Nonfinancial Corporate Social Responsibility Reporting and Firm Value: International Evidence on the Role of Financial Analysts

ABSTRACT

Using a large international sample of 24,293 observations from 3,991 unique firms in 56 countries, we examine the role of financial analysts in the relationship between voluntary corporate social responsibility (CSR) reporting and firm value. We find that after controlling for firms’ CSR performance ratings and other factors, voluntary CSR reporting increases firm value in countries worldwide, and analysts strengthen the positive relationship between CSR reporting and firm value. More importantly, our results show that the positive role of analysts in the relationship between CSR reporting and firm value varies with country-level institutional characteristics, such as the level of investor protection, the development of capital markets and the analyst profession, and stakeholder orientation. Furthermore, the positive role of financial analysts varies according to CSR reporting characteristics, including CSR reporting assurance, choice of assurer, CSR reporting coverage, CSR reporting quantity, and length of CSR report. We also find that financial analysts can strengthen the positive association between CSR reporting and firm value for CSR disclosures published in all types of media (standalone, annual report-based, and web-based CSR disclosures). Overall, our results present global evidence that shows the important role of financial analysts in improving the valuation implications of voluntary CSR reporting.

Acknowledgements

We thank Guochang Zhang (editor) and two anonymous reviewers for their helpful comments. We also thank Peter Clarkson, Mark Wilson, Sue Wright, Peter Wells, and conference participants at the 2020 Australian Summer Accounting Conference for their helpful comments on earlier versions of this paper. Kun Tracy Wang acknowledges financial support from the 2015/2016 Accounting & Finance Association of Australia and New Zealand (AFAANZ) Research Grant.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09638180.2022.2094435.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 For example, although some studies (e.g., Blacconiere and Northcut, 1997; Griffin and Sun, 2013; Matsumura et al., 2013) support the argument that investors consider voluntary nonfinancial disclosure provided by firms to be value enhancing, other studies show that the capital markets may not value CSR reports (e.g., Cho et al., 2015; Clarkson et al., 2019) and that in some circumstances, CSR reports even reduce firm value (e.g., Christensen et al., 2017; Chen et al., 2018).

2 For example, Dhaliwal et al. (2011) show that firms’ actual CSR performance is an important contextual factor that influences the effect of voluntary CSR disclosure on the cost of equity capital. Muslu et al. (2019) show that the effectiveness of voluntary standalone CSR disclosure in reducing information asymmetry varies with the quality of voluntary CSR disclosure (Christensen et al., 2022). Similarly, Chen et al. (2016) suggest that voluntary CSR disclosure issued by firms with higher audit fees accelerates the incorporation of future earnings information into current stock prices.

3 We obtain the CSR performance score from the ASSET4 data in our main tests. In our additional robustness test, we use CSR performance rating data from the Sustainalytics database and find that our conclusion is unchanged. Sustainalytics is another CSR database with global coverage; it is provided by Bloomberg, an alternative professional CSR performance rating agency.

4 The literature (e.g., Simnett et al., 2009) suggests that voluntary CSR disclosure with external assurance audited by Big 4 accounting firms or with higher CSR coverage is likely to cover broader CSR categories. Therefore, it is more difficult and costly for investors to process this information.

5 This finding is in line with the consensus suggesting that the more uncertain and complex the assessment of firms’ actions, the more likely it is that the opinions of other informed parties (such as analysts) will play a more influential role in these assessments (Graffin and Ward, 2010). This argument is plausible, especially considering that the largely voluntary and future-oriented nature of CSR reporting may increase investors’ reliance on analysts’ informational role and investor services (Lehavy et al., 2011).

6 For example, using data from 31 countries, Dhaliwal et al. (2012) find that issuing standalone CSR reports is associated with a decrease in analyst forecast error after considering financial reporting quality, suggesting that CSR reporting plays a crucial role in reducing information asymmetry in financial markets worldwide.

7 See “Value of Sustainability Reporting – A Study by EY and Boston College Center for Corporate Citizenship,” available at https://www.ey.com/Publication/vwLUAssets/EY_Value_of_Sustainability_Reporting/%24File/EY-Sustainability.pdf.

8 For example, CSR reporting serves as a valuable method for managers to signal their trustworthiness and communicate private information to investors about their firms’ CSR performance and future financial prospects (e.g., Hockerts and Moir, 2004; Lys et al., 2015; Christensen, 2016), which can positively affect investors’ expectations regarding future net cash flows.

9 Supporting this view, Petrenko et al. (2016) find that firms implement CSR activities for image management without truly focusing on improving their financial performance. Similarly, other studies find a negative relationship between environmental performance and environmental disclosure (e.g., Clarkson et al., 2008), which suggests that firms with low environmental performance tend to provide more environmental disclosure to manage investors’ impressions. Furthermore, firms may be motivated to report CSR activities even without substantive CSR activities because of the investment community’s increasing perception of corporate investment in CSR activities (Lee et al., 2018).

10 For example, Ryou et al. (2022) show that firms’ voluntary CSR disclosure increases in length over time and presents a large variation in readability across firms.

11 Hopwood contends that, “To the extent that such strategies work, it is possible that fewer questions might be asked of the legitimated organization and thereby less might be known of it.” (p.437, Hopwood, 2009.)

12 Survey data from CSR Europe, Deloitte, and Euronext (2003) also corroborate these findings, showing that approximately half of the fund managers and financial analysts surveyed indicate that they consider CSR information in their assessment of long-term firm value.

13 Corroborating the effect of financial analyst coverage on firms’ CSR performance, Jo and Harjoto (2014) show that the level of analyst following promotes value-enhancing CSR performance.

14 Using a survey of 401 US chief financial officers (CFOs), Graham et al. (2005) reveal that most CFOs are willing to sacrifice long-term investments to hit the short-term earnings targets set by financial analysts because of their own wealth and career concerns. He and Tian (2013) show that a higher level of analyst following hinders firm innovation.

15 According to ASSET4, CSR reporting data measure whether and when a company publishes a separate report on the environmental and social effects of its operations, a section of CSR information in its annual report, or information related to its CSR activities on its corporate website. As a robustness test, we construct CSR reporting variables using Sustainalytics data and find consistent results. According to Sustainalytics, CSR reporting data are created by analyzing a company’s CSR reporting and the extent to which it complies with international standards and best practices. In Section 4.2.3, we further test whether the role of analyst following in the relationship between CSR disclosure and firm value differs across CSR disclosures published in different media (standalone CSR reports, annual report-based CSR disclosures, and web-based CSR disclosures).

16 The four measures are as follows: (1) STAKELAW, which measures the legal environment of a country for the protection of labor rights; (2) CSRLAW, which measures the mandatory CSR disclosure requirement in a country; (3) PUBAWARE, which measures public awareness of CSR issues in a country; and (4) PUBAWARE1, which measures corporate executives’ awareness of CSR issues in a country. See Dhaliwal et al. (2014) for a detailed explanation of these four measures and the construction of the principal factor.

17 Our inferences are unchanged if we use industry-adjusted CSR performance scores.

18 Discretionary accruals (DA) are estimated using a performance-adjusted modified Jones model based on each country-industry-year cross-section (Kothari et al., 2005), with a higher value of DA indicating lower-quality financial reporting.

19 We exclude observations for financial firms because of regulatory differences with other firms in the sample.

20 Our inferences remain unchanged if we include firm-year observations without any analyst following in our sample; however, we cannot control for FError and FDispersion when including these observations in the sample.

21 In our robustness tests (see Sections 4.1 and 4.2.1), we exclude countries that contribute to a large percentage of the firms from our sample to address the potential large sample bias. In addition, we exclude all countries with fewer than 100 observations during the entire sample period from our analyses for robustness and find consistent results (untabulated).

22 In an additional test (untabulated), we further exclude all firms from the UK and Japan and find our conclusion unchanged.

23 The negative coefficient in Column 1 (nonsignificant coefficient in Column 3) on Report suggests that for firms with no analyst following (low level of analyst following), CSR disclosure generally has a negative (no significant) effect on firm value. This result is consistent with the argument that the market may react neutrally and even negatively to CSR disclosure for firms without sound intermediaries to interpret such information, because investors are unable to distinguish between greenwashing and substantive actions (Lyon and Maxwell, 2011) or to comprehend and incorporate CSR information into firm valuation because of the multidimensional and complex nature of CSR information (Luo et al., 2015; Christensen et al., 2021).

24 Our results also remain consistent if we exclude the UK and Japan, or countries with fewer than 100 observations.

25 We obtain consistent results when using LnFollowing and HiFollowing to measure the level of analyst following; therefore, for brevity, in all subsequent analyses we present only the results using LnFollowing as the key measure of the level of analyst following.

26 Supporting this assumption, Simnett et al. (2009) show that large firms and firms seeking to increase the credibility of their CSR reports are more likely to have their CSR reports assured and/or audited by a Big 4 accounting firm. Christensen et al. (2022) find that greater CSR disclosure leads to greater ESG rating disagreement among professional rating agencies, which is associated with higher return volatility, larger absolute price movements, and a lower likelihood of obtaining external financing.

27 We also conduct the analysis using the full sample by interacting Report_HiScope and Report_LoScope with LnFollowing, respectively, with nondisclosers serving as the base comparison group for Report_HiScope and Report_LoScope. We find that our inferences remain unchanged using this alternative sample and approach.

28 The negative coefficient of Report_HiScope supports the view that it is more difficult for investors to process and incorporate the information in CSR disclosure into firm valuation when a firm has a high level of CSR disclosure but no analyst following because greater CSR disclosure increases the possibilities of interpreting CSR information differently (Christensen et al., 2022), given its multidimensional and complex nature (Luo et al., 2015; Christensen et al., 2021).

29 For example, Bucaro et al. (2020) find that CSR information in standalone reports has a greater influence on investors’ decisions than that integrated into financial reports, because standalone CSR disclosure can help investors value CSR information by highlighting this information separately. de Villiers and van Staden (2011) note that firms tend to disclose vague, long-term environmental information integrated into financial reports, but they disclose timely and detailed environmental information on their websites.

30 The ASSET4 database provides a variable (ASSET4 Code: CGVSDP018) that indicates whether a firm provides nonfinancial ESG information in the management discussion and analysis section of its annual report.

31 To identify web-based CSR reporters, we follow previous research (e.g., Everaert et al., 2019) and review each sample company’s official website for disclosures relating to four general social areas: (1) fair business practices, (2) labor practices, (3) societal concerns, and (4) environmental issues. We classify a firm as a web-based CSR reporter in a year if its official website discloses information about at least one of the four general areas in the year. We do not include firms that only disclose general CSR information (e.g., their CSR strategies) on their website but do not provide the year of the disclosure, because we cannot attribute such CSR disclosures to a particular year.

32 Annual-report CSR reporters (standalone CSR reporters) that also provide CSR information in a standalone CSR report (annual report) or on their official websites in the year are excluded from this sample. This leads to an exclusion of 683 firm-year observations.

33 1,411 web-based CSR reporters that also provide CSR information in their annual reports in the year are excluded from this sample. We do not exclude web-based CSR reporters that also provide standalone CSR reports on their official websites because this is a very common practice; however, this will only bias against our finding of a significant difference in the role of analysts between web-based CSR reporters and standalone CSR reporters.

34 As with any study of the impact of regulatory shocks, one limitation of this analysis is that we cannot fully exclude the confounding effects of other unknown contemporaneous changes in other regulations when the mandatory CSR disclosure shock occurred. However, because our analysis includes five countries from geographically diverse locations (spanning Europe, Asia, and Africa), we suggest that any country-level unknown contemporaneous changes in non-CSR reporting regulations are unlikely to systematically affect our results.

35 According to Ioannou and Serafeim (2019), seven jurisdictions required CSR disclosure during our sample period: Denmark, China, Malaysia, South Africa, India, Hong Kong, and Brazil. Because Hong Kong and Brazil did not implement mandatory CSR reporting until 2016, which is the end of our sample period, we include only Denmark, China, Malaysia, South Africa, and India in this analysis.

36 Controlling for FError and FDispersion excludes the observations without any analyst coverage; thus, we do not control for the two variables in the model to allow the inclusion of firm-year observations with no analyst following. This leads to a larger sample for this analysis than the sample used in our main analyses. Excluding firms with no analyst following during the entire sample period from our final sample does not change our results.

37 To examine whether our results are sensitive to the choice of cutoffs and capture the largest firms in an economy that are more likely to be exposed to political pressures, we use an alternative cutoff by classifying a firm as large if its Size is in the top quartile of the country-industry distribution in a given year. We find that our inferences remain unchanged.

38 The following industries are classified as socially exposed industries: SIC codes 2111–2141 (tobacco products), 2812–2899 (chemicals and allied products), 1011–1499 (resource extraction), 2084–2085 (alcoholic beverages), 3761–3769 (defense), and 4011–4971 (utilities).

39 The literature highlights the insurance effects of CSR on firms’ financial performance during the crisis. For instance, Lins et al. (2017) find that firms with higher CSR performance tended to have higher stock returns than their peers during the financial crisis. In addition, as shown in Table 1, firms experienced substantial increases in CSR reporting during the latter half of our sample period, which was dominated by the 2008–2009 GFC.

40 $t=\frac{\overline{\text{β}}}{\mathrm{se}\left(\text{β}\right)}$, where $\overline{\text{β}}=\frac{1}{13}\sum _{t=2004}^{2016}{\widehat{\text{β}}}_t$

41 In untabulated tests, we also perform the same change analyses for the six CSR reporting properties that capture high CSR reporting scope (as defined in Section 4.2.2), and find that firms switching from CSR disclosure with low reporting scope to CSR disclosure with high reporting scope year in t-1 experience a significant increase in firm value in year t.

42 The three measures are NonDuality, BoardIndependence, and CorporateGovernance (see Appendix A for definitions). Firms with no CEO duality, or with a higher level of board independence or corporate governance score, are considered to have better corporate governance.

43 Because Sustainalytics starts its coverage in 2009 and our CSR reporting data must be lagged by one year, the first year of the sample for this additional analysis is 2010. The sample size of the tests using Sustainalytics CSR reporting data is less than one-third of our full sample because of its smaller coverage.