approach is based on the premise that materiality is a fundamental principle of mandated financial disclosure. In characterizing the nature of material information, the SASB explicitly refers to the U.S. Supreme Court’s definition, which considers (financial) information material if there is “a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the ‘total mix’ of information made available” (TSC Industries, Inc. v. Northway, Inc., 1976). When reflecting on the SASB approach, two aspects are noteworthy. First, it focuses on the investor’s perspective. Second, and consequently, the SASB’s understanding of the materiality of nonfinancial information mirrors the understanding of the materiality of financial information by recognizing that some information is more important to investors than other information when they make investment decisions. Any information that only has a noneconomic impact and affects decisions other than those related to investments is, by definition, not material.

The understanding of materiality according to the GRI³ is broader. The GRI determines topics that “reflect the organization’s significant economic, environmental and/or social impacts; or which substantively influence the assessments and decisions of stakeholders” (GRI, 2016, p. 10) as material. Thus, the GRI emphasizes the importance of stakeholder engagement shaping a firm’s sustainability reporting (Grushina, 2017). Similar to the definition of materiality in financial accounting and the SASB’s definition, there is a focus on whether the decision-making process of a user of an information item is influenced. The crucial difference compared with the SASB materiality concept lies in the definition of the potential users of the reported information. While the SASB focuses on (potential) investors, the GRI refers to a variety of different stakeholder groups. Consequently, in the realm of GRI standards, decision-making includes environmental and social considerations, which may not have a (direct) economic impact but are important to one or more stakeholder groups.

Regardless of these differences, both definitions suggest that materiality is influenced by the topic of nonfinancial information. Here, the SASB asks, “Is there a consensus that the proposed topics [. . .] are reasonably likely to have material impacts?” (SASB, 2017b, p. 10). Similarly, the GRI describes material topics that merit inclusion in a report as “those that can reasonably be considered important [. . .] for the decisions of stakeholders” (GRI, 2016, p. 10). This is especially relevant for nonfinancial information, because certain topics are more important to some companies and their stakeholders than to others. For example, energy consumption might be more important for manufacturing companies than for companies in service industries.⁴

Furthermore, the SASB and GRI explicitly rely on a user perspective. Thus, the materiality of information is judged by the potential effect of this information on the information user’s decision. Interestingly, most research about materiality in the context of sustainability reporting focuses on the preparers’ perspective. Fasan and Mio (2017), for example, analyze how companies report their materiality assessment. Moroney and Trotman (2016) report a different understanding of materiality by assurors confronted with financial versus nonfinancial information. Also, from an assuror’s perspective, Edgley et al. (2015) report a different understanding of materiality between accounting and nonaccounting assurors. The few previous studies that take a user perspective mainly rely on questionnaires (e.g., Deegan & Rankin, 1997; O’Dwyer, Unerman, & Hession, 2005) that cannot unequivocally clarify whether and to what extent users consider nonfinancial information in their decision-making process (see also Holm & Rikhardsson, 2008). Materiality can, however, also be measured by observing report users’ decision-making behavior in an experimental context. While also incorporating the direct approach, we apply an experimental research design to analyze how far and why materiality of nonfinancial information lies in the eye of the beholder.

Theory and Hypotheses Development

When examining materiality in sustainability reporting, the most fundamental questions are whether and to what extent stakeholders consider nonfinancial information in their decision-making processes. For a corresponding initial hypothesis on the materiality of nonfinancial information, the classical lens of decision usefulness theory often applied within the accounting domain (Staubus, 2000) is suitable. According to the decision usefulness criterion, information users generally react to positive and negative decision-useful information in the corresponding direction. Initially, the decision usefulness theory addressed capital market participants and their investment-related decisions as the focal point. However, the theory has also been applied in the realm of sustainability reporting (Reimsbach & Hahn, 2015; Tschopp & Nastanski, 2014).

The answer to the question of what is “decision-useful” and thus material depends on the nature of the decision at hand, so it potentially differs between the various stakeholder groups. Capital market participants, for example, usually focus on investment-related, financial decisions, such as investment decisions or stock recommendations. Nonfinancial information is increasingly expected to also influence the financial bottom line of a company (Orlitzky, Schmidt, & Rynes, 2003; Stefan & Paul, 2008) and is included in investment models (Solomon, Solomon, Norton, & Joseph, 2011). Nevertheless, the need for information in an investment context is still dominated by core financial aspects. Prior research has shown that the influence of nonfinancial information diminishes when investors make conscious choices regarding the attractiveness of an investment (Elliott et al., 2014). Even the growing (but still limited) share of socially responsible investors makes mainly financially oriented investment decisions supplemented by nonfinancial aspects (US SIF Foundation, 2018; von Wallis & Klein, 2015). Other stakeholder groups, however, have a more nonfinancial focus in the first place. For example, for potential employees, an important decision is whether to work for a company. Employees generally prefer to work for a company that aligns to their own values and enables them to be proud working for the company (Jones et al., 2014; Turker, 2009) and especially high performing individuals are even willing to accept lower wages when working for a socially responsible company (Burbano, 2016). Against this background of different decision foci, we hypothesize that

Hypothesis 1: Nonfinancial information is more material for potential employees than for capital market participants.

In the following, we build upon the dual-process theory of reasoning (Evans, 2011; Osman, 2004; Stanovich & West, 2000). The theory posits that there is no uniform reasoning in decision-making (Goel, 2007) and instead, “reasoning comprises two underlying systems that serve functionally separate roles” (Osman, 2004, p. 988): the affective, intuitive, or experiential system (also called System 1) or the deliberative, analytic, or reflective system (also called System 2; Evans, 2011; Menzel, 2013). While the first system offers quick and automatic decision-making in an intuitive manner, the second system is slower as it performs analytic tasks in a more formal and controlled way (Metcalfe & Mischel, 1999; Osman, 2004; Slovic, Finucane, Peters, & MacGregor, 2004). System 1 is based on subconscious, often automated aspects, such as emotions, feelings, and intuition, and is “independent of working memory and cognitive abilities” (Evans, 2011, p. 87). System 2 consciously follows logic and algorithms and is “heavily dependent on working memory and related to individual differences in cognitive ability” (Evans, 2011, p. 87). Depending on the decision at hand, the dual-process theory of reasoning presumes that the two systems can be of varying relevance in human decision-making (Menzel, 2013).

As capital market participants process information to make better investment decisions, they will rely more strongly on System 2 because they aim to transform the nonfinancial information into potential financial consequences in a rational and analytic manner. Accordingly, the topic of the provided information can be more or less important to capital market participants. In the following, we explain these different levels of importance based on the exemplary and contrasting topics of energy and biodiversity. Intensive debates about climate change have led to the development of detailed voluntary disclosure standards, such as the Carbon Disclosure Project, and to the implementation of emission trading schemes. The increased interest in this area of sustainability also led to capital market participants’ better understanding of financial consequences of this issue. The discussions about the International Financial Reporting Interpretations Committee (Ascui & Lovell, 2011) show that even accounting standards setters were considering how to monetarize carbon emissions. In addition, empirical research by Matsumura, Prakash, and Vera-Muñoz (2014) and Griffin, Lont, and Sun (2017) shows that the firm value of Standard & Poor’s 500 firms decreases with increasing carbon emissions. Information about energy consumption is relatively easy to measure, and a translation into potential financial consequences seems feasible, for example, based on energy prices. Accordingly, capital market participants are likely to consider the financial consequences of climate-related issues, such as energy consumption, in their decision-making processes.

We contrast information about energy with biodiversity-related information, which is typically more difficult to quantify and for which a translation into financial consequences is rather unclear. This is in line with industry-specific guidance from the SASB that differentiates between higher and lower levels of importance for various nonfinancial topics (SASB, 2017a). Specifically, the SASB frames energy-related issues as highly important in many industries, while biodiversity is mostly seen as a topic of lower or no importance. Accordingly, we expect that energy-related information is readily available in the working memory of professional capital market participants while biodiversity-related information is not.

For potential employees, we expect, based on the dual-process theory, that they rely more heavily on System 1 with its intuitive decision-making for two reasons. First, potential employees typically do not make decisions about their future employment on a regular basis. Their working knowledge and their algorithms relating to such a decision are much more fragmented, especially compared with professional capital market participants with their respective investment-related decisions or recommendations. Therefore, processes relevant for System 2 decision-making (see Evans, 2011; Metcalfe & Mischel, 1999; Osman, 2004; Slovic, Finucane, Peters, & MacGregor, 2004) recede into the background. Second, the rather intuitive decision-making in System 1 typically relies on moral judgments, trained behavior, or social norms. Furthermore, System 1 aims not only at self-interested behavior but also at preserving a person’s self-image (Evans, 2011). Potential employees likely (and unconsciously) consider whether the image of working for a new employer is consistent with their own values. In this regard, Rampl, Opitz, Welpe, and Kenning (2016) show that decision-making of potential employees is strongly linked to emotions, whereas the relevance of working memory and reasoning is decreasing in such situations.

We therefore expect that the “monetization” of sustainability topics as discussed above for the energy topic is less relevant in the decision-making process of potential employees than it is for capital market participants. Furthermore, a topic such as biodiversity likely also has an impact on the System 1–dominated decision-making of potential employees because biodiversity is concerned with the direct impact of a company on different species within its immediate surroundings. Irresponsible handling of the immediate ecological environment affected by a company could lead to stakeholder pressure (e.g., Wolf, 2014) and thus raise discrepancies with the values and self-images of the company’s employees (Kuruppu & Milne, 2010). Energy-related topics will have a considerable impact on the decisions of potential employees because of the topic’s extensive media coverage (Barkemeyer et al., 2017) and the corresponding high awareness about climate change impacts by companies. Due to this publicity, aspects of energy consumption should be readily available even in the more intuitive and affective mode of System 1 decision-making. Based on dual-process theory and the proposed differences in decision-making by professional capital market participants and potential employees, we hypothesize on the materiality of energy-related versus biodiversity-related information:

Hypothesis 2a: For capital market participants, nonfinancial information about energy is more material than nonfinancial information about biodiversity.

Hypothesis 2b: For potential employees, nonfinancial information about energy is as material as nonfinancial information about biodiversity.

Apart from the respective topic of nonfinancial information, the nonfinancial performance can also influence materiality. Consequentially, the SASB explicitly considers whether “performance on the sustainability topic is reasonably likely to have a material impact on companies in the industry” (SASB, 2017b, p. 10). Likewise, the GRI explicitly refers to a “threshold for influencing the [. . .] decisions” (2016, p. 10) of the information-users, and thus, also captures this quantitative performance aspect of materiality, which is also well established in financial reporting (Adams & Simnett, 2011).

In this study, we specifically hypothesize on the effect of reporting decreasing nonfinancial performance because prior research confirmed a stronger reaction to negative nonfinancial information than to positive nonfinancial information (Cho, Lee, & Pfeiffer, 2013). One explanation is that in a setting of voluntary disclosure, companies are expected to provide positively tinged information (Hahn & Lülfs, 2014). Therefore, stakeholders see negative information as more credible (Cho et al., 2013). We argued above based on dual-process theory that the analytic and deliberate System 2 dominates capital market participants’ decision-making. We now add that due to such analytic decision-making, the reactions of this stakeholder group to topics of high materiality will be amplified by the strength of the nonfinancial performance change. We base this expectation on the underlying arithmetic within analytic decisions illustrated by the following numerical example.

Let us assume that the energy topic is perceived as three times more material (i.e., relative materiality = 3) than the biodiversity topic (i.e., relative materiality = 1). Now, let us also assume that a strong decrease in nonfinancial performance is perceived as three times more material (relative performance = × −3) than a weak decrease in nonfinancial performance (relative performance = × −1). For a strong decrease in nonfinancial performance, the weighted materiality is thus 3 × (−3) = −9 for the energy topic and 1 × (−3) = −3 for the biodiversity topic. For a weak performance decrease, the weighted materiality is 3 × (−1) = −3 for the energy topic and 1 × (−1) = −1 for the biodiversity topic. The difference in weighted materiality between the two topics is much larger in the case of a strong decrease compared with the small decrease. For energy, the difference in weighted materiality is −6 (−3 for weak decrease versus −9 for strong decrease), while for biodiversity this difference is only −2 (−1 for weak decrease vs. −3 for strong decrease). We thus expect that in the analytic decision-making process of capital market participants, a strong decrease in nonfinancial performance leads to a stronger adjustment of the investment decisions for topics of high materiality compared with topics of low materiality because it is translated into a more pessimistic forecast of a company’s future financial success. Therefore, we hypothesize that

Hypothesis 3: For capital market participants, differences in materiality between an energy topic and a biodiversity topic are larger in the case of a strong decrease in nonfinancial performance.

For the group of potential employees, we hypothesized above that energy-related information is as material as biodiversity information (i.e., we expect no significant main effect of the nonfinancial topic for employees). Consequentially, a change of performance should not lead to different reactions for energy versus biodiversity information. Therefore, we do not offer a performance-related hypothesis for this group.

Method

Participants

The experiment focuses on two of the most important stakeholder groups, capital market participants and (potential) employees, which allows us to compare two similarly relevant groups from the same general type of stakeholder classification: (quasi) internal and primary stakeholders. For the group of professional capital market participants, we recruited financial analysts and advisors via e-mail, telephone calls, and personal contacts. Of the 121 final participants, 73 were male (2 participants chose not to answer this question). Seventy-one participants provided information on their age with a mean of 41.70 years and mean work experience in their professional role of 10.55 years from 99 responses. Most of the participants worked in Germany (62.80%), the remainder worked in Austria (14.05%), England (14.05%), Italy (3.31%), Spain (3.31%), and Switzerland (2.48%). Results of chi-square tests did not show statistically significant differences regarding the personal characteristics and the dependent variables between the group of Germany-based and other participants.

For the group of potential employees, we selected participants who typically already have and/or will regularly apply for a full-time position, so that they are familiar with the decision task at hand. Thus, we used business students for the sample of potential employees (see also, for example, Evans & Davis, 2011; Greening & Turban, 2000; Kuruppu & Milne, 2010; Turban & Greening, 1997).⁵ The 129 respondents in the sample of potential employees (62 female) were, on average, 23.16 years old and had a mean work experience of 0.92 years. The further demographics verify that the participants are indeed real-life potential employees: 108 (83.7%) had previously applied for a job (e.g., apprenticeship, internship, full- or part-time job) and 118 (91.5%) were planning to apply for a regular full-time job within the next 2 years.

Experimental Design and Procedure

The experiment followed a full-factorial, between-subjects design. We assigned all participants randomly to one of the experimental groups.⁶ Our main manipulated variable was a nonfinancial topic. Additionally, we manipulated the variable (nonfinancial) performance, resulting in a total of four experimental groups (see Figure 1), because we also address whether the nonfinancial topic interacts with changes in nonfinancial performance. A Kruskal–Wallis test did not reveal statistically significant differences between the four groups in each sample in terms of personal characteristics (p > .1). Randomization of participants was thus successful.

Figure 1.

Experimental groups.

We provided the same company information to the two different stakeholder groups as described above (see Appendices A and B). This mirrors the realistic setting in which a company discloses one sustainability report to address the firm’s many different stakeholder groups. We conducted the experiment online so that, apart from the questions asked about dependent variables for the different stakeholder groups, all circumstances were otherwise equal. All participants had access to the same general introduction to the fictitious “Alpha Company” and its financial highlights, including the statement of income and the statement of cash flows (see Appendices A and B for details).⁷ We modeled the experimental material following actual reports from a real-life company, and disguised the company’s identity to prevent any previous knowledge of the company affecting the participants’ judgment (similarly, e.g., to Holm & Rikhardsson, 2008; Reimsbach & Hahn, 2015). Following previous studies that focus on nonfinancial performance aspects (Elliott et al., 2014; Reimsbach, Hahn, & Gürtürk, 2018), we chose a company with an average financial performance so that the participants did not implicitly or explicitly focus on extraneous variables that were not the focus of the study. Alpha Company was described as a STOXX 50 listed multinational European chemical company that manufactures plastics, chemicals, and agricultural products. We deemed the chemical industry an appropriate background because sustainability-related topics have been on the agenda of the industry for a long time, so we expect stakeholders to be aware of these aspects.

For the main independent variable topic, the two groups in the left half of Figure 1 received nonfinancial information on the topic “biodiversity” while the two groups in the right half received nonfinancial information on the topic “energy.” We conducted a qualitative content analysis of 10 annual or sustainability reports from companies in the chemical industry issued in 2015 as a prestudy to determine these two topics. Specifically, we analyzed the materiality process and the justification for materiality issues in the reports, which included paragraphs on how the company determined the materiality of issues to report and a list or matrix of issues with higher and lower materiality. Constant throughout the reports, energy (or energy consumption) was presented as a topic of high importance and biodiversity was described as a topic of medium or low importance for the respective company in the chemical industry. Thus, these topics are appropriate as manipulations for topic materiality in the experiment. In addition, the SASB published a Materiality Map^TM (SASB, 2017a) focusing on the materiality of nonfinancial information for a number of industries (Khan et al., 2016). In line with our own assessment of company reports, the SASB identifies “energy management” as an issue of high importance and defines “biodiversity impacts” as an issue of low importance for chemical industries. We purposefully chose two topics from the environmental pillar of sustainability so that the results of the experiments are not influenced by a heterogeneous choice of topics, for example, from the environmental versus the social domain of sustainability. Nevertheless, although biodiversity and energy are environmental issues, the topics’ implications differ. Therefore, we can assess potential differences in the perceived materiality of the two stakeholder groups in this study.

Following Elliott et al. (2014), we report a rating score for Alpha Company’s nonfinancial performance. A rating score is suitable for the experimental manipulations because it can show a quantitatively identical change in the performance of the two nonfinancial topics independently of any scaling issues for specific indicators. Furthermore, providing a (third party) rating score as opposed to an unfiltered management disclosure avoids concerns about the credibility and/or reliability of the reported performance measures (see, Elliott et al., 2014), which is not the focus of the study. All participants received a table that reported this score in either of the two experimental conditions for the topic. For the variable performance, we manipulated the difference between Alpha Company’s nonfinancial performance score and the industry average score. After an identical initial value in all experimental conditions of 62 in 2014 (industry average 65), Alpha Company’s score declined to 60 in the weak decrease condition and 42 in the strong decrease condition (the industry average remained at 65) for 2015. Thus, Alpha Company’s nonfinancial topic score is always below the industry average, because, in line with previous research, we expect a higher (perceived) impact of negative information (Murray, Sinclair, Power, & Gray, 2006; Solomon et al., 2011; see the reasoning above).⁸

The information provided to the participants was accompanied by a narrative section explaining that the respective performance score relied on two key performance indicators (i.e., “carbon emissions” and “energy efficiency” for the topic energy, and “renewable resources” and “production site near protected areas” for the topic biodiversity). Thus, we kept the presentation format and the information extent constant over all experimental conditions. We pretested the experimental material to assess its internal consistency and plausibility (Wason, Polonsky, & Hyman, 2002). Twenty-eight professionals with a current capital market work background completed the pretest and suggested minor changes to the material to enhance its understandability and ensure that the material was appropriate and realistic.

The experimental task required the participants to complete several steps. After accessing the web page of the online study, the participants first read the instructions and the brief introduction to Alpha Company. Afterward, participants accessed the experimental material depending on the experimental condition to which they were randomly assigned. Then, the participants had to provide their judgments. Throughout these judgments and decision-making processes, participants had access to the reports to obtain the required information. Finally, participants could comment on their judgments. After the participants completed this task, the reports were no longer accessible, and participants were asked to respond to the following manipulation checks to ensure that our manipulations were successful. The questions concerned the topic in the participants’ respective reports and the nonfinancial performance score. For the topic manipulation, we asked participants to identify out of a list of six options (human rights, water, biodiversity, product stewardship, energy, and corporate governance) the nonfinancial topic included in their respective report. In the professional capital market participants’ sample, only two participants failed to answer this question correctly, while in the potential employee sample, six participants failed. We excluded these participants from the analysis.

To test the effectiveness of the performance manipulation, we asked the participants to rate how the nonfinancial performance changed from the previous to the current fiscal year on an 8-point scale ranging from extreme decrease (1) to extreme increase (8). Participants in the strong decrease conditions (mean of 1.76 for capital market participants and 2.97 for employees) and the weak decrease conditions (mean of 3.63 for capital market participants and 3.97 for employees) correctly indicated that the nonfinancial performance decreased.⁹ Furthermore, the mean difference between the two performance conditions was statistically significant (p < .01) for both samples, indicating that participants in the strong decrease condition correctly identified the stronger decrease in nonfinancial performance compared with participants in the weak decrease condition. Therefore, eliminating additional further candidates was not necessary, and the final samples consisted of 121 capital market participants and 129 potential employees. Finally, participants answered (nonmandatory) demographic questions.

Dependent Variables

As outlaid before, materiality from the report users’ perspective can be measured by asking the report users directly, as questionnaire-based research does, or by observing report users’ decision behavior in an experimental context. Our approach incorporates both perspectives. We asked participants in the postexperimental questionnaire to explicitly rate the relevance of nonfinancial information for their judgment and decision-making on a scale from 0 (not at all relevant) to 8 (extremely relevant). The focus on “relevance” follows the idea of the International Accounting Standards Board, which classifies materiality as an entity-specific aspect of relevance (and our study uses a specific entity/case), and we avoided potential confounding effects through the introduction and use of the term materiality. Thus, our first dependent variable for both stakeholder groups was nonfinancial relevance. Other than that, we used stakeholder group-specific dependent variables that represent the judgments of each group in our experiment. Hence, we followed the broader understanding of materiality of nonfinancial information as provided by the GRI. Naturally, different stakeholder groups analyze company disclosure in different decision contexts. While capital market participants primarily deal with buy or sell decisions of company shares, potential employees make rather noneconomic judgments, such as whether the company is an attractive employer.

The dependent variable for capital market participants was stock recommendation. We asked participants what recommendation they would deem fair and appropriate on a 5-point scale, ranging from strong sell to strong buy (for a similar procedure, see Ghosh & Wu, 2012). As the participants were professional capital market participants, they should be familiar with this rating task and scale.¹⁰ We asked participants in the potential employees’ group to rate the company in terms of its attractiveness as an employer (employer attractiveness) based on a 7-point Likert-type scale (Turban & Greening, 1997).¹¹

Results

Hypotheses Testing

Hypothesis 1 addresses the materiality of nonfinancial information for our two subject groups. Hypothesis 1 posits that nonfinancial information is more material for potential employees than for capital market participants. For the sample of capital market participants, the mean nonfinancial relevance was 3.68 (for the bio-weak condition), 4.81 (bio-strong), 4.76 (energy-weak), and 5.59 (energy-strong; see Table 1, Panel A).¹² For the group of potential employees, the respective means were 5.14 (bio-weak), 5.13 (bio-strong) 5.24 (energy-weak), and 5.62 (energy-strong; see Table 1, Panel B).¹³ The mean nonfinancial relevance for potential employees (5.28) over all groups was thus significantly higher (p < .01) than the nonfinancial relevance for capital-market participants (4.65), which is in line with Hypothesis 1.¹⁴

Table 1.

Nonfinancial Relevance.

Panel A: Descriptive statistics (professional capital market participants); mean [SD]
		Nonfinancial topic
		n	Biodiversity	n	Energy	n	Σ
Nonfinancial performance	Weak decrease	34	3.68 [1.25]	29	4.76 [1.24]	63	4.17 [1.35]
	Strong decrease	31	4.81 [1.11]	27	5.59 [1.08]	58	5.17 [1.16]
	Σ	65	4.22 [1.31]	56	5.16 [1.23]	121	4.65 [1.35]
Panel B: Descriptive statistics (potential employees); mean [SD]
		Nonfinancial topic
		n	Biodiversity	n	Energy	n	Σ
Nonfinancial performance	Weak decrease	28	5.14 [1.56]	41	5.24 [1.97]	69	5.20 [1.80]
	Strong decrease	31	5.13 [1.63]	29	5.62 [1.37]	60	5.37 [1.52]
	Σ	59	5.14 [1.58]	70	5.40 [1.75]	129	5.28 [1.67]

Note. We asked participants in the postexperimental questionnaire to rate the relevance of nonfinancial information for their judgment and decision-making on a scale ranging from 0 = not at all relevant to 8 = extremely relevant.

For Hypotheses 2 and 3, we plot the results in Figure 2 and report the results of the statistical hypotheses tests in Tables 2 to 4. Hypothesis 2a posits that for capital market participants, nonfinancial information about energy is more material than nonfinancial information about biodiversity. Our focus on a nonfinancial performance decrease implies that participants receiving energy information give less favorable stock recommendations compared with those who receive biodiversity information. In line with these expectations, the mean stock recommendation was lower for the energy group compared with the biodiversity group (2.27 versus 2.51; see Panel A, Table 2). These results are directionally consistent with Hypothesis 2a, and the ANOVA (analysis of variance) term for topic is statistically significant (Panel B in Table 2; p < .05). The plotted data (see Figure 3) and an additional series of post hoc tests¹⁵ are consistent with Hypothesis 2a as well. Our inquiry of the perceived nonfinancial relevance provides further support for Hypothesis 2a. Capital market participants judged the relevance of nonfinancial information significantly higher (p < .01) for the energy topic (mean of 5.16) than for the biodiversity topic (mean of 4.22).

Figure 2.

Plotted results.

Table 2.

Professional Capital Market Participants.

Panel A: Descriptive statistics (stock recommendation^a); mean [SD]
		Nonfinancial topic
		n	Biodiversity	n	Energy	n	Σ
Nonfinancial performance	Weak decrease	34	2.82 [0.46]	29	2.69 [0.54]	63	2.76 [0.50]
	Strong decrease	31	2.16 [0.58]	27	1.81 [0.74]	58	2.00 [0.68]
	Σ	65	2.51 [0.64]	56	2.27 [0.77]	121	2.40 [0.70]
Panel B: Results of ANOVA (stock recommendation^a)
Source		df	Sum of squares	Mean square	F	p value^d	Partial η²
Corrected model		3	19.54	6.51	19.37	.000	.331
Topic^b		1	1.73	1.73	5.14	.025	.042
Performance^c		1	17.74	17.74	52.65	.000	.310
Topic × Performance		1	0.34	0.34	1.01	.318	.009
Error		117	39.42	0.34
Panel C: Simple effect tests
Contrast: Effect of topic in the case of a weak performance decrease		1	0.28	0.28	0.83	.363	.007
Contrast: Effect of topic in the case of a strong performance decrease		1	1.73	1.73	5.14	.025	.042
Contrast: Effect of performance in the case of a biodiversity topic		1	7.11	7.11	21.12	.000	.153
Contrast: Effect of performance in the case of an energy topic		1	10.71	10.71	31.77	.000	.214
Contrast: Topic by performance interaction^e		1	14.49	14.49	42.59	.000	.331

Note. ANOVA = analysis of variance.

We asked participants what recommendation they would deem fair and appropriate on a 5-point scale, ranging from 1 = strong sell to 5 = strong buy. ^bTopic is 1 if participants received information on the topic energy and 0 otherwise. ^cPerformance is 1 if participants received information indicating a strong decrease in nonfinancial performance and 0 otherwise. ^dp values are two-tailed. ^eWe use contrast weights of −5 for the energy-strong condition, +1 for the bio-strong and the energy-weak conditions, and +3 for the bio-weak condition).

Figure 3.

Mediation analysis.

Hypothesis 2b posits that for potential employees, nonfinancial information about energy is as material as nonfinancial information about biodiversity. For this group, we analyzed the attractiveness of the reporting company as an employer. Potential employees judged this attractiveness almost identically no matter whether they received negative information on the energy topic or on the biodiversity topic (mean of 3.81 vs. 3.75; see Panel A, Table 3), and the respective ANOVA terms for topic are not statistically significant (Panel B in Table 3, p > .1), just as the series of post hoc tests indicate.¹⁶ Finally, our results show that the perceived nonfinancial relevance between the potential employees receiving information on the energy topic (mean of 5.40) and potential employees receiving information on the biodiversity topic (mean of 5.14) was statistically indistinguishable (p > .1). These results support Hypothesis 2b.

Table 3.

Potential Employees.

Panel A: Descriptive statistics (top-choice employer^a); mean [SD]
		Nonfinancial topic
		n	Biodiversity	N	Energy	n	Σ
Nonfinancial performance	Weak decrease	28	3.68 [1.34]	41	3.95 [1.45]	69	3.84 [1.40]
	Strong decrease	31	3.81 [1.40]	29	3.62 [1.29]	60	3.72 [1.34]
	Σ	59	3.75 [1.36]	70	3.81 [1.39]	129	3.78 [1.37]
Panel B: Results of ANOVA (top-choice employer^a)
Source		df	Sum of squares	Mean square	F	p value^d	Partial η²
Corrected model		3	2.247	0.75	0.39	.758	.009
Topic^b		1	0.06	0.06	0.31	.860	.000
Performance^c		1	0.32	0.32	0.17	.681	.001
Topic × Performance		1	1.66	1.66	0.871	.352	.007
Error		125	237.68	1.90
Panel C: Planned contrast tests
Contrast: Effect of topic in the case of a weak performance decrease		1	1.24	1.24	.65	.421	.005
Contrast: Effect of topic in the case of a strong performance decrease		1	0.52	0.52	.27	.603	.002
Contrast: Effect of performance in the case of a biodiversity topic		1	0.24	0.24	.13	.723	.001
Contrast: Effect of performance in the case of an energy topic		1	1.86	1.86	.98	.325	.008

Note. ANOVA = analysis of variance.

Participants were asked to rate whether the company would be one of their top choices as an employer on a 7-point Likert-type scale ranging from 1 = strongly disagree to 7 = strongly agree. ^bTopic is 1 if participants received information on the topic energy and 0 otherwise. ^cPerformance is 1 if participants received information indicating a strong decrease in nonfinancial performance and 0 otherwise. ^dp values are two-tailed.

Table 4.

Perceived Investment Risk of Professional Capital Market Participants.

Panel A: Descriptive statistics (perceived investment risk^a); mean [SD]
		Nonfinancial topic
		n	Biodiversity	n	Energy	n	Σ
Nonfinancial performance	Weak decrease	34	38.53 [16.68]	29	45.52 [11.52]	63 41.75	[14.84]
	Strong decrease	31	54.35 [15.48]	27	62.78 [17.72]	58 58.28	[16.95]
	Σ	65	46.08 [17.86]	56	53.84 [17.08]	121 49.67	[17.86]
Panel B: Results of ANOVA (perceived investment risk^a)
Source		df	Squares	Mean square	F	p value^d	Partial η²
Corrected model		3	10039.30	3346.43	13.86	.000	.262
Topic^b		1	1783.11	1783.11	7.38	.080	.059
Performance^c		1	8218.97	8218.97	34.04	.000	.225
Topic × Performance		1	15.46	15.46	0.06	.801	.001
Error		117	28247.48	241.43

Note. ANOVA = analysis of variance.

We asked participants to assess the risk of an investment in Alpha Company relative to the average firm of equivalent size in the same industry on a scale ranging from 0 = very low risk to 100 = very high risk. ^bTopic is 1 if participants received information on the topic energy and 0 otherwise. ^cPerformance is 1 if participants received information indicating a strong decrease in nonfinancial performance and 0 otherwise. ^dp values are two-tailed.

Hypothesis 3 posits for capital market participants, that differences in energy versus biodiversity materiality are larger in the case of a strong change in nonfinancial performance. A visual inspection of the patterns in Figure 2 shows a steeper slope in the energy conditions compared with the biodiversity conditions. Differences in investment-related judgments between participants who received information on the energy topic thus seem to be larger (smaller) if the respective report indicated a strong (weak) decrease in nonfinancial performance. This is in line with Hypothesis 3 and hints at a potential ordinal interaction between the reported topic and nonfinancial performance changes for professional capital market participants.

Simple effect tests provide quantitative support for an ordinal interaction (Panel C in Table 2). The difference in stock recommendation between the biodiversity topic and the energy topic conditions increases from 0.13 in the weak decrease treatment groups (= 2.82 for bio-weak − 2.69 for energy-weak) to 0.35 (= 2.16 for bio-strong − 1.81 for energy-strong) in the strong decrease groups. Furthermore, the mean difference in stock recommendation between the biodiversity and energy treatment groups is statistically significant only for the strong decrease condition (F = 5.14, p < .05), while the mean difference for the weak decrease condition is not statistically significant (F = 0.83, p = .36). These results provide initial support for an ordinal interaction.

We further analyzed the functional form of the interaction using planned contrasts (Buckless & Ravenscroft, 1990; see also, Lachmann, Stefani, & Wöhrmann, 2015), because for an ordinal interaction, ANOVA is less powerful as a statistical tool (Buckless & Ravenscroft, 1990). Planned contrasts increase the statistical power without increasing Type I error rates. We used contrast weights consistent with the expectation derived in Hypothesis 3 that reporting a nonfinancial performance decrease on an energy topic is always associated with a lower perceived stock recommendation compared with reporting the same decrease for a biodiversity topic. This effect is larger if the respective decrease in nonfinancial performance is strong instead of weak. When performing a planned contrast analysis, the sum of the chosen contrasts needs to equal zero (Buckless & Ravenscroft, 1990). Thus, we transformed the values for the weighted materiality we used when deriving Hypothesis 3 by subtracting the average weighting factor as follows:

\begin{array}{l} c o n t r a s t w e i g h t (t o p i c i, p e r f o r m a n c e j) = w e i g h t e d m a t e r i a l i t y (t o p i c i, p e r f o r m a n c e j) \\ - a v e r a g e w e i g h t e d m a t e r i a l i t y . \end{array}

For the planned contrast analysis, we used contrast weights of −5 for the energy (strong) condition, +1 for the bio (strong) and the energy (weak) conditions, and +3 for the bio (weak) condition.¹⁷ These contrasts test whether the difference in stock recommendation between firms reporting on an energy topic compared with firms reporting on a biodiversity topic is larger for firms reporting a strong compared with a weak nonfinancial performance decrease. The results of this test are reported in Panel C of Table 2. The planned contrast is statistically significant (p < .01). Thus, we find evidence of an ordinal interaction and our results support Hypothesis 3.

Additional Analysis: The Mediating Role of Risk Assessment for Capital Market Participants

We further explored what drove the results for capital market participants because this group showed relatively strong reactions to changes in the topic of nonfinancial information. Some authors suggest that capital market participants see sustainability information primarily from a risk perspective (Murray et al., 2006; Solomon et al., 2011). In the context of this study, decreasing nonfinancial performance indicates a potential risk leading to a negative reaction, while good nonfinancial performance merely indicates the absence of a related risk and does not necessarily lead to a positive reaction. Therefore, acknowledging the mediating role of the risk perspective of sustainability disclosures might help to explain the mechanism through which the disclosed information influences capital market participants’ decision-making process. In their analytical decision-making (System 2), capital market participants likely attribute higher risks to a company that reports a performance decrease on the energy topic as opposed to reporting the same performance decrease on the biodiversity topic. As argued above, the energy topic can be transformed into monetary terms more directly, so that financial risks (e.g., of stricter carbon regulation, carbon taxes, or lost sales due to increased customer awareness) are easier to identify as well. The more negative risk assessment for the energy topic would lead to more negative investment-related judgments. Therefore, the risk assessment of a company potentially has a mediating role for the impact of the topic dimension on investment-related judgments. To better understand whether and how the risk perspective influences capital market participants’ decision-making, we also analyzed their risk perception. Table 4 provides descriptive statistics and ANOVA analysis for perceived investment risk.¹⁸ Following Baron and Kenny (1986), we performed a three-stage regression (see also, Shen, Wu, & Chand, 2017) to test the mediating effect of the perceived investment risk (see Figure 3).

In Regression (1), the relationship between the independent variable topic and perceived investment risk is positive and statistically significant (a₁ = 0.218, p < .05), indicating that the energy topic is associated with a higher perceived investment risk compared with the biodiversity topic. In Regression (2), we find a negative and statistically significant association between the independent variable topic and stock recommendation (b₁ = −0.171, p < .1). In addition, the relationship between perceived investment risk and stock recommendation is negative and statistically significant in Regression (3) (c₂ = −0.593, p < .01). Finally, the effect of the independent variable on stock recommendation is smaller (i.e., less negative) in the third regression than in the second (c₁ = −0.042, p > .1 vs. b₁ = −0.171, p < .01). Overall, the results display all conditions that must hold to show a mediating effect according to Baron and Kenny (1986), and thus, support the notion that perceived investment risk mediates the effect of topic on stock recommendation.

Discussion

Theoretical Insights Into the Materiality Concept in Nonfinancial Reporting

This study contributes to the literature on the materiality principle for nonfinancial information (Edgley et al., 2015; Fasan & Mio, 2017; Khan et al., 2016; Moroney & Trotman, 2016). We took a user’s perspective and empirically captured different understandings of materiality in an experimental setting considering two different stakeholder groups (i.e., capital market participants and potential employees). First, and in line with the decision usefulness theory, we found that potential employees perceived nonfinancial information per se as more material than capital market participants. This is the first evidence that the materiality of nonfinancial information is stakeholder-dependent and thus lies in the eye of the beholder.

Second, in line with the dual-process theory of reasoning, we found that for capital market participants, the topic energy was more material than the topic biodiversity, while for potential employees both topics were equally material. The information on energy seems to be easier to translate into financial terms than biodiversity information and, therefore, is better suited for the more analytical System 2 decision-making processes (Evans, 2011; Menzel, 2013) apparently used by capital market participants. Specific to the case in this study, it is easier to identify costs and risks arising from excessive energy use than from a high impact on biodiversity. Thus, the capital market participants in this study regarded the former as a topic of high materiality, which is also in line with the SASB Materiality Map. Potential employees, instead, seem to rely more on System 1 decision-making. Accordingly, energy and biodiversity have a similar impact on decision-making of potential employees. This in is line with our arguments that translating nonfinancial performance into financial terms is less relevant for potential employees, while it is more important that company behavior is in line with the values of its potential employees.

Third, and again referring to the dual-process theory of reasoning, we found that for capital market participants, differences in materiality between an energy and a biodiversity topic are larger in the case of a strong change in nonfinancial performance. Accordingly, capital market participants considered the nonfinancial topic and performance simultaneously when making investment-related judgments and reacted most negatively to a strong decrease in the energy topic (since this translates to a larger negative impact on financial performance) and least negatively to a small decrease in the biodiversity topic. This is further support of an analytical System 2 decision-making process.

Fourth, our results reveal the underlying mechanism that drove capital market participants’ (implicit) materiality judgments. The risk assessment of a company mediated the impact of the topic dimension on investment-related judgments. More specifically, the previously statistically significant relationship between topic and stock recommendation was no longer statistically significant when perceived investment risk was considered a mediator. According to Baron and Kenny (1986), this result illustrates perfect mediation. In the context of the present study, this means for capital market participants, the impact of topic on perceived investment risk and the impact of perceived investment risk on stock recommendation accounts for all of the observed relationship between topic and stock recommendation. This finding strongly supports the notion that capital market participants look at nonfinancial information from a risk perspective (Murray et al., 2006; Solomon et al., 2011). Thereby, the capital market participants seem to reduce the information content of the energy topic to purely financial materiality interpretations.

Dilemmas of Extant Understandings of Materiality in Nonfinancial Reporting

Our results highlight the differences in stakeholder decision-making when confronted with identical nonfinancial topics and performance changes (see also, Byrch, Milne, Morgan, & Kearins, 2015) and they empirically underline that materiality is a highly ambiguous concept. Most important, materiality cannot be considered an inherent characteristic of any nonfinancial information, but it rather lies in the eyes of the beholder. This poses significant challenges for companies’ decisions on what to include in their nonfinancial reporting. A way out of the dilemma of what is material for whom, in principle, can be found in two ways.

First, materiality definitions can reduce complexity by focusing on few stakeholder groups with presumably homogenous decision-making patterns. This road is taken by the SASB whose approach exclusively focuses on investors. However, given that prior research framed sustainability reporting as “a new genre of organizational communication” (Grushina, 2017), this simplification hardly lives up to the general idea of materiality from a sustainability point of view. Natural scientific research, for example, illustrates the relatively higher importance of biodiversity compared with climate change aspects when it comes to reducing the risk of the destabilizing human impact on Earth (Steffen et al., 2015). Thus, a focus on financial materiality is limited to what Bansal and Knox-Hayes (2013) frame as “sociomateriality.” The important physical materiality of the natural environment, however, is lost in this perspective (Bansal & Knox-Hayes, 2013). Furthermore, even a focus on specific stakeholder groups does not necessarily solve the mentioned problem if there is a high degree of in-group heterogeneity in decision-making patterns. Prior research has, for example, identified such heterogeneity between sustainability-conscious and unconscious decision-makers for the stakeholder groups of investors (see von Wallis & Klein, 2015), potential employees (e.g., Burbano, 2016), and customers (e.g., Mohr & Webb, 2005). Thus, even for a single, well-defined stakeholder group, finding consistently material topics might not be a trivial task for the reporting company.

Second, an alternative avenue out of the dilemma is to acknowledge that stakeholder groups differ in their respective materiality assessments and that each company individually needs to engage in a stakeholder dialogue to prioritize the materiality of nonfinancial topics, for example, through a materiality matrix. This represents the procedure advocated by the GRI. However, this approach leads to a severe lack of comparability between reporting entities. Furthermore, it could still lead to a reporting in which certain stakeholder groups are marginalized if they are deemed as less important than others, and even if less than in the SASB approach.

Our results also should be considered in the light of the literature on the institutionalization of competing logics within a field (e.g., Dunn & Jones, 2010; Purdy & Gray, 2009). These studies show that, in institutional fields characterized by highly dichotomized parties that allow no possibility for a consensus to emerge, this conflict can be institutionalized in the form of (lasting) competing institutional logics. This applies to the case of standard setting for reporting nonfinancial information best illustrated by the different understanding of materiality of the SASB and the GRI. Their respective understandings of materiality represent the opposite sides of a spectrum ranging from the consideration of a single stakeholder group (SASB) to the consideration of virtually all possible stakeholder groups (GRI). Given that our results indicate that materiality always lies in the eye of the beholder, a synthesis of these opposite logics does neither seem realistic nor adequate so that competing logics on the materiality approach in the realm of reporting nonfinancial information will probably be a lasting phenomenon.

Conclusion

This study provides a deeper understanding of the materiality concept in nonfinancial reporting. While recent studies focus on companies (Fasan & Mio, 2017) and auditors (Edgley et al., 2015; Moroney & Trotman, 2016), we add the user perspective through the lens of decision usefulness theory and dual-process theory. In our experimental setting, we show how and why nonfinancial information affects stakeholder decisions and how the different decision-making processes lead to a state where materiality of nonfinancial information ultimately lies in the eye of the beholder. Consequently, and in line with studies on the institutionalization of competing logics (e.g., Dunn & Jones, 2010), different materiality concepts for reporting nonfinancial information have coevolved. They range from a narrow understanding focusing on a single, homogenous stakeholder group and their information needs (SASB approach), to a broader approach with the acknowledgment of a multitude of heterogeneous stakeholder groups and their respective decision foci (GRI approach). Only the latter approach, in which materiality must be determined through a complex process of stakeholder engagement, is able to capture the complex concept of sustainability, but it comes at the cost of impaired comparability of sustainability reports. Overall, our results empirically explain why competing institutional logics have co-evolved in the field of reporting nonfinancial information and why they are likely to remain.

We acknowledge that our findings come with several methodological limitations. Experiments inevitably involve an oversimplification of the real decision-making process because participants in the experiment cannot collect the information that they perceive to be useful themselves (Chan & Milne, 1999). In real-life decisions, the amount of information processed by stakeholders is likely to be much more extensive, reducing the relative attention paid to each piece of information. Although the limited amount of information enabled us to conduct the experiment and increase internal validity, it could have come at the cost of reduced external validity. Further research could focus on the information-seeking behavior by different stakeholder groups specifically regarding different types and sources of sustainability information. However, including and investigating the relevance of different sources of information was beyond the scope of this study. Furthermore, also for design reasons, we had to keep the financial performance of the reporting company constant over all experimental conditions. It is possible, that a change in financial performance mitigates the negative effects of a nonfinancial performance decrease. As these effects were beyond the scope of this study, we encourage future (experimental) research on the interplay of financial and nonfinancial performance changes.

Regarding the participants themselves, we directly approached capital market participants, and we used graduate business students as participants for the stakeholder group of potential employees. The use of students as participants is common in experimental research on corporate disclosure (Elliott et al., 2014; Holm & Rikhardsson, 2008) and, in our setting, the role of students as potential employees is realistic. However, the results may not be generalizable to other stakeholder groups of decision makers. Future research might focus on the materiality understandings of additional stakeholder groups to identify where homogeneous and heterogeneous understandings occur. Furthermore, we did not provide monetary incentives to the participants. Given that investment recommendations in real life are not immediately linked to a “performance-dependent” compensation and that (partially) noneconomic judgments such as employee choice are not approachable via “performance-dependent” compensation anyway, monetary incentives are a less powerful tool. Nevertheless, other decision situations could provide valuable scenarios for incentivized experiments.

Finally, although the experimental setting allows for a strong interpretation of the impact of the manipulated information, it restricts the analysis to this specific setting. Our study specifically aimed at the two topics of energy and biodiversity for chemical companies. This focus was necessary because materiality is sector specific (Eccles, Krzus, Rogers, & Serafeim, 2012). However, the results are not necessarily transferable to other settings. In addition, other forms of representing information through impression management (Solomon, Solomon, Joseph, & Norton, 2013) can impact the decisions of the users of such information. Further studies of how sector-specific characteristics influence materiality could follow, as well as studies addressing different topics or representations of nonfinancial information.

Footnotes

Appendix A

Appendix B

Acknowledgements

We would like to thank Markus Milne and Michael Jones for their feedback and valuable comments. Furthermore, we thank editor Michael Russo for his excellent guidance throughout the process and the anonymous reviewers for their valuable feedback.

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.

ORCID iDs

Frank Schiemann

Rüdiger Hahn

Notes

Author Biographies

Daniel Reimsbach holds a tenured position at Nijmegen School of Management at Radboud University Nijmegen, the Netherlands. He previously held positions at the Universities of Düsseldorf and Hohenheim, both Germany. His research focuses on the effects of financial and non-financial reporting, including CSR reporting and carbon disclosure on investors’ decision making. He published in internationally renowned journals such as European Accounting Review, Business & Society, Business Strategy and the Environment, Organizations & Environment and many others.

Frank Schiemann is a professor of Accounting at the Faculty of Business, Economics and Social Sciences of the University of Hamburg, Germany. His research focusses on the quality, determinants, and consequences of non-financial reporting, such as corporate social responsibility disclo-sure and carbon disclosure. His work has been published in leading international journals such as Business & Society, European Accounting Review, Journal of Accounting and Public Policy, Organization & Environment, Organizational Research Methods and many others.

Rüdiger Hahn is the Henkel-Endowed Chair of Sustainability Management at the Heinrich Heine University Düsseldorf, Germany. He previously held chaired positions at the Universities of Hohenheim and Kassel, both Germany. His research focuses on non-financial reporting, stakeholder behavior, hybrid businesses, and the Base of the Pyramid. He published in internationally renowned journals such as Business & Society, International Business Review, Journal of Business Ethics, Journal of Industrial Ecology, Organization & Environment and many others. Twitter:

Eric Schmiedchen was a PhD student at the Faculty of Business, Economics and Social Sciences of the University of Hamburg, Germany. Currently, he is a senior investor relations manager at OMV. His research focusses on the question how materiality of non-financial information is understood and considered by companies and capital market participants.

References

Adams

Simnett

(2011). Integrated reporting: An opportunity for Australia’s not-for-profit sector. Australian Accounting Review, 21, 292-301.

Ascui

Lovell

(2011). As frames collide: Making sense of carbon accounting. Accounting, Auditing & Accountability Journal, 24, 978-999.

Bansal

Knox-Hayes

(2013). The time and space of materiality in organizations and the natural environment. Organization & Environment, 26, 61-82.

Barkemeyer

Figge

Hoepner

Holt

Kraak

J. M.

P.-S.

(2017). Media coverage of climate change: An international comparison. Environment and Planning C: Politics and Space, 35, 1029-1054.

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Buckless

F. A.

Ravenscroft

S. P.

(1990). Contrast coding: A refinement of ANOVA in behavioral analysis. Accounting Review, 65, 933-945.

Burbano

V. C.

(2016). Social responsibility messages and worker wage requirements: Field experimental evidence from online labor marketplaces. Organization Science, 27, 1010-1028.

Byrch

Milne

M. J.

Morgan

Kearins

(2015). Seeds of hope? Exploring business actors’ diverse understandings of sustainable development. Accounting, Auditing & Accountability Journal, 28, 671-705.

Chan

C. C. C.

Milne

M. J.

(1999). Investor reactions to corporate environmental saints and sinners: An experimental analysis. Accounting and Business Research, 29, 265-279.

10.

Cho

S. Y.

Lee

Pfeiffer

R. J.

(2013). Corporate social responsibility performance and information asymmetry. Journal of Accounting and Public Policy, 32, 71-83.

11.

Cianci

A. M.

Kaplan

(2008). The effects of management’s preannouncement strategies on investors’ judgments of the trustworthiness of management. Journal of Business Ethics, 79, 423-444.

12.

Daudigeos

Valiorgue

(2011). Conditions for value creation in the marketplace through the management of CSR issues: A negative external effects framework. Business & Society, 50, 28-49.

13.

Deegan

Rankin

(1997). The materiality of environmental information to users of annual reports. Accounting, Auditing & Accountability Journal, 10, 562-583.

14.

Dunn

M. B.

Jones

(2010). Institutional logics and institutional pluralism: The contestation of care and science logics in medical education, 1967–2005. Administrative Science Quarterly, 55, 114-149.

15.

Eccles

R. G.

Krzus

M. P.

Rogers

Serafeim

(2012). The need for sector-specific materiality and sustainability reporting standards. Journal of Applied Corporate Finance, 24(2), 65-71.

16.

Edgley

Jones

M. J.

Atkins

(2015). The adoption of the materiality concept in social and environmental reporting assurance: A field study approach. British Accounting Review, 47, 1-18.

17.

Elliott

W. B.

Jackson

K. E.

Peecher

M. E.

White

B. J.

(2014). The unintended effect of corporate social responsibility performance on investors’ estimates of fundamental value. Accounting Review, 89(1), 275-302.

18.

Elliott

W. B.

Krische

S. D.

Peecher

M. E.

(2010). Expected mispricing: The joint influence of accounting transparency and investor base. Journal of Accounting Research, 48, 343-381.

19.

Evans

J. St. B. T

. (2011). Dual-process theories of reasoning: Contemporary issues and developmental applications. Developmental Review, 31, 86-102.

20.

Evans

R. W.

Davis

W. D.

(2011). An examination of perceived corporate citizenship, job applicant attraction, and CSR work role definition. Business & Society, 50, 456-480.

21.

Fasan

Mio

(2017). Fostering stakeholder engagement: The role of materiality disclosure in integrated reporting. Business Strategy and the Environment, 26, 288-305.

22.

Ghosh

(2012). The effect of positive and negative financial and nonfinancial performance measures on analysts’ recommendations. Behavioral Research in Accounting, 24(2), 47-64.

23.

Global Reporting Initiative. (2016). GRI 101: Foundation. Amsterdam, Netherlands: Author.

24.

Global Sustainable Investment Alliance. (2015). 2014 Global sustainable investment review. Retrieved from http://www.gsi-alliance.org/wp-content/uploads/2015/02/GSIA_Review_download.pdf

25.

Goel

. (2007). Anatomy of deductive reasoning. In: Trends in Cognitive Sciences, 11, 435-441. doi: 10.1016/j.tics.2007.09.003

26.

Greening

D. W.

Turban

D. B.

(2000). Corporate social performance as a competitive advantage in attracting a quality workforce. Business & Society, 39, 254-280.

27.

Griffin

P. A.

Lont

D. H.

Sun

E. Y.

(2017). The relevance to investors of greenhouse gas emission disclosures. Contemporary Accounting Research, 34, 1265-1297.

28.

Grushina

S. V.

(2017). Collaboration by design: Stakeholder engagement in GRI sustainability reporting guidelines. Organization & Environment, 30, 366-385.

29.

Hahn

Kühnen

(2013). Determinants of sustainability reporting: A review of results, trends, theory, and opportunities in an expanding field of research. Journal of Cleaner Production, 59, 5-21.

30.

Hahn

Lülfs

(2014). Legitimizing negative aspects in GRI-oriented sustainability reporting: A qualitative analysis of corporate disclosure strategies. Journal of Business Ethics, 123, 401-420.

31.

Hicks

E. L.

(1964). Materiality. Journal of Accounting Research, 2, 158-171.

32.

Holm

Rikhardsson

(2008). Experienced and novice investors: Does environmental information influence investment allocation decisions? European Accounting Review, 17, 537-557.

33.

Jones

D. A.

Willness

C. R.

Madey

(2014). Why are job seekers attracted by corporate social performance? Experimental and field tests of three signal-based mechanisms. Academy of Management Journal, 57, 383-404.

34.

Khan

Serafeim

Yoon

(2016). Corporate sustainability: First evidence on materiality. Accounting Review, 91, 1697-1724.

35.

Kuruppu

Milne

M. J.

(2010). Dolphin deaths, organizational legitimacy and potential employees’ reactions to assured environmental disclosures. Accounting Forum, 34(1), 1-19.

36.

Lachmann

Stefani

Wöhrmann

(2015). Fair value accounting for liabilities: Presentation format of credit risk changes and individual information processing. Accounting, Organizations and Society, 41, 21-38.

37.

Libby

Bloomfield

Nelson

M. W.

(2002). Experimental research in financial accounting. Accounting, Organizations and Society, 27, 775-810.

38.

Matsumura

E. M.

Prakash

Vera-Muñoz

S. C.

(2014). Firm-value effects of carbon emissions and carbon disclosures. Accounting Review, 89, 695-724.

39.

Menzel

(2013). Are emotions to blame?—The impact of non-analytical information processing on decision-making and implications for fostering sustainability. Ecological Economics, 96, 71-78.

40.

Metcalfe

Mischel

(1999). A hot/cool-system analysis of delay of gratification: Dynamics of willpower. Psychological Review, 106, 3-19.

41.

Mohr

L. A.

Webb

D. J.

(2005). The effects of corporate social responsibility and price on consumer responses. Journal of Consumer Affairs, 39, 121-147.

42.

Moroney

Trotman

K. T.

(2016). Differences in auditors’ materiality assessments when auditing financial statements and sustainability reports. Contemporary Accounting Research, 33, 551-575.

43.

Murray

Sinclair

Power

Gray

(2006). Do financial markets care about social and environmental disclosure? Accounting, Auditing & Accountability Journal, 19, 228-255.

44.

O’Dwyer

Unerman

Hession

(2005). User needs in sustainability reporting: Perspectives of stakeholders in Ireland. European Accounting Review, 14, 759-787.

45.

Orlitzky

Schmidt

F. L.

Rynes

S. L.

(2003). Corporate social and financial performance: A meta-analysis. Organization Studies, 24, 403-441.

46.

Osman

(2004). An evaluation of dual-process theories of reasoning. Psychonomic Bulletin & Review, 11, 988-1010.

47.

Purdy

J. M.

Gray

(2009). Conflicting logics, mechanisms of diffusion, and multilevel dynamics in emerging institutional fields. Academy of Management Journal, 52, 355-380.

48.

Rampl

L. V.

Opitz

Welpe

I. M.

Kenning

(2016). The role of emotions in decision-making on employer brands: Insights from functional magnetic resonance imaging (fMRI). Marketing Letters, 27, 361-374.

49.

Reimsbach

Hahn

(2015). The effects of negative incidents in sustainability reporting on investors’ judgments-an experimental study of third-party versus self-disclosure in the realm of sustainable development. Business Strategy and the Environment, 24, 217-235.

50.

Reimsbach

Hahn

Gürtürk

(2018). Integrated reporting and assurance of sustainability information: An experimental study on professional investors’ information processing. European Accounting Review, 27, 559-581.

51.

Shen

Chand

(2017). The impact of corporate social responsibility assurance on investor decisions: Chinese evidence. International Journal of Auditing, 55, 467.

52.

Slovic

Finucane

M. L.

Peters

MacGregor

D. G.

(2004). Risk as analysis and risk as feelings: Some thoughts about affect, reason, risk, and rationality. Risk Analysis, 24, 311-322.

53.

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.

54.

Solomon

J. F.

Solomon

Norton

S. D.

Joseph

N. L.

(2011). Private climate change reporting: An emerging discourse of risk and opportunity? Accounting, Auditing & Accountability Journal, 24, 1119-1148.

55.

Stanovich

K. E.

West

R. F.

(2000). Individual differences in reasoning: Implications for the rationality debate. Behavioral and Brain Sciences, 23, 645-726.

56.

Staubus

G. J.

(2000). The decision usefulness theory of accounting: A limited history. Routledge new works in accounting history. New York, NY: Routledge.

57.

Stefan

Paul

(2008). Does it pay to be green? A systematic overview. Academy of Management Perspectives, 22(4), 45-62.

58.

Steffen

Richardson

Rockström

Cornell

S. E.

Fetzer

Bennett

E. M.

. . . Sörlin

(2015). Planetary boundaries: Guiding human development on a changing planet. Science, 347, 1259855. doi:10.1126/science.1259855

59.

Sustainability Accounting Standards Board. (2017a). SASB Materiality Map™. Retrieved from http://materiality.sasb.org/

60.

Sustainability Accounting Standards Board. (2017b). SASB’s approach to materiality for the purpose of standards development (Staff Bulletin No. SB002-07062017). San Francisco, CA: Author.

61.

TSC Industries, Inc. v. Northway, Inc., 426 U.S. 438 (June 14, 1976), No. 74-1471.

62.

Tschopp

Nastanski

(2014). The harmonization and convergence of corporate social responsibility reporting standards. Journal of Business Ethics, 125, 147-162.

63.

Turban

D. B.

Greening

D. W.

(1997). Corporate social performance and organizational attractiveness to prospective employees. Academy of Management Journal, 40, 658-672.

64.

Turker

(2009). How corporate social responsibility influences organizational commitment. Journal of Business Ethics, 89, 189-204.

65.

US SIF Foundation. (2018). Report on US sustainable, responsible and impact investing trends 2018. Retrieved from https://www.ussif.org/files/Trends/Trends%202018%20executive%20summary%20FINAL.pdf

66.

von Wallis

Klein

. (2015). Ethical requirement and financial interest: A literature review on socially responsible investing. Business Research, 8, 61-98.

67.

Wason

K. D.

Polonsky

M. J.

Hyman

M. R.

(2002). Designing vignette studies in marketing. Australasian Marketing Journal, 10(3), 41-58.

68.

Wolf

(2014). The relationship between sustainable supply chain management, stakeholder pressure and corporate sustainability performance. Journal of Business Ethics, 119, 317-328.