Introduction

Information about firms’ environmental, social, and governance (ESG) activities is increasingly deemed value relevant (e.g., Amel-Zadeh and Serafeim 2018; Grewal et al. 2019; Christensen et al. 2021).Footnote 1 Firms have responded to demand for ESG performance transparency by voluntarily issuing ESG reports. While the frequency and depth of these reports have grown dramatically, with 92 percent of S&P 500 firms issuing reports in 2020 (G&A Institute 2021), prior research shows that investors demand higher quality ESG disclosures (Cohen et al. 2015; Christensen et al. 2021).

Our study considers, in addition to a firm’s disclosure of forward-looking ESG targets, its disclosure of progress made toward those targets. Disclosed ESG targets aid investors in assessing a firm’s strategic direction. However, ESG targets alone (“target-only disclosures”) do not provide a full picture of ESG performance.Footnote 2 The inclusion of progress reporting improves the relevance and faithful representation of an ESG disclosure by making it more complete and improving its predictive and confirmatory value (see the "Background and related literature" section). As a result, a firm’s disclosure of progress toward its stated ESG goals should enhance the quality of the firm’s ESG reporting.

We conduct a survey experiment which uses responsible cocoa bean sourcing issues as a setting to examine whether a firm’s disclosure of progress alongside its targets (as compared to target-only reporting) influences investment decisions made by nonprofessional investors. Despite a large literature examining ESG reporting effects, any effect of progress reporting is ex ante unclear. On the one hand, progress reporting may incrementally affect investment allocation because higher quality ESG disclosures can reduce uncertainty and improve estimations of firm value. For example, Plumlee et al. (2015) find a positive relation between the quality of voluntary ESG disclosures and the cost of equity. Extant research also suggests that investors are skeptical of firms that do not follow through on their stated ESG values (see Martin 2019), an impression that could be avoided through progress reporting (the "The potential impact of progress reporting" section provides details).

On the other hand, for a variety of reasons, ESG progress reporting may have no significant effect on investment. For example, due to the affective nature of ESG activities, investment decisions may be influenced by the idea of eliminating child labor, regardless of the information content of a particular disclosure. Prior studies show that ESG reports are likely to elicit emotional responses (Elliott et al. 2014) and that investors focus not only on valuation effects, but on societal benefits as well (e.g., Martin and Moser 2016).

Previous research also demonstrates that investors have difficulty processing ESG disclosures (Martin 2019). The findings in Elliott (2006), for example, suggest that nonprofessional investors are unconsciously influenced by a firm’s emphasis on pro forma earnings in earnings announcements. Whether an ESG disclosure is value relevant depends critically on attributes such as presentation choices, as well as how investors acquire, evaluate, and weigh ESG information (Gödker and Mertins 2018). Further, reliably assessing the direction and magnitude of an ESG activity’s impact on firm value is challenging; for example, though a firm’s progress toward its ESG goals may contribute to the firm’s success (Flammer 2015), ESG activities are costly (Serafeim 2018). Finally, factors such as variation in measuring ESG performance and perceived materiality may influence the weight investors assign to progress disclosures in their capital allocation choices.

Our survey experiment participants consist of undergraduate and graduate business students enrolled at a private college in the Northeast region of the United States. Removing 41 participants who did not complete the survey and 39 participants who did not pass the two manipulation checks results in a sample of 119 participants. Participants are given a hypothetical $10,000 endowment and are asked to make an investment decision regarding the stock of a fictional chocolate manufacturing firm. First, each participant views brief financial highlights for the firm and determines an investment allocation as a percentage of their endowment (Fig. 1). As all participants are shown the same financial highlights, this first decision provides an investment baseline.

Fig. 1
figure 1

Financial highlights information. Displays the financial highlights provided to all survey experiment participants prior to making a baseline investment allocation decision

Full size image

Participants are then randomly assigned to view one of two versions of an ESG disclosure that discusses the firm’s activities with respect to child labor and other human rights issues in cocoa production (Fig. 2). Participants assigned to the control group view an ESG disclosure containing only targets whereas treated participants view an ESG disclosure that includes progress metrics in addition to those same targets. After viewing the ESG disclosure, participants have the opportunity to adjust their initial investment allocation. Our hypothetical ESG disclosures are motivated by the 2020 ESG reports for The Hershey Company and Mondelēz International (Figs. 3 and 4 of the appendix contain excerpts reflecting target-only and progress disclosures, respectively).

Fig. 2
figure 2

Survey experiment ESG target-only versus progress reporting disclosures. Figs. 2a and 2b show the ESG disclosure provided to participants who were randomly assigned to the control (target-only) and treatment (progress) groups, respectively

Full size image
Fig. 3
figure 3

Target-only disclosure excerpts from 2020 ESG reports. Figs. 3a and 3b provide ESG target-only disclosure excerpts from Mondelēz’s and Hershey’s 2020 ESG Reports, i.e., disclosures of targets without corresponding progress reporting. These excerpts were retrieved from Mondelēz International (2021, p. 9) and The Hershey Company (2021, p. 7), respectively

Full size image
Fig. 4
figure 4

Progress disclosure excerpts from 2020 ESG reports. Figs. 4a and 4b provide excerpts from Mondelēz’s and Hershey’s 2020 ESG Reports that capture ESG progress reporting, i.e., disclosures of progress alongside stated targets. These excerpts were retrieved from Mondelēz International (2021, p. 10) and The Hershey Company (2021, p. 30), respectively

Full size image

We then conduct a difference-in-differences (DiD) analysis to examine whether the firm’s reporting of ESG progress incrementally influenced our participants’ investment decisions. While our analysis shows that all participants increased their investment on average after viewing the ESG disclosure, we do not find evidence that the increase statistically differs between the treatment and control groups. The experiment results therefore suggest that ESG disclosures lead nonprofessional investors to increase their investment allocation similiarly on average irrespective of progress reporting. Responses to a post-experiment questionnaire indicate that treated participants were more likely than control participants to agree that the ESG disclosure they viewed provided progress metrics, providing evidence of the effectiveness of the manipulation.

Post-experiment questionnaire responses also provide insights suggesting that the null treatment effect is possibly attributable to participants not explicitly assessing disclosure quality (Elliott et al. 2020) and making value judgments based upon their affective reactions rather than deliberate information processing (Elliott et al. 2014). As examples, the progress disclosures did not lead participants to view the reported ESG activities as more relevant to financial or societal value, and there was no significant difference in treated and control participants’ perceptions of disclosure credibility. Participants also tended to perceive the activities discussed as important, suggesting an emotional connection to the issues. Further research is required to form definitive conclusions regarding the mechanisms underlying our study’s results.

We make several contributions to the literature. To our knowledge, this is the first study to investigate the effect of ESG progress reporting (relative to target-only reporting) on investment decisions. Our survey experiment, which finds no evidence that ESG progress reporting incrementally affects investment allocation choices, sheds additional light on nonprofessional investors’ perceptions and use of ESG disclosures. Our study suggests that more complete disclosures (i.e., those that report progress) do not significantly change retail investors’ capital allocation decisions, relative to disclosures of targets only. These insights are informative for managerial decision-making and can help make investors aware of biases and heuristics that may hinder their objectivity. This research further has relevance for standard setters and regulatory bodies, as well as for preparers of ESG disclosures, e.g., in contemplating presentation methods for making progress disclosures more salient to nonprofessional investors. The goal of ESG disclosures is to help investors understand how a firm manages their ESG risks and opportunities. If existing approaches for reporting progress do not contribute to this goal, other options should be considered; for example, explicitly linking metrics to their effects on earnings and cash flows.

We contribute to a stream of research examining how retail investors process voluntary ESG reports and factor ESG disclosures into their decision-making (e.g., Martin and Moser 2016; Guiral et al. 2020), and to the experimental voluntary disclosure literature more generally (e.g., Koonce et al. 2016). Overall, our results in combination with post-experiment questionnaire insights appear most in line with investors not explicitly assessing disclosure quality (Elliott et al. 2020) and using an affect-as-information heuristic to evaluate ESG disclosures (Elliott et al. 2014), though additional studies are necessary to draw definitive conclusions. Our study suggests that nonprofessional investment allocation choices are affected by the perceived social benefits of ESG activities (see, e.g., Hong and Kostovetsky 2012; Elliott et al. 2014; Martin and Moser 2016).

The remainder of our paper is organized as follows: The "Background and related literature" section provides background as well as discussions of prior literature and the potential effect of progress reporting on investment decisions. The "Setting" section provides an overview of our setting. The "Method" section describes our participants, study design, and survey experiment procedures. The "Findings" and "Additional tests" sections discuss results of our primary and additional analyses, and the "Conclusion" section concludes.

Background and related literature

Background on ESG target and progress disclosures

The creation of the International Sustainability Standards Board (ISSB) was announced in November 2021 by the International Financial Reporting Standards (IFRS) Foundation Trustees. The ISSB is a standard-setting board that is intended to deliver a global baseline of sustainability disclosure standards, i.e., the IFRS Sustainability Disclosure Standards (IFRS SDS). The ISSB Exposure Draft issued in March 2022 proposes that a firm disclose “the targets it has set to assess progress toward achieving its strategic goals” as well as “performance against its disclosed targets” within its general-purpose financial reporting package (ISSB 2022, p. 29). The objective of the IFRS SDS metrics and target disclosures is:

To enable users of general purpose financial reporting to understand how an entity measures, monitors and manages its significant sustainability-related risks and opportunities. These disclosures shall enable users to understand how the entity assesses its performance, including progress toward the targets it has set (ISSB 2022, p. 12).

In the current voluntary reporting environment, however, firms’ ESG reports do not always present progress alongside targets (Fig. 3 of the appendix provides examples). This presents an issue given that investors are limited in their information processing capacity and the mean length of a 2020 voluntary ESG report is 9312 words (Rouen et al. 2022) and about 70 pages, ranging between 12 and 243 pages (Filosa et al. 2021).

Drawing from accounting conceptual frameworks, reporting progress metrics alongside targets results in higher quality disclosures, as compared to reporting targets only.Footnote 3 By allowing investors to evaluate how a firm’s ESG performance compares with its targets, progress reporting improves the faithful representation of reporting; its inclusion facilitates more complete disclosures of ESG risks and opportunities. Disclosures that include progress metrics are also more relevant—they enhance predictive value (by increasing the probability that investors correctly predict firm outcomes) and confirmatory value (by helping investors confirm or correct their expectations).Footnote 4 For these reasons, it is possible that disclosures that include progress reporting may be more likely to influence investment decisions, relative to those that report targets only.

Related literature

Empirical research on the effects of voluntary ESG disclosures

An emerging and extensive empirical literature examines the effects of ESG reporting on a variety of outcomes (Christensen et al. 2021). Evidence from a number of prior studies suggests that the voluntary disclosure of ESG information is relevant in valuation assessments (e.g., Dhaliwal et al. 2011, Clarkson et al. 2013, Griffin and Sun 2013, Cheng et al. 2014, Matsumura et al. 2014, Bonetti et al. 2023, Grewal et al. 2021).

Other studies, however, find limited evidence of significant voluntary ESG disclosure effects, e.g., Cho et al. (2015) find no evidence that investors value corporate social responsibility (CSR) disclosures, and interpret their findings as suggesting that such disclosures are not value relevant because they are used as tools to enhance firms’ perceived legitimacy. That is, a firm’s ESG disclosures are intended to promote its image, e.g., to reduce exposure to societal or political pressures, rather than to cultivate meaningful accountability (see also Gray 2006). Consistent with a symbolic reporting approach, Michelon et al. (2015) examine 112 firms in the United Kingdom over the period 2005–2007 and find that, despite disclosing more CSR information, stand-alone CSR report issuers do not provide higher quality disclosures. Griffin et al. (2017) show similar effects on valuations for firms that voluntarily disclose greenhouse gas emissions compared to non-disclosers.

Prior archival evidence also demonstrates that any impact of reporting on investment choices may depend upon a number of attributes, such as the content and type of disclosure (e.g., Ng and Rezaee 2015; Plumlee et al. 2015). For example, Plumlee et al. (2015) find variation in the association between voluntary environmental disclosure quality and firm value components (cash flow and cost of equity) based upon the nature of the disclosure (i.e., whether it pertains to positive, neutral, or negative issues) and its type (i.e., whether it presents hard or soft information).

Experimental research on the effects of voluntary ESG disclosures

Findings from experimental studies suggest that challenges in information processing create barriers to investors’ use of ESG information in decision-making (Gödker and Mertins 2018; Martin 2019).Footnote 5 For example, Koonce et al. (2016) find that investors are misled by partial explanations of the firm’s financial outcomes, despite having knowledge that the explanation is incomplete. In addition, findings from Elliott et al. (2014) are consistent with the use of an affect-as-information heuristic, in which investors who view CSR disclosures unintentionally base their value judgments on affective reactions rather than deliberate information processing.Footnote 6 Similarly, Martin and Moser (2016) find that investors assign a higher value to firms that disclose a green investment (with a focus on its societal benefits), relative to firms that do not disclose a green investment, despite having awareness that the green investment is net costly. Additionally, the evidence in Elliott et al. (2020) is consistent with investors ascribing incremental value to higher financial reporting quality only when commentary in the auditor report reveals the auditor’s assessment about the firm’s reporting quality, suggesting that investors may not explicitly assess disclosure quality. Other experimental research finds little or no overall effect of voluntary ESG disclosures on investment allocation (e.g., Teoh and Shiu 1988; Milne and Chan 1999).

Still, evidence from other studies suggests that investors are sensitive to variation in firm disclosures (Martin 2019). For example, Hirst et al. (2003) find that prior-year reconciliations between estimated and actual results help investors distinguish between opportunistic and accurate forward-looking disclosures. In the ESG context, Johnson et al. (2020) show that nonprofessional investors assign a lower value to firms that disclose an emissions strategy that emphasizes purchasing offsets relative to one that emphasizes operational changes (for firms with historically below-average CSR performance). Evidence from Elliott et al. (2017) suggests that when CSR disclosures reflect an alignment between the strategy frame and presentation style, less numerically-inclined investors believe they are better able to process the information, leading to a positive emotional response and thus a greater investment allocation.

Overall, studies using experimental and empirical methods offer mixed evidence regarding the use of ESG information in investment decision-making. Prior literature also draws attention to challenges associated with investors’ perceptions and processing of voluntary ESG disclosures.

The potential impact of progress reporting

Whether a firm’s voluntary disclosure of ESG progress (relative to target-only disclosures) affects nonprofessional investors’ investment allocation choices is ex ante unclear.

Why progress reporting may influence investment decisions

Voluntary ESG disclosures may facilitate a more comprehensive assessment of firm value and thereby affect investment allocation (see Christensen et al. 2021). For example, progress reporting may reduce uncertainty about a firm’s prospects. A survey of senior investment professionals conducted by Amel-Zadeh and Serafeim (2018) finds that most investors view ESG information to be financially material, primarily because it aids in the assessment of a firm’s legal, regulatory, and reputational risks. Progress reporting may signal that a firm strategically manages their regulatory and reputational risk exposures (see, e.g., Blacconiere and Patten 1994).

In general, higher quality disclosures may assist investors in estimating firm value. Prior studies find, for example, a negative relation between the quality of voluntary ESG reporting and the cost of equity (Plumlee et al. 2015).Footnote 7 Further, findings in Elliott et al. (2020) suggest that investors inherently value the credible use of higher financial reporting quality. If an ESG disclosure includes progress metrics, investors who value high quality reporting may be less likely to conclude that the disclosure is symbolic, i.e., that it simply intends to “construct an image of commitment that is designed to positively influence stakeholders’ perceptions” (Michelon et al. 2015, p. 60). Voluntary disclosure theory predicts that firms with better ESG performance will produce higher quality disclosures to reveal their superior performance (Verrecchia 1983; Gödker and Mertins 2018).

Because forward-looking information is uncertain, progress reporting may enhance the credibility of ESG disclosures. That is, investors may view disclosures of targets only as more susceptible to opportunistic reporting incentives than those containing progress. Previous research suggests that investors value firms that follow through on their ESG values (Martin 2019), which may be demonstrated by progress reporting. Investors may also view disclosures containing both progress and target metrics as more credible as they are more precise than target-only disclosures.Footnote 8 Legitimacy theory predicts that firms with poor ESG performance will use ESG disclosures to influence the public perception of their ESG activities. Such firms will prefer incomplete ESG disclosures that are difficult to evaluate, i.e., to disguise their poor performance (Deegan 2002; O’Donovan 2002). If investors view more complete information as more legitimate, progress disclosures may affect investment decisions (incremental to target-only reporting).

For these reasons, it is possible that ESG disclosures containing both targets and progress metrics influence investment allocation decisions, relative to target-only disclosures.

Why progress reporting may not influence investment decisions

On the other hand, prior literature demonstrates that investors have difficulty evaluating firm disclosures (see the "Related literature" section). Elliott (2006), for example, shows that nonprofessional investors’ decisions are unconsciously influenced by a firm’s emphasis on pro forma earnings unless they are provided with a quantitative reconciliation of pro forma to GAAP earnings. Target-only disclosures, in addition to focusing on uncertain forward-looking events, provide only partial information. Though ESG progress disclosures constitute higher quality reporting compared to target-only disclosures (see the "Background on ESG target and progress disclosures" section), investors may fail to recognize ESG reporting quality. Further, even if investors perceive progress reporting as high quality, disclosure quality may not necessarily impact value estimates (e.g., Elliott et al. 2020).

In addition, evaluations of ESG metrics are complex regardless of their value relevance. While ESG progress may be strategically important to a firm’s success (e.g., Flammer 2015), ESG activities are often costly in terms of use of both financial and human resources (Serafeim 2018). In addition, the content of ESG reports varies widely across firms (e.g., due to industry factors, materiality, and disclosure content and presentation), which may influence the relative importance of progress reporting in investors’ estimations of future cash flows. Despite the potential for ESG disclosures to incorporate information relevant to fundamental value, their usefulness is hampered by the extent of such challenges even in the absence of opportunistic disclosure incentives. For example, a firm’s ESG report preparation tends to be costly, particularly in the presence of multiple and evolving reporting frameworks, and it can be difficult to verify and quantify ESG risks and assess their materiality (e.g., Radhakrishnan et al. 2018). Additionally, given their voluntary nature, ESG disclosures may overall lack credibility (e.g., Gray 2010) irrespective of progress reporting.

Prior studies further demonstrate that investors focus not only on the impact of ESG activities on financial returns, but also on their societal and moral implications. ESG reports often induce affective responses due to “their imagery-provoking and value-laden nature” (Elliott et al. 2014, p. 276). As discussed, Martin and Moser (2016) find that investors respond favorably to green investment disclosures despite knowledge that the investment is unprofitable, suggesting that they trade off wealth for societal benefits. In our setting, for example, it is possible that allocation decisions are based simply upon a positive emotional reaction to the notion of eradicating child labor, regardless of the disclosure’s content. For these and the above reasons, investors may not differentiate between target-only disclosures and disclosures that include progress reporting.

Overall, our discussion highlights that it is unclear whether investors perceive progress reporting as informative to their investment decisions, relative to disclosures of targets alone.

Setting

Child labor issues in cocoa production

The ESG disclosures used in our survey experiment are issued by a hypothetical chocolate manufacturing firm and focus on responsible sourcing in cocoa production—a major human rights concern. A report by the University of Chicago National Opinion Research Center found that in 2018–2019 there were 1.56 million child laborers working in the cocoa-growing regions of Ghana and Côte d’Ivoire, where over two-thirds of the world’s cocoa is produced (Wexler 2020), with 1.48 million children exposed to hazardous work (Sadhu et al. 2020). To satisfy the demands of large, multinational chocolate manufacturers, cocoa farms in these countries have been known to engage in human rights violations (Aktar 2013; Balch 2018, 2021; Chang 2021). Over the ten-year period through 2019, cocoa production increased by 62%, increasing the demand for workers (US Department of Labor 2021). The European Union and the United States are two of the largest importers of cocoa sourced from Côte d’Ivoire and Ghana, responsible for 3.4 million and 650,000 tons of cocoa imports in 2018–2019, respectively (Ermgassen 2021).

Global chocolate sales amount to an estimated $103 billion each year and the industry has spent $150 million over 18 years to address child labor (Whoriskey and Siegel 2019). The issue of child labor in cocoa farming gained prominence in 2001 when former senator Tom Harkin and former US representative Eliot Engel negotiated the Harkin–Engel Protocol, an international voluntary agreement with eight major chocolate companies, with the aim of eliminating the “worst forms of child labor” (International Labour Organization 2011). The nonbinding agreement kept the US government regulating the chocolate supply, as Eliot Engel had introduced legislation that would mandate a labeling system to disclose whether child slaves had been used in their cocoa production (Whoriskey and Siegel 2019). However, two decades later, these companies have failed to reach their targets. The 2005 deadline originally specified in the Harkin–Engel Protocol along with interim extended targets in 2010 and 2015 were missed. Firms also fell short of the latest (less ambitious) goal—to eradicate 70% of the worst forms of child labor by 2020 (Whoriskey and Siegel 2019; Walt 2020).

ESG reporting by chocolate manufacturers

Given the above issues, non-governmental organizations (NGOs) and other stakeholders increasingly call for chocolate manufacturers to responsibly source their cocoa beans and take actions to assist cocoa-growing communities. In response, multinational chocolate companies’ voluntary disclosures discuss programs aimed at addressing child slavery and other human rights concerns. Transparency about a firm’s social actions can influence capital allocation, as investors may not only assess the impact of these actions on financial value but may also make socially conscious decisions even when it is unprofitable (see the "Related literature" and "The potential impact of progress reporting" sections). In the consumer context, Aktar (2013) finds that, when consumers have limited awareness of social issues in cocoa production, a firm’s voluntary disclosure of positive or negative information about these issues does not affect their willingness-to-pay for the firm’s chocolate. Interestingly, the study also finds that when consumers are highly aware of these issues, even the disclosure of negative information is more beneficial than no disclosure if the firm expresses a commitment to eliminate its unethical practices.

Firms disclose this information in voluntary ESG reports and elsewhere, e.g., on their websites. To motivate the development of our experimental ESG disclosures, we review the 2020 stand-alone ESG reports of the two largest publicly traded US chocolate companies—The Hershey Company and Mondelēz International.Footnote 9 Hershey’s ESG report discloses, for example, information pertaining to their Cocoa for Good strategy and Child Labor Monitoring and Remediation System (The Hershey Company 2021), and indicates a commitment to investing $500 million in cocoa-growing communities by 2030. Similarly, Mondelēz’s report discusses their Cocoa Life program, including a $400 million commitment over ten years to “empower 200,000 cocoa farmers and improve the lives of more than one million people in cocoa communities” (Mondelēz International 2021, p. 43).

Of the 70 (60) total pages contained in Hershey’s (Mondelēz’s) ESG report, 25 (17) pages include discussions or metrics pertaining to social issues in cocoa production, i.e., 38.5% (28.3%) of pages. Our page count pertaining to social issues in cocoa production considers issues that concern human rights issues, e.g., cocoa sourcing from suppliers, child labor, forced labor, traceability, farming community economics and living wage issues, and farm safety conditions. Pages discussing purely environmental issues are not included in the count. While the ESG reports are voluminous, the disclosed metrics often repeat information contained on other pages and omit important traceability information. For example, they do not disclose sourcing details regarding cocoa bean supplier tiers, where tier-1 includes direct suppliers (e.g., cocoa bean exporters such as Cargill), tier-2 includes suppliers at the cooperative/farmer group level, and tier-3 includes suppliers at the farm level. Better traceability and transparency of firms’ cocoa supply chain is a key means of improving accountability in the cocoa and chocolate sector (IDH, GISCO, C-lever.org 2021).

Both ESG reports include various metrics capturing targets and the progress made toward those targets. In each report, certain portions contain target-only ESG disclosures whereas other portions disclose progress alongside targets (see Figs. 3 and 4 of the appendix). Because stand-alone ESG reports are typically lengthy and investors have limited information processing capacity, we are interested in whether capital allocation choices are influenced by the inclusion of progress reporting.

Method

Participants

Our target population was undergraduate and graduate business students from a private college in the Northeast region of the United States.Footnote 10 We used this population as their previous courses and knowledge make them relatively informed, but their lack of occupational experience makes them representative of a nonprofessional investor. The survey was designed in Qualtrics and distributed, following best practices (Dillman et al. 2014), to 829 students through personalized email invitations and four reminders during the Spring 2022 semester. Participants were incentivized with the opportunity to enter a raffle to win one of ten $25 Amazon gift cards. After the distribution, 199 participants responded to the survey, for a response rate of about 24 percent, which is in line with typical student survey response rates (Annabi et al. 2018; Cheng and González-Ramírez 2021). Dropping participants who did not finish the survey resulted in a sample of 158 participants. Table 1 presents sample characteristics for those who completed the survey.

Table 1 Sample descriptive statistics (N = 158)
Full size table

The sample includes both undergraduate and graduate students from every academic year. It includes a majority of students with grade point averages above 3 (equivalent to a B letter grade). Around a third of the sample is first-generation, which is in line with the population of the college. The race and ethnicity distribution includes mainly White students (about 70 percent), followed by Hispanic students (about 18 percent). About 56 percent and 43 percent of male and female students, respectively, completed the survey.

Table 2 includes sample statistics related to investment experience. The two most popular majors within the sample are finance and accounting, followed by marketing and management. We targeted this sample as we expect they have additional investing knowledge—about 34 percent took a high school-level finance or accounting course and more than 90 percent have taken at least one college-level finance or accounting course. Notably, about 53 percent of the sample reports experience investing in the stock market. Table 2 includes the distribution of cumulative investment amount ranges among those who have investing experience. Approximately 30 percent have invested between $1000 and $4999, about 13 percent have invested between $5000 and $9999, and about 14 percent have invested above $20,000.

Table 2 Investment-related sample descriptive statistics
Full size table

Design and procedures

Our survey experiment investigates the effect of ESG disclosures containing progress metrics (relative to target-only ESG disclosures) on investment decisions made by nonprofessional investors. We use a randomized experimental approach rather than an empirical approach to avoid selection problems inherent in examining the effects of voluntary ESG disclosures. That is, voluntary ESG reporting is endogenous as it is influenced both by firms’ voluntary ESG activities and their reporting choices (Christensen et al. 2021).

We first provide all participants with a description of the hypothetical chocolate manufacturer, Neville Premium Chocolates, Inc., along with its recent financial highlights. The financial highlights, presented in Fig. 1, include the firm’s sales growth, net profit margin, and price-to-earnings ratio for the most recent three years (2019, 2020, and 2021) as well as an analyst recommendation. The buy rating provides a heuristic for unsophisticated individual investors who tend to rely on expert recommendations (e.g., Daniel et al. 2002; Malmendier and Shanthikumar 2007; Kelly et al. 2011). This can alleviate cognitive load and reduce the time required to complete the survey, consistent with best practices (Dillman et al. 2014). Participants are then asked to allocate a percentage of a hypothetical $10,000 endowment to invest in the stock, which reflects the baseline investment decision. All participants, regardless of whether they are subsequently randomly assigned to view the target-only or progress ESG disclosure, review the same initial set of financial highlights in order to establish an investment baseline.

After making an initial investment decision, participants are randomly assigned to either the treatment or control group. Each is shown a disclosure that discusses the firm’s efforts toward eradicating child labor in their cocoa bean supply chain. We randomly assign participants to view one of two ESG disclosures: (1) A disclosure containing targets only (control); or (2) A disclosure containing both targets and progress made toward those targets (treatment), as shown in Fig. 2.

Both disclosures provide metrics related to: (1) attaining full coverage of child labor monitoring and remediation in farming communities where the firm’s cocoa beans are sourced, (2) investing $500 million in education for children in cocoa growing communities, and (3) increasing the number of children receiving daily nutritional support (granola snack packs) to 100,000. Importantly, both disclosures list the same activities and targets for 2025. The only difference is that the progress disclosure includes progress toward those targets as of 2021. Our experimental ESG disclosure does not make mention of applicable child labor regulations in Ghana and the Ivory Coast as such information was not provided in either Hershey’s or Mondelēz’s stand-alone 2020 ESG report. We expect that individual investors are unlikely to have this knowledge prior to making investment choices.

After viewing the ESG disclosure, each participant is given the opportunity to adjust their initial investment allocation. This design is intended to causally estimate any effect of ESG progress reporting (relative to only information regarding ESG goals) on capital allocation choices in this hypothetical experiment. Upon conclusion of the experiment, we ask participants a series of post-experiment questions to gain insight into any factors that may have influenced their decisions. Participants state their agreement, on a 5-point Likert scale, with seven statements related to the presented ESG disclosure. The survey concludes by collecting investment-related and demographic information.

Empirical methodology

Our main analysis is based upon the difference-in-differences (DiD) model in Eq. 1, where β3 represents the mean change in investment allocation from before to after viewing the ESG disclosure for treated participants (targets and progress) relative to control participants (targets only). A positive and statistically significant estimate of β3 would indicate that, relative to control participants, the ESG progress reporting induces treated participants to increase their investment allocation in the stock of the hypothetical chocolate manufacturing firm.

$${\text{Investment}}_{it} = \beta_{0} + \beta_{{1}} {\text{Treat}}_{i} + \beta_{{2}} {\text{Post}}_{t} + \beta_{{3}} {\text{TreatPost}}_{it} + \varepsilon_{it}$$
(1)

The dependent variable Investmentit equals the investment allocation, i.e., the percentage of the $10,000 endowment chosen by participant i in time t, where time t is either 0 for the initial investment or 1 for the final investment. Treati is an indicator variable equal to one if participant i is in the treatment group and equal to zero if participant i is in the control group (as described in the "Design and procedures" section). Postt is equal to one (zero) for the final (initial) investment decision. TreatPostit, the DiD variable, is equal to Treati × Postt. We then estimate an additional model that includes controls for demographic characteristics and variables related to schooling and investment experience. Data for control variables are retrieved from post-experiment survey questions (see Tables 1 and 2). We also estimate a model that includes participant fixed effects. Our regressions use robust standard errors to account for potential heteroskedasticity and autocorrelation.

Findings

Manipulation checks

Our survey includes two manipulation checks. After providing a baseline investment allocation, participants answer a multiple-choice question regarding their investment endowment: “In the previous scenario, what was the total amount of money you had available to invest?” The Qualtrics survey is set up such that participants are unable to navigate back in the survey to verify this information. Thus, responses to this manipulation check allow us to confirm whether the participant paid attention to the baseline investment task. We further include a post-experiment manipulation check to assess the attention paid to the ESG disclosure presented prior to the second investment decision. Specifically, we ask, “In the supplemental report you just reviewed, the company provided information about:” Participants were able to choose between the following answers: (A) Deforestation and responsible palm oil, (B) Greenhouse gas emissions and renewable energy, (C) Responsible cocoa sourcing and child labor (correct answer), (D) Water usage and biodiversity statistics, or (E) I don’t know.

As discussed, our initial sample included 199 participants and of those, 41 did not finish the survey. Removing 39 additional participants who failed one or both manipulation checks results in a sample of 119 participants.

Covariate balance

Prior to performing our analyses, we assess the covariate balance of the treatment and control groups. As Table 3 shows, the covariate means do not statistically differ, with the exception of the first-generation indicator variable, which has more representation within the treatment group.

Table 3 Covariate balance (N = 119)
Full size table

Overall effect of ESG reporting on investment allocation

Table 4 summarizes the average initial and final investment allocation for control and treated participants. For the control and treatment groups, the average initial investment is approximately 28 and 33 percent of the $10,000 endowment, respectively. The last column reports the difference in means between the two groups. As expected, the baseline investment allocation for treated participants is not statistically different from that for control participants. For the final investment allocation (after viewing the ESG disclosure), control and treatment groups allocate approximately 36 and 40 percent of the hypothetical endowment, respectively. The difference between groups for the final investment decision is also not statistically significant.

Table 4 Baseline and final investment allocations
Full size table

Both control and treated participants increase their allocation percentage in the firm after viewing the ESG disclosure. The control group’s average allocation increased by 7.33 percent (p < 0.01) and the treatment group’s average allocation increased by 6.45 percent (p < 0.01). The difference-in-differences estimate (− 0.885) is not statistically significant.

Regarding our participants’ allocation choices, many studies show that American retail investors lack rudimentary financial literacy and do not understand the value of portfolio diversification and other financial concepts (US Library of Congress 2011; US SEC 2012). Though the majority of our participants have investing experience and have taken college-level finance and accounting courses, a final allocation of over 35% of their endowment on average to a single investment is consistent with studies indicating that the average individual investor holds an underdiversified portfolio (Statman 2004; Barber and Odean 2013). In a survey of undergraduate college students, Annabi et al. (2018) find that only 51.54% of participants correctly answered a question about the basic concept of diversification, consistent with previous studies.

Results of main analysis

We estimate the three econometric regressions shown in Table 5. The first column presents results from estimating the difference-in-differences (DiD) model in Eq. 1 without controls. The second column reports results after adding controls to the first model. Lastly, we estimate a participant fixed effect panel regression and include the results in the third column. As expected, our regression results are consistent with Table 4. The − 0.885 estimated coefficient on TreatPostit is small in magnitude and not statistically significant. Thus, we do not find evidence that treated participants invest differently than

Table 5 Difference-in-differences regression results
Full size table

control participants, i.e., we find no evidence that progress reporting leads to a significant change in investment allocation (incremental to target-only reporting).

While not the focus of our study, it is noteworthy that participants increase their investment allocation by approximately 7 percentage points on average after being presented with the ESG disclosure (see Table 4). Thus, viewing ESG target disclosures (regardless of whether progress was presented alongside those targets) increased participants’ investment allocation. Our findings suggest that, while nonprofessional investors tend to allocate more capital after viewing a firm’s ESG disclosure, progress reporting does not incrementally influence their investment decisions over target-only reporting.

ESG disclosure perspectives and discussion

After the experiment concludes, we ask participants a series of 5-point Likert scale questions to gain insight into their perceptions of the firm’s ESG disclosure. Participants were asked to rate their agreement with seven statements. The statements and distribution of answers are summarized in Table 6. We ascertain the effectiveness of our ESG disclosure manipulation by asking participants to state their agreement with the statement “The supplemental report I reviewed after making my initial decision discussed progress made toward Neville’s societal goals.” Relative to the control group, the treatment group was significantly more likely to agree that the report discussed progress made toward the firm’s societal goals (p < 0.01).

Table 6 Post-experiment responses (N = 119)
Full size table

The remaining questions ask participants to rate their agreement with six statements regarding their perceptions of the ESG disclosure. As Table 6 shows, there is no statistically significant difference between treatment and control groups for these statements. More than half of the sample either strongly agreed or agreed that the supplemental report discussed activities that will have a positive impact on the firm’s value. Participants’ financial impact assessments may encompass subjective judgments concerning the tradeoffs between the long-term valuation benefits associated with responsible sourcing and the immediate negative profit and cash flow effects of the ESG expenses. Alternatively, these assessments may simply reflect expectations based upon affective reactions.

Over 80 percent either agreed or strongly agreed that the report discussed activities that will have a positive impact on society. Similarly, around three quarters of the sample felt the report discussed societal issues that are important to them. Interestingly, about 50 percent appeared indifferent as to whether the report made truthful claims. A sizable portion of the sample (32.8 percent of control participants and 51.7 percent of treated participants) had neutral perceptions with respect to whether the firm’s disclosure reflected its genuine intentions. Lastly, the majority of the sample agreed or strongly agreed that the ESG goals presented were achievable/attainable.

Our main test indicates that progress reporting does not incrementally affect participants’ investment allocation, though all participants increased their investment after viewing the ESG disclosure. These findings taken together demonstrate that viewing ESG disclosures increases nonprofessional investors’ investment allocation on average, irrespective of progress reporting. The absence of a significant treatment effect possibly aligns with investors failing to explicitly assess ESG disclosure quality, and the potential use of an affect-as-information heuristic. Though further research is needed to definitively form conclusions regarding underlying mechanisms, the below discussion aims to provide some exploratory insights.

For example, regardless of progress reporting, most participants perceived the ESG activities as positively affecting financial and societal value. Progress reporting therefore did not lead treated participants (relative to control participants) to view the ESG activities as more relevant with respect to future firm value or societal impact. Progress reporting further did not significantly affect perceived disclosure credibility, based upon responses to statements focused on the truthfulness of the disclosure’s claims and genuineness of its intentions. Importantly, participants on average do not hold strong beliefs about these statements, suggesting that the null treatment effect is not attributable to progress metrics lacking credibility. Interestingly, treated participants were less likely to agree that the ESG disclosure reflected the firm’s genuine intentions—41.4 (60.7) percent of treated (control) participants agreed or strongly agreed whereas 51.7 (32.8) percent neither agreed or disagreed. Similarly, investment allocation increased on average after viewing the ESG disclosure despite over half of participants neither agreeing or disagreeing that the disclosure made truthful claims.

Sample mean responses were highest for statements focused on social value and whether the societal issues were important to participants, suggesting an emotional connection to the ESG activities. In addition, though rational assessments of the likelihood of achieving targets should be influenced by progress, there is no significant difference between treated and control participants’ average perception of attainability. Participant views are charitable despite a lack of information available to evaluate whether targets are realistic—75.9 (85.2) percent of treated (control) participants either agree or strongly agree that the ESG goals were achievable. The similar perceptions of attainability for treated and control participants is potentially consistent with an affect-as-information explanation. That is, one might objectively have expected the company’s progress metrics (which indicate that it has already, with four years remaining, passed the halfway point for all targets) to instill confidence in its ability to achieve its goals.

Additional tests

We conduct further tests to explore potential effects for subsets of participants based on their ESG disclosure perceptions. For example, progress reporting may influence investment allocation among participants who view the disclosure as truthful. We estimate six additional regressions using the model in Eq. 2. For each of the six post-questionnaire perception responses discussed in the "ESG disclosure perspectives and discussion" section, Perceptioni is an indicator variable equal to one if the participant strongly agrees or agrees with the statement, and zero otherwise. For each regression, \(\widehat{\alpha }\) 7 is our main estimated coefficient of interest.

$$\begin{aligned} {\text{Investment}}_{it} = & \alpha_{0} + \alpha_{{1}} {\text{Treat}}_{i} + \alpha_{{2}} {\text{Post}}_{t} + \alpha_{{3}} {\text{Perception}}_{i} + \alpha_{{4}} {\text{TreatPerception}}_{i} \\ & + \alpha_{{5}} {\text{TreatPost}}_{it} + \alpha_{{6}} {\text{PerceptionPost}}_{it} + \alpha_{{7}} {\text{TreatPostPerception}}_{it} + \epsilon_{it} \\ \end{aligned}$$
(2)

Untabulated results show that \(\widehat{\alpha }\)7 is not statistically significant in any of the six regressions, suggesting that these perception variables do not differentially influence the effect of progress reporting on investment allocation.

Conclusion

Our research aims to better understand how forward-looking ESG target disclosures, compared with disclosures that also include progress metrics, influence investment decisions. We conduct a survey experiment to investigate whether ESG progress reporting affects capital allocation choices made by nonprofessional investors. While viewing ESG disclosures leads to greater investment allocation on average for both groups of participants, we do not find evidence of a significant incremental impact of progress reporting. While our main analysis, in combination with the post-experiment responses, possibly suggests that nonprofessional investors do not explicitly evaluate ESG disclosure quality and instead base their capital allocation decisions upon affective judgments, further research is required to arrive at more conclusive takeaways.

Several caveats and limitations apply to our study. First, our use of student participants may not be generalizable to the broader investor population. Professional investors may react differently to disclosures compared to nonprofessional investors, and prospective investors may react differently compared to current investors (Martin 2019). Additionally, our study involves the use of a hypothetical endowment, which may affect participants’ behavior. Lastly, our null findings should be interpreted with caution, as they could potentially be attributed to matters related to our experimental design. However, these null results present an opportunity for future research to further explore the impact of voluntary ESG target and progress disclosures on investors’ decision-making.