Investors’ Perceptions of Sustainability Reporting—A Review of the Experimental Literature

Abstract

Prior research has shown that companies’ sustainability reporting in relation to environmental, social, and governance disclosures influences investors’ investment decisions. Since the credibility of these disclosures is often questionable, it is important to understand how investors perceive sustainability reporting and include it in their decision-making process. Although the high relevance of this topic has already been clarified in research, the extant literature is heterogeneous and poorly connected on several levels. Against this background, we conducted a systematic literature review of 27 experimental studies on this topic published in leading accounting journals between 2000 and 2021. By clustering the results according to Mercer’s credibility factors, we synthesize the research on investors’ perceptions of sustainability reporting in a novel way, and derive suggestions for future research. We find that the interest in experimental research on sustainability reporting perception has grown in recent years. Researchers so far have examined sustainability performance and external assurance as the most relevant factors determining the credibility of sustainability information. Other factors, such as disclosure precision and inherent plausibility, are sparsely explored. We provide avenues for future research to investigate the perception of sustainability disclosures more comprehensively, by focusing on understudied credibility factors and on new theories and heuristics. Additionally, we suggest considering diverse experimental settings, such as different investor groups, company characteristics, or experimental procedures in general.

1. Introduction

External corporate reporting is undergoing broad changes. Stakeholders were previously mainly interested in companies’ financial reporting (FR), but there is growing interest in firms’ sustainability reporting (SR) [1]. Investors particular request SR, with 82% considering environmental, social, and governance (ESG) aspects in their decision-making processes [2]. The high interest in and usage of SR lead to numerous economic effects, such as reduced financing costs and a facilitated access to capital [3,4,5]. Investors, at the same time, often perceive SR as a marketing instrument that conveys a positive image of the organization and increases its legitimacy or strengthens its reputation [6]. Although the disclosure credibility (from now on, credibility) of ESG disclosures is often questionable, prior research indicates that SR changes investment behavior [7].

How investors perceive and consider corporate information in their decision-making process is an essential question in accounting, finance, and sustainability research. Unlike FR, SR has a partly qualitative character and underlies regulatory requirements. Diverse methods and a lack of clarity in international standards aggravate SR. Thus, significant differences exist between different firms’ SR, relating to the publication venue, scope, and precision of these sustainability disclosures [8,9,10].

Behavioral experiments help us to understand the role of SR in investors’ decision-making. By reducing the complexity and isolating effects in a controlled setting, researchers can investigate the impacts of various SR design aspects on investment behavior [11,12,13,14]. Behavioral experiments are purposeful, particularly in the context of the critical aspect credibility. Researchers can model credibility as a dependent variable and thus observe the effect of potentially influencing factors as independent variables. Specifically, Mercer (2004) distinguishes four conceptual factors: situational incentives, management credibility, external and internal assurance, and disclosure characteristics [14], which together can determine the credibility of SR perceived by investors.

Despite SR’s high economic importance for the sustainable development of financial markets, behavioral experimental research on this topic is far from mature. First, legislation is still under development, e.g., in Europe and the U.S. Changes in legislation lead to changes in reporting behavior, which induce a further need for research [15]. Second, several studies criticize the variable quality of behavioral experiments. This view is related to different issues, such as the lack of psychological theories as a fundament of experimental designs [12]. Moreover, extant studies have not developed a common understanding of SR’s most relevant constructs and determinants so far. This leads to problems in identifying, comparing, and interpreting different studies [16]. Third, the nature of investors’ decision-making processes and SR is complex [17]. Combining these two multifaceted aspects leads to various research approaches whose individual components are difficult to bring together.

Within the past few years, some researchers have synthesized previous findings and formulated proposals for future research. Malik (2015) reviewed literature that investigates the role of corporate social responsibility (CSR) disclosure in firm value [18]. Gödker and Mertins (2018) also synthesized CSR research [19]. Martin (2019) particularly studied experimental research on investor perception, but without solely focusing on SR [20]. We extend and complement this literature by addressing the following limitations: both Malik and Gödker and Mertins examined several empirical research methods in their reviews, such as experiments, archival studies, and surveys. Considering several methods enables a comprehensive overview of the empirical status in general, but at the same time, cannot shed light on the more nuanced aspects of experiments. In contrast, Martin solely analyzed behavioral experiments, but did not focus on SR. Due to the dominance of FR research, a broad synthesis and discussion of SR aspects are missing. Another limitation of the extant literature regards the publication date. None of the studies have considered articles published after 2017. Since SR is a highly relevant topic that is in constant transformation, for instance, due to changing legal regulations, there is a lack of recent experimental findings. Finally, all three studies review the literature in a predominantly qualitative way. In contrast, our bibliometric review integrates both qualitative and quantitative analyses. However, what makes our study stand out is that we focus on the crucial aspect of disclosure credibility. By clustering the results according to Mercer’s theoretical credibility factors, we synthesize this research on investors’ perceptions of sustainability reporting in a novel way. To the best of our knowledge, we are the first to include a comprehensive consideration of the credibility of SR. This allows us to gather new knowledge and discover new connections amongst the SR literature. Given this background, we conduct a systematic bibliometric literature review on behavioral experiments, investigating how investors perceive SR and consider it in their decision-making process. We focus on articles published between 2000 and 2021 in leading accounting journals as the primary outlets of this stream of research. The main aim of this review is to conceptually structure the experimental SR literature in order to (1) understand its current status, (2) identify research gaps, and (3) provide researchers with guidance on experimental designs to increase the validity and quality of their findings. As a result, we aim to answer the following two research questions: What (credibility) factors have already been studied and to what extent? What conclusions can we draw about the (interaction) effects of the factors?

To address the critical credibility aspect of SR, we structure this review by following Mercer’s credibility factors framework. This structure additionally enables us to:

• Understand what role SR credibility has in behavioral experiments;
• Synthesize credibility research on the credibility factor level, also including articles that do not focus specifically on investigating credibility factors;
• Explore the relationship between credibility factors, perceived credibility, and investment decisions;
• Identify gaps in credibility research.

Our findings show that current research on how investors perceive SR has grown in recent years. Although SR research is criticized for its great heterogeneity on several levels, we find that, according to experimental designs, there are a lot of patterns and similarities between studies. Interestingly, we find that all studies contain credibility factors within their experimental design. In other words, while some researchers purposely examine credibility in the SR context, other articles unconsciously provide results regarding the influence of credibility factors on investment decisions. Most experiments investigate the impact of SR performance and assurance on investors’ perceptions. We also find common patterns in theoretical foundations, company characteristics and investor characteristics. Considering that SR is a young research field, we show that researchers have already made a great contribution to the understanding of investor perceptions of SR. The findings show that investors initially perceive the SR of companies with high situational incentives as not credible, but other credibility factors, such as management credibility or assurance, can mitigate this effect. Our findings indicate that the combination of credibility factors determines how the investor views the overall credibility. SRs that are more credible influence the investment decision to a stronger extent than those that are less credible. Nevertheless, we also highlight key research gaps that should be investigated in the future. Especially in the area of multifaceted disclosure characteristics, more research is needed. For example, future behavioral experiments may investigate disclosure precision and inherent plausibility, since there are many quality differences between companies’ SRs in real life. Furthermore, researchers have rarely investigated more than two credibility factors in one scenario. Since context plays a significant role, it would be interesting to show how Mercer’s four credibility factors work together. We also intend to draw attention to the interplay between SR and FR. Since little research investigates the interplay between these two reporting types, we suggest future researchers might manipulate the credibility factors for both SR and FR to investigate spillover effects. This paper and its findings contribute to the literature in at least five ways, as it:

• Enriches the narrow but increasing literature on SR in accounting, finance, and sustainability research in a novel way by identifying new patterns amongst the literature;
• Provides researchers with potential research gaps and suggestions for future research on SR and SR credibility;
• Highlights that all experiments—explicitly or implicitly—examine credibility, and motivates researchers to be more aware and courageous in addressing this crucial aspect in future studies on SR;
• Supports lawmakers with an overview of evidence from behavioral experiments on how individual SR attributes (which are still legislatively under development) affect investor perceptions;
• Helps companies to understand their stakeholders’ perceptions and decision-making to provide higher-quality sustainability disclosures.

The rest of this paper is organized as follows: First, we provide the theoretical background of this review. This includes a clear terminology definition and an explanation of the credibility theory this review is based on. Next, we present our methodological procedure and highlight the search strategy step by step. We then present the results, followed by a comprehensive discussion. Finally, we provide a conclusion with limitations and implications for both practice and researchers.

2. Theoretical Background

2.1. Terminology

2.1.1. Sustainability Reporting

So far, no consistent definition of SR has been given in the extant literature. Instead, researchers use many terms synonymously, such as non-financial information, ESG disclosures, or CSR disclosures [16,21,22]. Following Erkens et al. definition of non-financial information, we use SR in this article to express:

“Disclosure provided to outsiders of the organization on dimensions of performance other than the traditional assessment of financial performance from the shareholders’ and debt-holders’ viewpoints.”

[16]

This definition includes, but is not limited to, ESG disclosures and CSR disclosure. However, this paper does not include information disclosed outside a report. For example, ESG information communicated in conference calls is not an object of consideration.

2.1.2. Credibility

Analogous to the inconsistency of the SR term, prior studies also use the concept of “credibility” in various ways [23]. Since we structure our review following Mercer’s credibility factors, we also refer to her credibility terminology:

“Disclosure credibility is defined as investors’ perceptions of the believability of a particular disclosure.”

[24]

Credibility does not necessarily have to represent reality or “real” credibility. Instead, credibility is an investor’s subjective perception. It is necessary to acknowledge that the extant literature often uses the terms “credibility” and “reliability” in almost the same way, although other strands of research differentiate these vocabularies [25]. Against this background, we consider studies that name reliability but mean credibility within this article. For consistency and to prevent confusion, we avoid using the two words synonymously, and use the term “credibility” throughout the paper.

2.2. Regulatory Requirements

To increase the understanding of SR and to interpret the results of the review in a broader context, it is indispensable to consider the regulatory requirements for SR. As with the regulations for FR, SR regulations vary considerably across countries [15]. The EU and the U.S. have both introduced new SR requirements in past years.

In April 2021, the EU adopted a proposal to replace the requirements under the Non-Financial Reporting Directive (NFRD). Currently, the NFRD requires large public-interest companies with more than 500 employees to publish environmental-, social-, and employee-related aspects within either their management report or a separate sustainability report. The proposed Corporate Sustainability Reporting Directive (CSRD) extends the scope and requires a more comprehensive SR, but only within the management report. In addition, the CSRD also introduces an external assurance requirement. There are two levels of assurance: reasonable and limited assurance. Reasonable assurance provides a high level of assurance, providing sufficient and appropriate evidence that the SR disclosures conform in all material respects with applicable legislation and accounting standards. Limited assurance provides a lower level of assurance, providing sufficient appropriate evidence that the SR disclosures are plausible. The proposed CSRD requires limited assurance. Unspecified EU considerations include a later movement toward reasonable assurance [26,27].

In March 2022, the United States Securities and Exchange Commission (SEC) also published a proposal to provide greater transparency for investors in SR. So far, SEC requires public companies to report certain ESG information if it is material to an understanding of the business. Mandatory sustainability disclosures are significantly lower in the United States compared to other markets. The new proposal extends the scope of especially climate-related reporting. It obligates companies to use mandatory standards to reduce the problem of selective reporting and a lack of comparability in SR. Similarly to the EU proposal, the SEC proposal also requires limited assurance first, considering a move toward reasonable assurance in the near future [28].

In summary, the regulatory changes in the EU and U.S. address several critical points in SR, such as the heterogeneity of SR across different companies. These changes at the legislative level provide additional evidence that SR is of high importance. However, it is essential to note that many companies still fall outside the scope of companies with SR obligations, and report voluntary sustainability disclosures without mandatory requirements.

2.3. Credibility Factors

In 2004, Mercer concretized the concept of perceived disclosure credibility by highlighting several of its influencing factors. She summarized her findings within a framework that has provided a theoretical foundation for numerous research articles in the field of FR, but also SR [29,30]. The framework consists of four credibility factors. These can be described as follows:

Situational incentives are the first factor in Mercer’s credibility factor framework. The situation of a firm at the time of its disclosure is crucial to how individuals perceive the credibility of the disclosure. Individuals perceive messages that are inconsistent with the source incentives as more credible than consistent disclosures, and vice versa. In scenarios with inconsistencies between information messages and situational incentives, individuals principally assume that the sources are communicating their underlying beliefs, instead of trying to benefit from misleading or untruthful messages.

Management credibility is Mercer’s second factor. It influences source credibility, which affects disclosure credibility. Individuals perceive disclosure credibility as higher in the presence of high management credibility. For instance, Hirst et al. (1999) found that when management provide accurate forecasts in the previous periods, this positively affects investors’ perceived management credibility, resulting in investors relying on and considering management disclosures more intensively [31].

The third credibility factor is external and internal assurance. External assurance is conducted by various external parties, such as auditors and industry experts. Internal assurance differs from external assurance in that it involves corporate governance mechanisms as a combined assurance approach, or internal audits. Disclosure credibility increases with the level of external and internal assurance.

Finally, Mercer’s last credibility factor is disclosure characteristics. Disclosure characteristics consist of several aspects. The precision differs significantly between different disclosures. Since imprecise disclosures signal that management is unwilling or incapable of providing better information, and individuals perceive high precision as more credible. For example, some companies provide qualitative sustainability disclosures with concrete quantitative information (high precision), while others only provide vague qualitative statements (low precision).

Next, corporate management publishes disclosures in various venues. This includes, for instance, financial statements, conference calls, and special press releases. Regarding SR, the company provides sustainability information either in the management report or as an additional sustainability report. The venue affects the perception, either positively or negatively.

Another factor is the time horizon. Mercer explains that information about immediate outcomes (such as interim earnings forecasts) is more certain than information on later outcomes (such as annual earnings forecasts), which results in shorter time horizons being more credible. According to SR, this dimension can also apply to concrete sustainability disclosures, such as short-term and long-term SR performance forecasts.

A significant amount of supporting information can also positively affect investors’ perception of credibility. For instance, a company that surprisingly changes its manufacturing site might provide supporting information, such as better working conditions for that strategic decision. One explanation is that supporting information can serve as ex-post verification of the disclosure.

Finally, inherent plausibility is also an essential factor that influences credibility. If inherent plausibility is not present, credibility decreases. This perception can be explained by individuals having individual expectations or prior beliefs when reading a piece of information. If the content of the information does not match these expectations or beliefs, individuals perceive the message as inherently implausible, resulting in skepticism, which leads to lower perceived credibility. For example, an individual could expect a sustainable fashion company to produce its clothes under high labor standards. However, if the company admits child labor issues in the SR, this can lead to inherent implausibility and skepticism. Figure 1 summarizes the framework graphically.

3. Materials and Methods

For this systematic bibliometric literature review, we developed a search strategy to identify potentially relevant articles and to limit the review scope to a manageable size. This strategy included the following three criteria: First, according to the research topic, this review only considered articles that address the investors’ decision-making process in relation to SR. Articles focusing on investors’ investment decisions or perceived credibility of SR were relevant in this context.

Second, this review only inspected articles published between 2000 and 2021. A time horizon of 22 years allowed the analysis of current research and simultaneously included older established articles. This enabled us to identify trends or changes in experimental designs over the past years without losing focus on the context for this review.

Third, this review only considered articles published in one of the most influential accounting journals. Since several studies have criticized the volatile quality of studies investigating SR, we followed previous literature and considered the article’s origin as a key criterion. Therefore, according to the ratings for accounting literature given by the German Academic Association of Business Research (VHB), only A+-, A-, and B-ranked journals were the object of consideration. The rating is the most influential journal assessment approach in German-speaking countries, and has assigned 35 journals an A+, A, or B grade. To keep the focus on the role of SR in investors’ decision-making process, unsuitable journals with different key areas, such as management accounting, were eliminated. Consequently, this review included the following 21 journals (in alphabetical order): Abacus, Accounting and Business Research (ABR), Accounting, Auditing, and Accountability Journal (AAA), Accounting, Organizations and Society (AOS), Auditing: A Journal of Practice and Theory (AJPT), Behavioral Research in Accounting (BRIA), Contemporary Accounting Research (CAR), Critical Perspective on Accounting (CPA), European Accounting Review (EAR), International Journal of Auditing (IJA), Journal of Accounting & Organizational Change (JAOC), Journal of Accounting and Economics (JAE), Journal of Accounting and Public Policy (JAPP), Journal of Accounting Research (JAR), Journal of Accounting, Auditing & Finance (JAAF), Journal of Business Ethics (JBE), Journal of Business Finance & Accounting (JBFA), Journal of International Accounting Auditing and Taxation (JIAAT), Review of Accounting Studies (RAS), Review of Quantitative Finance and Accounting (RQFA), and The Accounting Review (TAR).

Considering these three scoping criteria, we developed a search string (see Appendix A) and searched the full text of all articles published in the 21 selected journals by visiting the journals’ or corresponding journal publishers’ websites (since we have already identified the journals as part of the search strategy, we did not search a large database such as Web of Science). This search resulted in 1045 potentially relevant articles.

The next step appraised the 1045 articles by performing the initial inclusion check. The initial inclusion check aimed to identify potentially relevant articles within the review scope. Therefore, we analyzed the titles, abstracts, and keywords according to the following criteria.

• Must be an original research article;
• Must employ an experimental research method;
• Must address investors’ perceptions or decision-making processes relating to SR.

The initial inclusion check identified 57 potentially relevant articles. Finally, we performed the final inclusion step by screening the full texts of all 57 articles. In this screening procedure, the final inclusion check confirmed whether the articles were suitable for this review or not, considering the initial inclusion check criteria again. Since researchers use the term SR inconsistently, it was necessary to check the criteria within the full text again to ensure the papers were suitable. The final inclusion check found 13 relevant articles.

This review also considered a forward and backward search to avoid the risk of missing relevant articles. For the forward search, we identified all articles that had cited the 13 relevant articles. For the backward search, we collected all academic paper references for the 13 relevant articles. In the SR research, consensus on terms does not exist, and academic studies use different terms, although the studies refer to similar constructs. To raise the quality of the review, we performed forward and backward searches considering the identified initial and final inclusion criteria again. This additional procedure identified another 14 articles, resulting in a total of 27 relevant papers.

Table 1. Numbers of included and excluded articles per process step.

Description	Number
Articles identified with the search string	1045
Excluded after initial inclusion check	(−) 988
Potentially relevant articles after initial inclusion check	57
Excluded after final inclusion check	(−) 44
Articles included after final inclusion check	13
Additional relevant articles identified in the reverse research	(+) 14
Total number of articles included in this review	27

summarizes the numbers of included and excluded articles per process step.

4. Results

4.1. Overview

Between 2000 and 2021, 12 of the 21 examined journals published experimental articles on investors’ perceptions of SR. Overall, AJPT published the most articles (five), followed by EAR (four), JBE (four), AAA (three), CAR (three), ABR (two), AOS (one), BRA (one), IJA (one), JAPP (one), JBE (one), and TAR (one). Figure 2 shows the distributions of articles grouped in two-year periods, with the journal ranking.

We identified a growing number of publications over the last few years. While there was only one article published between 2000 and 2007, there has been at least one article published in each two-year period since 2008. Figure 2 shows two peaks. The first peak was reached in 2014–2015, with five articles (18.5%), followed by the second in 2020–2021 with nine articles (33.3%). These years are congruent with the publishing years of the NFRD, CSRD, and SEC proposals. Interestingly, the number of papers in A-ranked journals increased significantly in the second peak, while in the past two-year periods, a maximum of one paper was published in an A-ranked journal, and there were five A-ranked journal articles in the second peak. This indicates that academic researchers are increasingly interested in how investors perceive SR. In addition to the increasing number of publications, we discovered that the summed frequency of citations of these papers per year was also increasing. We identified that the citation frequency increased not linearly, but with an exponential growth rate. Figure 3 shows the increasing citation frequencies per year. While in 2016 there were only 207 citations, five years later, there were already 522 citations. This represents a growth of about 150% in 5 years. This finding also indicates that interest in investor perception of SR is increasing.

4.2. Research Topics

The independent variables used in the experimental studies provide insights into investors’ specific determinants of SR perception. Within the 27 relevant articles, researchers conducted 37 experiments. Figure 4 graphically shows the quantities of the six independent variables researchers examined most often in their behavioral experiments on sustainability disclosures. The SR performance is the independent variable most often used in the experiments (in 46% of all experiments, researchers manipulated this factor), followed by assurance (32%), explicit assessment (16%), disclosure characteristics (14%), non-financial strategy (11%), and investor characteristics (11%).

The dependent variables used in the experimental studies also provide insights into the investigated elements of investors’ SR perceptions. Figure 5 presents the six key investigated dependent variables within the 37 experiments.

The most frequently applied dependent variable measuring investors’ perceptions of SR is the willingness to invest (within 54% of all experiments, researchers observed this factor), followed by firms’ value estimate (22%), perceived credibility (19%), allocation decision (16%), perceived reliability (8%), and perceived importance (5%) of SR. The results show a strong focus on certain independent and dependent variables. A popular area of research on the level of independent variables is SR performance and assurance. In the case of the dependent variables, researchers prefer to measure investment-related variables (willingness to invest, value estimate, allocation decision), followed by credibility-related variables (reliability, credibility).

4.3. Theoretical Foundation

So far, many theories have been established regarding SR and investor perception. Because behavioral experiments were criticized in the 1980s and early 1990s due to their lack of psychological or economic theory, the theoretical foundation of experiments is of high interest [12]. Since the use of theoretical models can also reflect certain research trends,

Table 2. Theoretical psychological principles applied in the review articles.

Name	Description	Author(s) (Year)	Reference
Attribution Theory	Individuals evaluate the incentives of information sources to determine the message’s credibility. Individuals tend to assume that messages inconsistent with the source’s incentives are more credible than consistent messages [24].	Coram, P. J. et al. (2009)	[32]
		Wang, L./Tuttle, B. (2014)	[33]
		Brown-Liburd, H./Zamora, V. (2015)	[34]
		Cheng, M. et al. (2015)	[35]
		Cohen, J. et al. (2017)	[36]
		Shen, H. et al. (2017)	[37]
		Stuart, A. et al. (2019)	[38]
		Sheldon, M./Jenkins, J. (2020)	[25]
		Hoang, H./Trotman, K. (2021)	[39]
		Hoang, H./Phang, S. (2021)	[40]
Signaling Theory	The sender chooses whether and how to communicate (signal) the disclosures. The receiver must choose how to interpret the signal [41].	Cheng, M. et al. (2015)	[35]
		Baier, C. et al. (2021)	[42]
Affect-as- Information Theory	Individuals attend to the affect or feelings they are experiencing as a source of information to inform their judgment and, unintentionally, their subsequent decisions [43,44].	Elliott, W. B. et al. (2014)	[45]
		Elliott, W. B. et al. (2017)	[46]
		Guiral, A. et al. (2020)	[47]
		Austin, C. et al. (2020)	[48]
		Hoang, H./Trotman, K. (2021)	[39]

and

Table 3. Theoretical economic principles applied in the review articles.

Name	Description	Author(s) (Year)	Reference
Legitimacy Theory	Companies use SR to legitimize certain actions, change stakeholder perceptions or prevent negative stakeholder reactions [49].	Milne, M./Patten, D. (2002)	[50]
		Holm, C./Rikhardsson, P. (2008)	[51]
		Zahller, K. A. et al. (2015)	[52]
		Stuart, A. et al. (2019)	[38]
Stakeholder Theory	Companies use SR to manage and affect relationships with important stakeholder groups, such as investors [53].	Holm, C./Rikhardsson, P. (2008)	[51]
		van der Laan Smith, J. et al. (2010)	[54]
		Zahller, K. A. et al. (2015)	[52]
Agency Theory	Companies provide SR to reduce uncertainty among stakeholders because of information asymmetry, and stakeholders ask for sustainability information to pursue personal goals, such as ethical investing [55].	Holm, C./Rikhardsson, P. (2008)	[51]
		Zahller, K. A. et al. (2015)	[52]
		Stuart, A. et al. (2019)	[38]
		Austin, C. et al. (2020)	[48]
		De Villiers, C. et al. (2021)	[56]

provide an overview of the psychological and economic theories applied in this area of research.

In terms of psychological theories, researchers most often use the attribution theory, followed by signaling theory and affect-as-information theory. In terms of economic theories, researchers rely on legitimacy theory, stakeholder theory, and agency theory.

The review results show that each of the 27 review articles draws on at least one theory. Researchers sometimes refer to the main idea of a theory without mentioning the theory’s name. Additionally, the intensity with which researchers incorporate their theories varies from article to article. For example, Guiral et al. build their whole article on affect-as-information theory [47]. In contrast, Austin et al. mention affect-as-information theory only marginally in their article [48]. Therefore, in Table 2 and Table 3, we underline all articles that use a theory more intensively. Many researchers use not only one, but multiple theories, such as Hoang and Trotman using two psychological theories (attribution theory and affect-as-information theory). Although the results in Table 2 and Table 3 show that researchers predominantly focus on six theories, some articles use additional theories. These theories are usually very specific and rarely occur, such as cognitive cost theory and liberal feminist theory [48,57].

4.4. Credibility Factors

4.4.1. Situational Incentives

This chapter clusters the following content analysis into four sections, according to Mercer’s four credibility factors. The most common credibility factor researchers have integrated into their experiments so far is situational incentives. Eleven papers investigated the effects caused by this factor, via 14 experiments. The experiments mostly manipulated situational incentives for SR performance by distinguishing between positive and negative or strong and weak performance.

The recurring findings show that positive SR performance significantly lowers the investors’ perceived credibility. Consistent with attribution theory, individuals tend to assume that messages consistent with the source’s incentives are less credible than inconsistent messages. Since companies have an incentive to provide strong SR performance, reporting that strong performance leads to lower credibility. Strong SR performance also extends the positive effect of other credibility factors. For instance, high management credibility has a positive effect on SR credibility. Thus, the positive effect is stronger when SR performance is strong than when it is weak. Although investors question the credibility of positive SR performance, the value of investment decision estimates increases when SR performance is positive.

Besides SR performance, Pflugrath et al. manipulated firms’ industry type in their experiment. They compared firms in the retail industry with the mining industry, and provided evidence that the extent of the positive assurance effect on SR disclosures is industry-specific. For firms in the mining industry, credibility increases more strongly when SR is assured than for retail industry firms. Pflugrath et al. suggested that companies in the mining industry have more substantial incentives to report positive SR. Consistent with strong SR performance intensifying the positive effects of other credibility factors, industry types with high incentives also showed increased positivity in other factors [58].

4.4.2. Management Credibility

The second factor, management credibility, appears in eight experimental papers. In these experiments, this factor appears in various forms. The most intuitive way of modeling management credibility is probably the manipulation of corporate governance itself. Berens et al. showed that for stock preferences, good SR can compensate a poor corporate ability [59]. Wang and Tuttle showed that management credibility perception influences investors’ perceptions of financial disclosure credibility and stock price expectations [33]. However, contrary to these findings, the results from Cohen et al. demonstrate only a marginal influence of governance strength on investment decisions when SR performance is strong [36].

Another possibility for integrating management credibility into an experimental investigation is to model a press release. Hoang and Phang modeled a press article as either positive or negative [40]. Stuart et al. used a negative press article that was either present or absent [38]. Both were company-specific articles, so the researchers connected the press article directly with corporate management. The results show that a negative newspaper article lowers both credibility and willingness to invest. However, an assurance of the reported SR information can eliminate this effect.

4.4.3. External and Internal Assurance

In addition to SR performance, external assurance is the second aspect that researchers analyze most intensively. Eleven articles consider this aspect. Again, researchers manipulate different characteristics.

First, some studies manipulate the presence of assurance. The findings indicate that external assurance positively affects SR credibility. Studies also provide evidence that external assurance of SR raises stock price estimates. However, positive effects only appear when SR performance is positive. In other words, the presence of external assurance only has significant positive effects when a situational incentive is present.

Furthermore, several studies manipulate the assurance level by differentiating between reasonable and limited assurance. Here, the extant research has not provided consistent findings. Hoang and Trotman show that both limited and reasonable assurance increase credibility, but in value estimates, reasonable assurance has a stronger positive effect [39]. In contrast, Sheldon and Jenkins surprisingly identify that investors perceive positive performance reports with limited, but not reasonable, assurance as more credible than SR with no assurance [25].

Finally, experiments also examine external assurance’s impact by differentiating between external auditors and other parties as assurance providers. Pflugrath et al. show that the impact of different assurance provider types (professional accountant versus sustainability expert) on SR credibility is context-specific. While U.S. investors perceive higher SR credibility in the presence of a professional accountant, investors from Australia or the U.K. do not significantly differ in their SR credibility perception [41]. Shen et al. support Pflugrath et al.’s observations that perception is context-specific. However, contrary to Pflugrath et al., they find no significant impact of assurance provider type on investors’ decision-making [48].

One study has also examined internal assurance. Hoang and Phang show that combined assurance positively influences investors’ perception of credibility and willingness to invest in the presence of significant reporting credibility risks [40].

4.4.4. Disclosure Characteristics

Disclosure characteristics consist of multiple aspects. Researchers integrate these characteristics in various forms into their experiments. Nine studies consider this credibility factor.

The first subfactor of disclosure characteristics in Mercer’s framework is precision. Zahller et al. investigate high versus low precision in their study. Although they do not measure credibility or investment preferences, they show that participants perceive a high level of precision positively. Zahller et al. show that high precision leads to higher legitimacy, increasing the firm’s social resilience to exogenous shocks [52].

Zahller et al. conducted the only experimental study investigating precision, but more studies consider the subfactor of venue or, more precisely, reporting format for SR. Reimsbach et al. and Bucaro et al. examine the influence SR has on investor behavior when it is presented in a single report together with FR information, compared to separate reports. Since the NFRD offers discretion to companies on where they publish their NFI information, both are standard practices. Reimsbach et al. show that placing sustainability disclosures in a single report increases the investors’ access to sustainability disclosures, because a substantial number of professional investors showed no interest in a separate SR. However, according to investment attractiveness, they find that the choice of reporting format interacts with the assurance factor. In the case of SR without assurance, integrated reporting leads to higher investment-related judgments [57]. In contrast, Bucaro et al. show that SR has a more significant influence on investor judgments when SR appears in a separate report [60].

The third subfactor of disclosure characteristics is the time horizon. The analysis identifies no papers that examine this aspect. For the next subfactor, supporting information, Brown-Liburd and Zamora find that supporting information influences investors’ stock price assessments positively only when it is combined with independent assurance [30]. De Villiers et al. also investigate how different disclosure levels affect shareholders’ perceptions, and whether they are willing to pay for additional disclosures. Therefore, they manipulate financial, environmental, and social disclosures as either “minimum required by regulations”, “average for firms in the industry”, or “additional disclosures around future growth, risk, cash flow”. They show that shareholders are willing to pay for financial and environmental disclosures. They do not find evidence that shareholders are willing to pay for social disclosures [56]. The last subfactor is inherent plausibility. None of the experimental studies explicitly investigate this factor.

4.5. Further Characteristics

4.5.1. Investor Characteristics

Besides the credibility factors, other characteristics, such as investor characteristics, have an essential influence on investor perceptions [61]. Most experimental designs consider non-professional investors as participants. At least half of these designs use students as proxies for non-professional investors. Only a few articles investigate professional investors’ decision-making processes. A similar (but not equal) differentiation can be made among the experiences of individual investors. Two studies consider investors’ experiences within their experiments. Dong shows that SR disclosure affects highly experienced investors to a stronger degree than less-experienced investors [62]. Holm and Rikhardsson observed the same effects [51]. Nevertheless, the main effect of positive SR applies to all investors.

Besides the investor type, the nationality of the investors is also an important investor characteristic. Participants in the 27 examined papers are mostly from the U.S. Researchers also often focus on participants from other states, such as New Zealand, Australia, and the United Kingdom. The results show that differences between populations concerning investment behavior exist. For instance, Cohen et al. show that Lebanese participants appear more sensitive to weak SR performance than U.S. participants [36]. Pflugrath et al. show that the preferred assurance provider differs between countries [58].

Besides the participants’ nationality, researchers also investigate the role of investment horizon (long-term versus short-term) in the investors’ SR perception. Holm and Rikhardsson show that investors make different portfolio decisions based on different investment horizons [51]. Dong found that SR increases investment interest for investors with long-term investment horizons, but not for investors with short-term investment horizons [62].

Elliott et al. also investigated how the investors’ numeracy influences the investors’ perceptions. Results show that a fit between strategy frame and presentation style significantly increases less numerate investors’ willingness to invest, compared to when there is no fit, while more numerate investors are not affected [46].

4.5.2. Company Characteristics

The previous literature shows that in addition to investor characteristics, firm characteristics can also significantly impact investor perceptions [63]. Researchers have a clear choice of industries to assign to the fictitious companies in their experiments. They mainly use textile and mining companies. The experiments also sporadically consider other industries, such as Berens et al. using a financial service provider, or Shen et al. integrating the pharmaceutical industry [37,59]. Besides specifying the industry, most experiments provide specific information on the sustainability and financial performance of the fictitious company. Many experiments provide [58] positive and negative SR performance by manipulating SR performance as an independent variable. In the remaining articles, the sustainability disclosures are mostly positive. In terms of financial performance, researchers provide mostly positive financial information. Since a firm’s industry and SR performance can be understood as situational incentives, the findings have already been presented above. All in all,

Table 4. Theoretical economic principles in the review articles.

Field	Operationalization	Results
Situational Incentives	Sustainability Reporting Performance	Positive performance lowers credibility. It also extends the positive effect of other credibility factors.
	Firms’ Industry Type	Industries with more incentives to report positive disclosures have lower credibility. This also extends the positive effects of other credibility factors.
Management Credibility	Corporate Governance	Inconsistent findings. Corporate governance has a significant versus marginal effect on investment decisions.
	Press Release	A negative press release lowers both credibility and willingness to invest.
External and Internal Assurance	External Assurance (Presence)	External assurance positively affects credibility and raises stock price estimates. Positive effects only appear when situational incentives are present.
	External Assurance (Level)	Inconsistent findings. Both limited and reasonable assurance increase credibility versus limited assurance, but the lack of reasonable assurance increases credibility. Reasonable assurance has the strongest positive effect on value estimates.
	External Assurance (Provider)	Inconsistent findings. Assurance provider has no impact on decision-making, versus assurance provider has an impact that depends on the specific context.
	Internal Assurance	Combined assurance positively influences investors’ perception of credibility and willingness to invest.
Disclosure Characteristics	Precision	Participants perceive a high level of precision positively.
	Venue	Inconsistent findings. Sustainability disclosures have a more significant impact on investors‘ judgments when it appears in a separate versus integrated report.
	Time Horizon	No results.
	Supporting Information	Supporting information only has a positive impact on stock price assessment when it is combined with independent assurance. Shareholders are willing to pay for financial and environmental disclosures. They do not find evidence that shareholders are willing to pay for social disclosures.
	Inherent Plausibility	No results.
Investor Characteristics	Profession and Experience	The main effect of positive sustainability disclosures applies to all investors. However, it affects high-experience investors to a stronger degree.
	Nationality	Investors with different nationalities react differently to sustainability disclosures and differ in assurance provider preferences.
	Investment Horizon	Sustainability reporting increases investment interest for the long-term, but not for short-term investment horizons.
	Numeracy	While less numerate investors are affected by a fit between a firm‘s strategy frame and presentation style, more numerate investors are not affected.
Company Characteristics	Industry	See above (situational incentives)
	Reporting performance	See above (situational incentives)

summarizes the comprehensive results of this chapter (Appendix B provides a supplementary overview summarizing the research question and main results).

5. Discussion

In this chapter, we discuss the results of our review and put them in context with the previous literature. Our discussion follows our two research questions. The first research question is: What (credibility) factors have already been studied and to what extent?

Looking at the theoretical foundations first, all experimental papers refer to or use the ideas of theoretical concepts, with three psychological and three economic theories dominating the theoretical foundation in this research area. However, the review shows that researchers sometimes refer to the main idea of a theory without mentioning the theory’s name. In the future, researchers may consider naming the theory precisely so that other researchers can easily access and compare the empirical results. Since SR and investor perception benefit greatly from interdisciplinary collaboration, researchers may consider other theories from various research areas that are not, or are only rarely, considered in SR experiments, such as the proximity compatibility principle [57,64,65]. Researchers may also focus more on investigating possible heuristics, such as recognition heuristic and fluency heuristic [46,66,67,68].

According to Mercer’s credibility factors, we identified variables in each experiment that we could assign to the theoretical factors. Many researchers study credibility-related aspects only implicitly. We deduce this from the fact that although they did not use the term “credibility” or “reliability” once in their article, they have manipulated aspects that we identified as a potential operationalization of one of Mercer’s theoretical credibility factors. In contrast to the previous literature, we are the first to claim that every paper—whether credibility is studied implicitly or explicitly—makes a contribution to credibility research.

The review results show that researchers primarily manipulate SR performance and external assurance. Both manipulations are standard examples of the two credibility factors of situational incentives and assurance. Although not all studies contain credibility as a dependent variable in this context, there seems to be a general trend in the literature and prevailing relevance for both factors. The trend of investigating assurance may result from the fact that assurance was not required by regulation for a long time, although several parties, such as investors and politicians, demanded it. The trend of investigating SR performance may result from the fact that, for many firms, SR is voluntary, so it is questionable to what extent it is worthwhile for companies to publish positive or negative sustainability disclosures. As for firms, both ordering external audits and providing sustainability information are cost-intensive. Thus, future research might focus primarily on the financial implications rather than the credibility perception.

In contrast to, for instance, Martin (2019), who suggested investigating assurance in more detail, we think future research should use both forms of manipulation with caution [20]. Manipulating SR performance in terms of positive versus negative performance does not represent most SR in the real world. Instead, SR legislation demands realistic corporate reporting, since SR comprises positive and negative performance aspects. Since experiments provide evidence that observations in a positive performance setting fundamentally differ from those in a negative one, it is necessary to understand the role of mixed settings. Since it is questionable to what extent the results of both extremes can be applied to mixed scenarios, experiments should consider those mixed settings more intensively.

Manipulating external assurance gives the impression that researchers have focused on external assurance as a trust-building solution. This is consistent with legal developments, such as the proposed regulations in the EU and in the U.S., which require companies to provide external assurance as a credibility guarantee. Unfortunately, this solution involves several risks. First, as long as sustainability audit processes and SR legislation are still under development, external assurance only shifts uncertainties to a third party. The auditor might not have the experience and (technical) knowledge to ensure a high-quality audit of SR. An external audit can suggest more credibility to an investor than it actually provides. Second, notwithstanding the complex interplay of various credibility factors, treating external assurance as a panacea cannot overcome the lack of deep understanding regarding how investors perceive credibility.

While SR performance and external assurance are very common operationalizations for situational incentives and assurance, there is no single dominant operationalization for management credibility and disclosure characteristics. We suppose that in comparison to assurance and SR performance, these factors are less connected with financial implications and, consequently, have not been the subject of many experiments so far.

Besides these credibility factors, several additional factors, such as investor or company characteristics, influence how investors perceive SR. Researchers have mainly distinguished whether investors are non-professional or professional. Researchers have also differentiated investors by their expertise, nationality, or concrete characteristics, such as numeracy. We encourage researchers to explore the influence of more investor characteristics to understand SR’s impact better. For example, following research from the FR field, investors can be divided into equity and debt investors [61]. Since previous studies have primarily examined non-professional investors, we encourage more experiments to include professional investors. Non-professional investors are often represented by students, and thus allow for easy acquisition and low experimental remuneration. Nevertheless, professional investors differ substantially in their acquisition and perception of information. Since professional investors comprise a large portion of investment funds worldwide, we appeal to researchers to focus on professional investors with equivalent frequency. The motivations of investors also influence the decision-making process. For example, future experiments can investigate the influence SR has on company valuation vs. performance evaluation of managers.

Regarding company characteristics, we show that most experiments have manipulated SR performance (positive vs. negative). In the above discussion, we already recommended using mixed settings more often in the future. Besides SR, firms’ FR performance is mostly modeled as positive in the experiments. Therefore, we advise exploring poor or mixed FR performance in this context. Another company characteristic is industry affiliation. We included this aspect in conjunction with situational incentives because some industries have more situational incentives to report positive sustainability disclosures than others. Since the extant focus is mainly on the retail sector, we recommend more variety in this regard. For example, industries with questionable SR performance, such as the chemical or oil industry, could be examined.

After discussing what (credibility) factors have already been studied and to what extent, this chapter continues by discussing the second research question: What conclusions can we draw about the (interaction) effects of the factors?

For the most part, previous research has consistently found that investors initially perceive the SR of companies with high situational incentives as not credible. Other credibility factors can mitigate investors’ perceptions, whereby the other factors affect the perception more when investors initially perceive credibility to be low. The combination of credibility factors determines how the investor views the overall credibility. The experiments also provide evidence that SRs that are more credible influence the investment decision to a stronger extent than those that are less credible. We recommend that researchers consider additional credibility factors in future experiments to increase the generalizability of these findings. Disclosure characteristics in particular are very multifaceted, and have received little attention so far. Furthermore, researchers have rarely investigated more than two credibility factors together. Since context plays a significant role, it would be interesting to show how Mercer’s four credibility factors work together in one context. We also intend to draw attention to the interplay between SR and FR. So far, little research has been conducted on how FR credibility affects SR credibility, and vice versa. We therefore recommend manipulating the credibility factors for both SR and FR to investigate spillover effects.

6. Conclusions

In this paper, we review the experimental literature on how investors perceive SR. The main aim of this review was to conceptually structure the literature in order to (1) understand its current status, (2) identify research gaps, and (3) provide researchers with guidance on experimental designs to increase the validity and quality of their findings. To address the critical aspect of sustainability disclosure credibility, we additionally linked the review findings via Mercer’s four credibility factors. We focused on articles published between 2000 and 2021 in leading accounting journals as the primary outlets of this stream of research. We can summarize what we have learned from this review following our three aims.

First, in our attempt to understand its current status, we identified that the interest in exploring the investors’ perception of SR has grown in recent years. Besides publication frequency, we show that citation frequency has also increased. With regard to Mercer’s framework, all experiments of this review examine credibility factors. Despite the criticized heterogeneity in sustainability accounting research, our results show that there are similarities between the experimental designs, which we interpret as research trends. This involves, for instance, a focus on the impact of SR performance and assurance. Another trend is using one or two out of six prevailing theories as the foundation for the experiments, or to make the companies studied in the experiments similar according to industry or company performance. We show that there are also similarities in the results, meaning that the findings are not always, but are often, consistent. Overall, the experiments indicate that the combination of credibility factors impacts how investors perceive credibility. The credibility perception also influences investment behavior. More credible disclosures have a stronger impact than less credible ones. Whether or not investors perceive SR initially as credible depends on whether situational incentives exist. In the case of existing situational incentives, resulting in lower-credibility perceptions, the other credibility factors can mitigate this effect. Overall, we find that researchers have contributed much to understanding the effect of SR information on investors in recent years. However, we also note that the current state of research is currently insufficient to understand the effects of NFI across the board, because several major research gaps exist.

Second, in order to identify research gaps, we classified several aspects that experiments have not considered closely enough so far. The experiments do not, or only rarely, use other theories than the six dominant ones in this research area. Some credibility factors, such as inherent plausibility (disclosure characteristics), are not studied. Additionally, we found that little research has been done on how more than two credibility factors interact. Although FR research already shows that a wide range of company and investor characteristics influence investor perceptions, there is little knowledge regarding whether these characteristics also differ in SR scenarios. Research gaps also arise where the results of two studies provided inconsistent results.

Third, in order to provide researchers with guidance on experimental designs, we recommend that future research endeavors extend SR behavioral experiments by considering the research gaps that we have presented above. This includes, for instance, using new theories. We suggest providing more space for other credibility factors, but also more diversification according to investor and company characteristics. We also recommend that future researchers address aspects that have not previously yielded consistent results, such as presentation venue. Finally, we encourage researchers to be bold and develop diverse experimental designs to strongly promote research in SR perception by investors. At the same time, we appeal to researchers to consciously anticipate the credibility aspect and to target research gaps.

The unique contribution of our study lies in exploring the investor perception of SR in a novel way. By shedding light on the crucial factor of credibility, our study connects studies on a new level and generates new knowledge about the status of, as well as the gaps in, SR research. The extant literature in sustainability accounting and the regulatory changes often focus on external assurance as the credibility-building solution. We show that there exist more credibility aspects, and it is negligent to only consider assurance. This gives rise to the following implications for practice.

As regards policy implications, we suggest regulators and standard-setters consider all credibility factors within regulatory adjustments to address the multifaceted credibility aspect. This could be arranged in such a way that directives provide more precise specifications for the presentation of SR information (→ disclosure characteristics). Recent changes in the law affect this to a certain extent. For example, the CSRD states that companies are no longer free to choose the location, but must publish the SR together with the FR in a single report. However, besides the venue, there are other disclosure characteristics, such as the scope and precision, which the legislators should define more clearly. In turn, companies should also consider the interplay of all four credibility factors to improve the decision usefulness of the SR.

This study is subject to several limitations. First, we only consider experimental behavioral research. This decision allows us to analyze the experimental studies in more depth. It also allows us to derive detailed recommendations for future experiments. However, other empirical research, such as archival studies, could deepen our understanding and explain additional dimensions. Second, we focused on articles published in accounting journals as the primary outlet for research on SR perceptions by investors. Nevertheless, some relevant studies may have been published in other journals that were outside the scope of this review.

Regardless of these limitations, we expect our review to be useful to SR researchers as an overview and point of departure for debates on the future. In addition to the recommendations contributed to the literature listed above, this review has further implications for research. On the one hand, future reviews could try to overcome the limitations of this study by increasing the scope of experimental studies and linking them to the results from other studies derived with other methods. On the other hand, future research could enhance the heterogeneity between the SR papers, such as by using different wording, and develop concrete best practices for researchers.