Mandated Social Disclosure: An Analysis of the Response to the California Transparency in Supply Chains Act of 2010

Abstract

In this study, we examine investor and firm response to the California Transparency in Supply Chains Act (CTSCA) of 2010. The CTSCA requires large retail and manufacturing firms to disclose efforts to eradicate slavery and human trafficking from their supply chains and is a rare example of mandated corporate social responsibility disclosure. Based on a sample of 105 retail companies subject to the CTSCA, we find a significant negative market reaction to the passing of the CTSCA. Furthermore, we find that the reaction is significantly more negative for larger firms and companies facing greater supply chain risks (apparel and footwear retailers), suggesting that investors place a negative value on exposure to legitimacy threats in the social domain. With respect to company disclosure response, we document relatively high compliance with the legislation, although we also find that the disclosure response appeared to be more symbolic than substantive in nature. Finally, our analysis indicates that both disclosure choice and disclosure extensiveness were significantly higher for the high-supply chain risk companies, suggesting that the response was influenced by concerns with strategic legitimation. Overall, the limited quality of disclosure suggests that, without additional rules and guidance, mandates alone may not lead to meaningful social disclosure.

Introduction

While the treatment of workers has long been an issue of ethical concern in Western economies (Krueger 2008), emphasis on the working conditions and potential abuses in supply chain operations, particularly those located in lesser developed countries, is a newer phenomenon. Fueled by exposures generated through both media investigations of “notorious labor practices in global factories” (Yu 2008, p. 513) and growing pressures from non-governmental organizations such as Amnesty International (Preuss and Brown 2012), the public has increasingly demanded corporations improve responsibility for the oversight of their supply chains. Firm response to these demands has included both the adoption of voluntary codes of conduct (see, e.g., Roberts 2003; Sethi et al. 2011) and attempts at being more transparent with respect to overseas factory locations (Doorey 2011). In this study, we focus on a different initiative related to oversight of potential human abuses in corporate supply chains, the California Transparency in Supply Chains Act of 2010 (CTSCA).

In 2010, the California State Legislature passed the CTSCA requiring certain large firms to provide on their websites disclosures related to efforts the companies are taking to eradicate slavery and human trafficking from their supply chains. The new reporting rules became effective January 1, 2012. Mandates for corporate social disclosure are extremely rare,¹ and we believe that the legislation provides an excellent opportunity to examine how both investors and affected companies responded to the new law. Prior studies, although limited to issues of an environmental nature, provide evidence that events increasing the social and political exposure² of companies are valued negatively by investors, and that differences in those exposures explain variation in the market reaction across firms (see, e.g., Bowen et al. 1983; Blacconiere and Patten 1994; Blacconiere and Northcut 1997). We are aware of no studies to date, however, that investigate market perceptions of potential legitimacy threats arising from non-environmental social concerns. Further, and owing largely to the limited existence of mandated corporate social responsibility (CSR) disclosure requirements, only a few recent studies explore company disclosure response to mandated requirements (Bebbington et al. 2012; Chauvey et al. 2015; Chelli et al. 2016). As such, we follow this prior research and examine the disclosure response from the perspectives of normativity—the degree to which actors see rules as binding (Chelli et al. 2016)—and legitimacy.

Based on a sample of 105 retail companies subject to the CTSCA, we document, first, significant negative market reactions to the final legislative events resulting in the passage and implementation of the law in 2010. We further find that the reaction is significantly more negative for apparel and footwear retailers—firms we argue face greater supply chain exposures—and larger firms, suggesting that the market negatively values the increased legitimacy threats imposed by the new legislation. We next reviewed the sample company websites over the first 10 days of January, 2012, and we found that 87 of the 105 sample companies made CTSCA disclosures, indicating a relatively high level of normativity. However, assessments of the extensiveness of information provided indicate that the disclosure response appeared to be more symbolic than substantive in nature. While most of the responding companies did include mention of all of the specific areas of concern laid out in the CTSCA requirements, extensive disclosure for any of the specific categories was quite limited, and only four firms included extensive information disclosure across all five required items. Further supporting the symbolic nature of the response, nearly half of the disclosing companies failed to comply with the legislation’s requirement for a prominent link on the website’s home page. Our analysis also reveals that both the choice to include CTSCA disclosures and the extent of information provided are positively related to higher supply chain risk (apparel and footwear retailers), suggesting that concerns with strategic legitimacy were at play in the disclosure choice.

In general, our findings both complement and extend prior research in the social and environmental accounting arena, and help also shed light on the ethical tensions companies face with respect to transparency regarding social exposures. First, the negative investor response to the passage of the CTSCA is consistent with the prior investigations of the market reaction to environmental-related events (e.g., Blacconiere and Patten 1994), and this suggests that the market appears to be as concerned with potentially increased legitimacy threats in the supply chain area as it is with those related to environmental issues. This also suggests that as firms consider their response to demands for more transparency with respect to social issues, in this case, information on efforts to reduce potential supply chain labor abuses, they may need to weigh the apparent investor beliefs that disclosure can have negative value implications. Indeed, this may help explain why, while we provide some evidence that the retail firms affected by the CTSCA did, for the most part, adopt the supply chain reporting, the actual information provided appeared to be quite symbolic in nature. Managerial concerns with investor perceptions of supply chain disclosure may have induced them to be less forthcoming in terms of details on their supply chain activities regarding protections against slavery and human trafficking. Given this potential managerial reluctance, it appears that, without additional rules and guidance, mandates alone may not be sufficient to bring about meaningful social disclosure and, more importantly, better supply chain safeguards. We begin with a discussion of the recent investigations of mandated CSR disclosure.

Background and Hypotheses Development

Prior Research

In this study, we build primarily on three recent investigations of CSR disclosure that bring the concept of normativity into their analyses. The first of these, Bebbington et al. (2012), focuses on “the ways in which actors come to see rules as binding” with respect to corporate environmental reporting by comparing the process of normativity in Spain and the U.K. They document low levels of compliance with Spain’s Plan General de Contabilidad (PGC), legislation mandating certain environmental disclosures, and argue that unless “norms are congruent with previous practice; a hierarchy of secondary rules define how the rule is to be made and applied; and the rule is well designed for its intended purpose,” and legal mandates will fail to meet the “test of legitimacy” necessary to invoke normativity (Bebbington et al. 2012, p. 79).³ Interestingly, they document that environmental disclosure in the U.K., driven by non-mandated incentives, exhibited higher levels of normativity than was the case in Spain.

Rather than focusing on the process of normativity, Chauvey et al. (2015) adopt what Chelli et al. (2016) refer to as an end-results perspective of normativity. That is, they attempt to determine the degree to which actors abide by the rules relative to mandated CSR disclosure requirements. Chauvey et al. (2015) explore the disclosure by French companies relative to the passage of the Nouvelles Régulations Économiques #2001-420 (NRE). Examining CSR disclosure in 2004 and again in 2010, Chauvey et al. report increases in both the space allocated to the topics and the quality of the information provided, a finding they argue suggests greater normativity toward the rules at the latter date. However, the analysis also shows that differences in disclosure in 2004 were associated with factors reflecting greater legitimacy exposures (firm size, industry membership, and levels of negative performance information) and that these relations continued to hold for the 2010 disclosures.

Most closely related to our analysis, Chelli et al. (2016) also take an end-results perspective and focus on changes in environmental disclosure for samples of French and Canadian companies. They note that in France the NRE, as well as requirements included in the Grenelle II Acts, represent official parliamentary legislation toward disclosure, whereas in Canada disclosure requirements are more limited and come from the Canadian Securities Administrators and thus represent a reliance on market mechanisms. Chelli et al. note that requirements in both countries are considered ‘soft laws’ in that they include only limited penalization for non-compliance. They report that the French firms showed significantly better improvement in information provision than the Canadian companies suggesting that governmentally mandated requirements appear to induce greater normativity.

Similar to Chauvey et al. (2015), Chelli et al. (2016) also assess disclosure from a legitimacy theory perspective, but they are more specific in their analyses. Chelli et al. argue that most prior environmental disclosure work drawing upon legitimacy theory views legitimacy as a strategic resource (strategic legitimacy) whereby managers use disclosure as a tool for garnering societal support. In contrast, they focus on an institutional view (Chen and Roberts 2010; Beck et al. 2015) and the notion of regulative legitimacy. From this perspective, firms achieve regulative legitimacy through compliance with regulations (Zimmerman and Zeitz 2002), and Chelli et al. argue that, because the sample firms (in both France and Canada) from environmentally sensitive industries did not exhibit higher levels of improvement, institutional legitimacy theory explains disclosure choice in their setting better than strategic legitimacy arguments. Following Hrasky (2012), Chelli et al. further examine whether the environmental disclosure response appeared to be a more symbolic or substantive approach toward legitimation. They note that substantive legitimacy is enhanced when companies describe corporate initiatives that lead to positive environmental outcomes, as opposed to merely providing information in an attempt to foster favorable perceptions of the organization. Chelli et al. find that, while French firms included more substantive disclosure than did the Canadian companies, provision of substantive information was still very limited. As such, while the French legislation appeared to bring about higher levels of normativity, the resulting disclosure remained largely symbolic in nature.

In general, both Chauvey et al. (2015) and Chelli et al. (2016) provide evidence that governmental mandates for CSR disclosure can induce normativity in the reporting. However, both studies’ evidence relates only to the French setting, and neither focuses on a specific area of social concern outside of the environmental arena. Accordingly, we extend this body of research by examining reactions to the CTSCA.

The California Transparency in Supply Chains Act of 2010

In 2010, the California State Legislature took on the issue of slavery and human trafficking in companies’ supply chains. Although a crime at state, federal, and international levels, the Legislature noted that the practice exists in every country, and even in the state of California, largely because it is hidden from view and difficult to uncover. The legislation states that without sufficient disclosure, consumers are unable to determine differences in the extent to which companies are making efforts to uncover and eradicate slavery and human trafficking in their supply chains, and thus may be inadvertently supporting its existence. Accordingly, the Legislature passed the CTSCA.

The CTSCA focuses on the slavery and human trafficking issue by requiring manufacturing and retail firms doing business in California and having worldwide sales in excess of $100 million to disclose on their web pages their efforts to eradicate the practice from their direct supply chains. The law specifically requires companies to address whether they (1) verify supply chains relative to slavery and human trafficking risk, and whether the verification was performed by a third party, (2) conduct audits of suppliers to ensure compliance with company standards on slavery and human trafficking, (3) require direct suppliers to certify that materials incorporated into the products comply with the laws of their country, (4) maintain standards and procedures for employees or contractors failing to meet company standards, and (5) provide training on slavery and human trafficking issues to employees and managers. Further, the law directs each affected firm to make this information available through a conspicuous link on the company’s home page. Companies failing to meet the new regulation’s requirements are subject to an action for injunctive relief. Although signed into law on September 30, 2010, the regulation did not take effect until January 1, 2012.

The CTSCA is one of only a limited number of mandated requirements for corporate social disclosure and accordingly offers an interesting case for examining how various parties responded to the new law. In particular, the passage of the legislation likely increased legitimacy concerns for the affected companies. Suchman (1995, p. 574) defines legitimacy as the perception “that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions,” and Milne and Patten (2002, p. 374) ague that where “the actual or perceived behavior of an organization departs from the social values and norms… its legitimacy is threatened.” Particularly to the extent that companies face exposures within their supply chains, requirements to be more transparent about their efforts to eliminate slavery and human trafficking from the supply chains increase the possibility that firms will be seen as departing from social values and norms with respect to this issue. Accordingly, in this study, we focus, first, on the investor response to the legislation, and second, company adoption of the reporting requirements.

Investor Reaction

Although limited almost exclusively to environmental issues, a number of prior studies document that events potentially increasing the regulatory costs or political exposure of firms are viewed negatively by market participants. For example, both Hill and Schneeweis (1983) and Bowen et al. (1983) explore the market reaction for utility companies following the Three Mile Island nuclear accident in 1979, and both show significant negative responses for those firms with nuclear power generation. Similarly, Blacconiere and Patten (1994) report a significant decline in market value for U.S. chemical firms following Union Carbide’s chemical leak in Bhopal, India in 1984. Blacconiere and Patten also show that companies with greater reliance on chemical sales suffered more negative reactions, while more extensive environmental disclosure prior to the event appeared to mitigate investor response. They thus argue that concerns with regulatory cost exposure drove the market reaction. Patten and Nance (1998), although finding increased market returns, on average, for petroleum firms following the 1989 Exxon Valdez oil spill in Alaska’s Prince William Sound, document that regulatory cost exposures again influence the reaction negatively in that larger companies and firms with operations in Alaska suffered less positive market reactions, while higher levels of environmental disclosure were associated with more positive adjustments. Similarly, Freedman and Patten (2004) report an overall positive market reaction surrounding the first President Bush’s unexpected call for changes to the Clean Air Act in June of 1989, but also document that companies with higher levels of airborne toxic releases suffered more negative adjustments, while differences in the reaction were positively related to levels of prior environmental disclosure.

Perhaps, most closely related to our investigation, Blacconiere and Northcut (1997) identify the market reaction to a series of events related to the passage of the Superfund Amendments and Reauthorization Act of 1986. Focusing exclusively on companies with chemical operations, Blacconiere and Northcut find negative cumulative returns, on average, for the events examined, and they further report that when only legislative events are considered, the market reaction is negative and statistically significant. Finally, the authors also document that firms facing greater Superfund exposures suffered more negative losses, while, consistent with Blacconiere and Patten (1994) and Patten and Nance (1998), prior environmental disclosure appeared to mitigate the negative market response.

Although the CTSCA relates to slavery and human trafficking issues in corporate supply chains as opposed to environmental issues, the passage of the law could be expected to increase the regulatory cost exposures for affected companies. And while disclosure compliance costs would likely be relatively low,⁴ if investors believed that the new legislation would substantially increase those costs, a negative market reaction would be anticipated. However, even if market participants do not believe that the cost of disclosure compliance would be high, if they believe that the increased transparency would increase legitimacy threats through social and political exposure for the companies (perhaps in turn leading firms to expand their efforts and activities with respect to addressing supply chain exposures), a negative market adjustment would likewise be anticipated. Based on the evidence of investor reactions to other social cost-inducing events, therefore, we state our first hypothesis as:

H ₁

Investors will react negatively to the legislative events culminating in the CTSCA.

While we anticipate a negative investor response to the passage of the CTSCA, the evidence from prior studies (e.g., Bowen et al. 1983; Blacconiere and Patten 1994) also suggests that the reaction would be expected to vary across different exposure factors. More specifically, the prior evidence suggests that companies facing greater exposures to the potential legitimacy threats experience more negative market impacts. This thus leads to the following hypothesis regarding the investor reaction:

H _1a

Ceteris paribus, market reactions to the legislative events culminating in the CTSCA will be more negative for companies facing higher regulatory cost exposures.

Company Disclosure Adoption

The second aspect of our analysis focuses on how companies subject to the CTSCA responded to the new reporting requirements and what might explain differences in that response. Results of the prior analyses of company response to mandated CSR disclosure, as summarized above, are mixed. Bebbington et al. (2012) report low levels of Spanish company compliance with the environmental disclosure requirements laid out in the PGC, while both Chauvey et al. (2015) and Chelli et al. (2016) find relatively high levels of compliance with France’s NRE mandates. However, it is important to note that the higher disclosure levels reported in the latter studies are limited to analyses of disclosure several years after the requirements took effect. Indeed, both Chauvey et al. (2015) and Chelli et al. (2016) indicate that initial levels of compliance were quite limited (also see Delbard 2008). Because we are examining the initial response to the new CTSCA disclosure requirements, we would therefore also anticipate relatively low levels of compliance. We state this hypothesis as:

H ₂

Company compliance with CTSCA reporting requirements will be limited.

Following Hrasky (2012) and Chelli et al. (2016), we also explore the extent to which the disclosure response to the CTSCA appears to be symbolic or substantive. Given Chelli et al.’s findings that, even when normativity was relatively high, substantive disclosure remained very low, we anticipate that company disclosure response to the CTSCA requirements will be more symbolic than substantive in nature. This hypothesis is formally stated as:

H _2a

Disclosure response to the CTSCA will be more symbolic than substantive.

Finally, we also examine the CTSCA disclosure response in terms of strategic as opposed to institutional legitimation. As noted above, Chelli et al. (2016) argue that their failure to find higher levels of environmental disclosure changes for sample companies in environmentally sensitive industries suggests that concerns with institutional rather than strategic legitimacy may explain disclosure choice. However, we believe that the lack of a significant difference in their analysis may be a function of both a small sample size (only 40 firms—20 from France and 20 from Canada) and, including as environmentally sensitive, industries not normally classified as such. To illustrate, Berthelot et al. (2003, p. 18), in their review of environmental disclosure research, note that the oil and gas, chemicals, forest and paper products, and utilities industries are typically considered as being environmentally sensitive, but Chelli et al. also code firms in the airline, transportation, construction, and electrical components and equipment industries as facing higher environmental sensitivity. Whether the lack of differences in disclosure would hold for an expanded sample, and one that classifies environmental sensitivity in line with prior research is not clear. But, given the breadth of studies documenting that firms facing greater social and political exposures consistently include more extensive CSR disclosure,⁵ we expect the disclosure response to the CTSCA to similarly be related to attempts at strategic legitimation. We state this final hypothesis as:

H _2b

Disclosure response to the CTSCA will be related to concerns with strategic legitimation.

Methods

Sample

Although the CTSCA applies to both manufacturing and retailing firms, the difficulty of identifying the presence in California for the former led us to limit our investigation to retail companies. Store locator information on retailers’ websites allowed us to verify that all sample firms did indeed have outlets in the state and as such, where worldwide sales were sufficiently high, subject to the law. Accordingly, sample firms had to meet the following criteria:

• (1) They had to be publicly traded retail firms with operations in the state of California.

• (2) They had to have worldwide sales (based on 2009 fiscal year sales) of $100 million or more.

• (3) They had to have necessary data available on the Center for Research in Security Prices (CRSP) and Research Insight databases.⁶

In total, 105 companies met our search criteria and constitute our final sample. Firms ranged in size from $176 million to $406,103 million with a mean (median) of $14,541 million ($2631 million). The sample consists of 50 apparel and footwear retailers, 30 specialty retailers, 17 general merchandisers, and 8 food and drug store chains.⁷

Investor Reaction

The first goal of our investigation is to assess the market reaction to the CTSCA law. Similar to prior studies focusing on legislative actions in the environmental domain (Blacconiere and Northcut 1997; Cahan et al. 1997), we identify multiple events in the legal process and calculate market-adjusted abnormal returns surrounding the chosen actions. More specifically, we focus on the date the law was officially passed by the California Senate (August 30, 2010), and the day then Governor Arnold Schwarzenegger signed the legislation into law (September 30, 2010).⁸ Following both Blacconiere and Northcut (1997) and Cahan et al. (1997), we combine the market responses for an overall reaction measure.

We rely on data in the CRSP database to calculate our market-adjusted abnormal returns. For each sample company, we retrieve the cumulative 3-day return centered on each of the two event dates and subtract the corresponding 3-day cumulative market return using the New York Stock Exchange value-weighted index yielding a cumulative abnormal return (CAR) for each sample firm. We then compute the portfolio return as the mean of combined CAR observations.

We use ordinary least squares (OLS) regression analysis to assess whether potential legitimacy threats in the form of exposures to social and political pressures explain differences in the market reaction across firms (based on individual company CARs), and we rely on two proxy variables to capture that exposure. The first of these, firm size, has been used extensively in social and environmental disclosure research as a measure of exposure (see, e.g., Patten 1991, 1992; Hackston and Milne 1996; Cho et al. 2012). Larger companies, presumably owing to greater visibility, are assumed to be subject to greater political scrutiny (Watts and Zimmerman 1986), and we accordingly anticipate more negative market reactions for these firms. We measure firm size in this model as the natural log of each company’s 2009 fiscal year sales.

Our second proxy for social and political exposure relates more specifically to potential differences in concerns with slavery and human trafficking in companies’ supply chains. We argue that, while almost all retail firms face some type of exposure to these issues, the apparel and footwear sector, in particular, has faced intense scrutiny regarding working conditions (see, e.g., Park-Poaps and Rees 2010). Garment production is labor intensive, automation is limited, and the supply chain is complex and multi-layered, making direct management difficult (Park-Poaps and Rees 2010; Sneed 2014). High-profile cases of human rights violations exist within the apparel industry dating back to 1911, when the devastating Triangle Shirtwaist Factory fire led to the death of 146 people.⁹ Further, supply chain concerns within the apparel and footwear industry have received exposure through non-governmental organization reports (e.g., Not for Sale 2012), popular press articles (e.g., Sneed 2014; White 2015), and academic research (e.g., Doorey 2011; Islam and Deegan 2010; Yu 2008), all suggesting greater public scrutiny and social pressure for these retailers. Therefore, we classify apparel and footwear retail firms as high risk, and we expect the market reaction to the legislation to be more negative for these companies.

In addition to our political cost exposure variables, we include a control for prior CSR reporting. Because the requirements of the CTSCA are related to social information disclosure, investors could believe that companies with established CSR reporting systems in place would be likely to incur fewer costs associated with meeting the CTSCA mandate. We assume that companies having already issued a standalone CSR-type report as of the passage of the CTSCA to be perceived as having more developed social reporting systems. Following Dhaliwal et al. (2011), Guidry and Patten (2010), and others, we reviewed Corporate Register, CSR Newswire, and sample company websites to identify whether firms had issued a standalone CSR report as of August, 2010, and we use a one/zero indicator variable to designate the sample firms with prior reporting. We identified 18 companies with a standalone CSR report issued prior to the passage of the CTSCA. Because adoption of the legislation would be expected to be less costly for companies with established social reporting systems, we expect this variable to be positively related to companies’ market reactions.

Based on the above discussion, we state our first model (with expected relations noted parenthetically beneath each variable) as

$${\text{CAR}}_{i} = a_{1} + B_{1} \, \mathop {{\text{firm}}\;{\text{size}}_{i} }\limits_{( - )} + B_{2} \, \mathop {{\text{high}}\;{\text{supply}}\;{\text{chain}}\;{\text{risk}}_{i} }\limits_{( - )} + B_{3} \mathop {{\text{prior}}\;{\text{CSR}}\;{\text{reporting}}_{i} }\limits_{( + )} .$$

Company Disclosure

The second aspect of our investigation centers on company disclosure in response to the CTSCA. The requirements of the legislation took effect January 1, 2012. Accordingly, to assess company compliance with the legislation, we accessed all sample company websites over the first 7 days of January, 2012 and searched for CTSCA disclosure. If companies had no prominent link to CTSCA information on the home page, we did a search using terms including ‘California Transparency in Supply Chains,’ ‘CTSCA,’ ‘supply chains,’ ‘human trafficking,’ and ‘slavery.’ If search results failed to identify the CTSCA information, we followed all active links on the website to assure that disclosure did not exist. In all cases where no CTSCA disclosure was found in the initial searches, we returned to the websites on January 10, 2012 and repeated the search. Our first measure of compliance is a yes/no delineation where companies with any CTSCA disclosure as of the first 10 days of January, 2012 were coded one.

Similar to Hrasky (2012) and Chelli et al. (2016), we next attempted to more carefully classify the disclosure response as symbolic or substantive. Hrasky (2012) investigated carbon footprint disclosures and classified them as substantive if they identified (1) internal corporate initiatives, (2) involvement with external initiatives, or (3) actions taken to reduce carbon footprint. Chelli et al. (2016) similarly coded disclosure segments as substantive if they described initiatives and set out positive environmental impacts. However, by definition, the CTSCA specifically requires firms to provide information on the policies and practices they have in place relative to the five areas of concern, and as such any disclosure made relates to initiatives of the company. As such, the approach of Hrasky and Chelli et al. is not viable in our situation. Instead, we assess the extensiveness of disclosure across each of the five areas of information and argue that where companies provide more extensive information relative to actions being taken, the disclosure is more substantive than symbolic.

We used content analysis to assess the extent of the CTSCA information provided by the sample companies including disclosure. Content analysis has been used broadly in social and environmental accounting research and involves reviewing the disclosure for the presence of specific items of information provision. Similar to Wiseman (1982), Warsame et al. (2002), and others, we used a weighted disclosure scoring where general disclosure of an item was scored one and more extensive disclosure within the topic was given two points. The weighted disclosure scores thus had a range from 0 to 10. “Appendix” Section provides examples of general and more extensive disclosure items across each of the five CTSCA categories. To aid in coding, all CTSCA disclosures were printed to hard copy. Further, given the inherently subjective nature of assessing disclosures as general or more extensive, all items were reviewed independently by at least two members of the research team. All differences in coding across reviewers were discussed and reconciled.

To explore whether legitimacy concerns relate to differences in CTSCA disclosure choice, we estimate two forms of the following multiple regression model (with expected relations noted parenthetically beneath each variable):

$${\text{Disclosure}}_{i} = a_{1} + B_{1} \, \mathop {{\text{firm}}\;{\text{size}}_{i} }\limits_{( + )} + B_{2} \, \mathop {{\text{high}}\;{\text{supply}}\;{\text{chain}}\;{\text{risk}}_{i} }\limits_{( + )} + B_{3} \mathop {{\text{prior}}\;{\text{CSR}}\;{\text{reporting}}_{i} }\limits_{( + )} .$$

In our first estimation, Disclosure _i is a one/zero indicator variable where one designates that firm i included CTSCA disclosure on its web page as of January 10, 2012. For the second estimation, we use firm i’s weighted CTSCA disclosure score as the dependent variable. We estimate the first disclosure model using logistic regression analysis and the second using OLS regression. As with the market response analysis, we consider firm size and high supply chain risk as factors associated with greater legitimacy threats, and we again control for prior CSR reporting experience. Because disclosure occurred in January of 2012, our firm size measure in these analyses is calculated as the natural log of each company’s fiscal year 2011 sales. Similarly, our prior reporting metric in the disclosure analyses is based on having had a standalone CSR report issued as of the end of 2011 (23 firms). The High Supply Chain Risk variable again identifies sample companies classified as apparel and footwear retailers. We expect all three explanatory variables to be positively related to difference in disclosure.

Results

Investor Reaction

In the first stage of our analysis, we focus on the investor response to the passage of the CTSCA, and panel A of Table 1 identifies the average market reaction for our sample of retail firms across the final legislative events related to the act. As summarized in the table, mean market-adjusted returns were negative for both legislative events and the combined reaction amounts to −2.29 % which is statistically significant at p < .01, one-tailed. This finding is consistent with the prior studies examining market reactions to events potentially increasing social and political costs for affected companies. We also find evidence that the reaction is more negative for companies presumed to face greater legitimacy threats in the form of social and political exposures. As reported in panel B of Table 1, our regression analysis indicates that both the firm size and higher supply chain risk variables are negatively signed and statistically significant, although the former at only the .088 level, one-tailed. In contrast, the supply chain risk variable is significant at <.01, one-tailed. Although, as expected, our control for prior CSR reporting is positively related to differences in mean abnormal returns, it is not statistically significant at conventional levels.¹⁰ Overall, the model explains only a modest 3.6 % of the variation in market response across sample firms. In order to more fully explore the impact that supply chain risk appears to have played in the investor reactions to the CTSCA legislation, we present in panel C of Table 1 the results of tests for differences in the market reaction across the higher supply chain risk firms and other retailers in the sample. As highlighted in the panel, the average abnormal return for the former is −3.22 % in contrast to an average negative return of only 1.44 % for the other retail companies, and this difference is statistically significant at .017, one-tailed. Overall, the results of our investigation of investor perceptions support Hypotheses 1 and 1a, although supply chain risk appears to play the largest role in explaining differences in reaction.

Table 1 Market response to the passage and signing into law of the CTSCA

Company Disclosure

We next explore the extent to which our sample of retail companies adopted the requirements of the CTSCA and the degree to which the disclosure response appears to be symbolic or substantive. In contrast to the expectation laid out in H₂, it appears that compliance with the legislation was fairly high. As noted in panel A of Table 2, 87 of the 105 firms in our sample (82.9 %) provided a CTSCA disclosure on their website as of the first 10 days of January, 2012. Further, analysis of the content of the disclosures across CTSCA categories, summarized in panel B of the table, shows that 78 of the 87 disclosing companies provided at least some information relative to each of the five disclosure categories, although only one category—‘evaluate and address risks in the supply chain’—was included by all disclosing firms. Our compliance level is consistent with the level of French firms taking up NER and Grennelle II environmental disclosure requirements as reported by Chelli et al. (2016) and thus provides additional evidence that ‘soft law’ initiatives, at least in some cases, can induce a certain level of normativity.

Table 2 Adoption of CTSCA disclosure by retail firms

Although general compliance with the CTSCA was relatively high, analysis of the extensiveness of the disclosure suggests that, overall, the responses tend to be more symbolic than substantive. As summarized in panel B of Table 2, extensive disclosure across the five CTSCA categories ranged from only eight firms doing so at the low end (for disclosures on ‘supply chain audits’ and ‘provide employee and manager training’) to 30 at the high end (for disclosure on ‘direct supplier certifies compliance’). Only four sample firms included extensive disclosure across all five CTSCA classifications. Perhaps more troubling, 50 of the 87 disclosing firms had no topics including extensive disclosure, and only 13 companies included extensive information provision for more than one of the CTSCA areas. Further supporting the lack of substantive disclosure, the mean weighted disclosure score, as summarized in panel C of Table 2, was 4.70 across the total sample, but rose to only 5.67 when averaged across disclosures only. Finally, and also suggesting a more symbolic disclosure response, almost half of the disclosing companies (43 of the 87) did not follow the letter of the law and include a link to the CTSCA information on their website’s home page (see panel A of Table 2). Overall, the results provide support for Hypothesis 2a.

Table 3 presents the results of our analyses of the choice to comply with CTSCA (panel A) and differences in the extent of disclosure included (panel B). Relative to the former, we find that, while companies deemed as exposed to higher supply chain risk are more likely to have complied with the CTSCA (High Supply Chain Risk is statistically significant at .002, one-tailed), neither the firm size nor the prior standalone CSR reporting variables are statistically significant at conventional levels. In comparison, as highlighted in panel B of the table, both High Supply Chain Risk and Prior CSR Reporting are positively and significantly(at p < .002 or better, one-tailed) associated with weighted disclosure scores, although firm size, while also positively signed, remains statistically insignificant. Consistent with the results of prior studies of other social and environmental disclosure (e.g., Hackston and Milne 1996; Chauvey et al. 2015; Cho et al. 2015), our analyses of company reporting adoption suggest that firm-specific attributes do explain differences in the reporting, and, in our case, higher supply chain risk appears to be the largest factor explaining differences in CTSCA disclosure decisions. These results provide at least a limited degree of support for Hypothesis 2b.

Table 3 Examination of factors relating to differences in CTSCA disclosure

Discussion

The 2010 passage of the CTSCA offers a rare opportunity to examine investor and firm response to mandatory CSR disclosure outside of the environmental domain. Focusing on a sample of 105 retail companies subject to the law, we find a negative market reaction, on average, to the enactment of the legislation. We further show that the market response was significantly more negative for apparel and footwear retailers, a finding we attribute to the higher supply chain exposures of these firms relative to other retailers. The market reactions were also negatively related to firm size, indicating that larger companies suffered more negative reactions to the legislative events associated with the passage of the CTSCA. Thus, while prior studies (e.g., Bowen et al. 1983; Blacconiere and Patten 1994; Freedman and Patten 2004) document similar investor responses to potential increases in social and political exposure arising from environmental-related events, our findings suggest that the market appears to be equally concerned with the negative consequences arising from legitimacy threats in the social domain.

Results of our investigation of the retail company disclosure response to the CTSCA indicate that, in contrast to evidence of initial response to mandated CSR disclosure in other settings (Larrinaga et al. 2002; Delbard 2008; Bebbington et al. 2012; Chauvey et al. 2015; Chelli et al. 2016), our sample of retail firms appeared to comply with the new law at relatively high rates. This unexpected result could potentially be a function of differences in the U.S. setting, relatively low costs of implementing disclosure, the legislation’s focus on a specific aspect of social concern, or some combination of these, or other factors. We leave exploration of this to future research. However, our analysis also indicates that the companies facing greater supply chain risks were more likely to include CTSCA disclosure than other retail firms subject to the law, suggesting that strategic legitimacy concerns may have influenced disclosure choice. This is further supported by our results with respect to differences in the extent of information provided. Overall, and consistent with Chelli et al. (2016), we find that disclosure seemed to be more of a symbolic response in that more extensive disclosure across the CTSCA items was quite limited. However, the finding that differences in the weighted disclosure scores were positively associated with classification as a higher supply chain risk firm again suggests that attempts at strategic legitimacy were at play in the response.

Aside from the evidence with respect to disclosure and firm legitimacy, it is important to highlight that the quality of the information presented in the CTSCA disclosures was, on average, quite limited, and we believe that the results of our initial market reaction tests may help to explain this. While consumers, NGOs, and other stakeholder groups clearly seem to want richer information on corporations’ efforts to ensure more ethical performance within their supply chains (see, e.g., Bhaduri and Ha-Brookshire 2011; Doorey 2011; Park-Poaps and Rees 2010), investors appear to interpret increased disclosure as potentially costly in terms of firm value. Accordingly, managers may be reluctant to be more transparent with respect to their supply chain activities. This tension between stakeholder desires and investor concerns, if it does limit information provision, would seem ultimately to be potentially harmful. Although couched only in terms of informing consumers, it seems likely that the CTSCA is meant to bring additional pressures on firms to enhance their efforts at reducing slavery and human trafficking in their supply chains. Doorey (2011, p. 587) notes that “transparency has long been used as a means to influence corporate behavior” and he cites Loss’s (1988, p. 33) quote that “people who are forced to undress in public will presumably pay some attention to their figures.” But if firms can adopt the CTSCA requirements without providing real transparency, as seems largely to be the case, it appears unlikely that the disclosure will induce better corporate efforts at safeguarding their supply chain activities.

The lack of meaningful disclosure in the social and environmental domain is not a new phenomenon. Indeed, one of the major criticisms of CSR reporting is that, due to its largely voluntary nature, the information provided is not comparable across firms (see, e.g., Dingwerth and Eichinger 2010) and does not allow for stakeholder assessment of actual company performance (see, e.g., Aras and Crowther 2009; Gray 2010; Moneva et al. 2006). As such, mandated disclosure such as that required under the CTSCA ought to, in theory, help alleviate that problem. Unfortunately, our results suggest that the company response, on average, was largely symbolic as opposed to substantive which would seem to suggest that meaningful stakeholder assessment of corporate actions, even in this mandated case, will likely remain elusive. Without additional guidance and rules for reporting, factors Bebbington et al. (2012) argue are necessary for inducing higher levels of normativity with respect to CSR reporting, we fear that the primary goal of the CTSCA—allowing consumers to make choices that are better informed with respect to companies’ supply chain efforts—will not be met, and in turn incentives for improved corporate performance will likewise remain reduced.

Conclusion

In this study, we explored the market and company responses to the CTSCA, one of the few pieces of legislation mandating CSR disclosure relative to a specific social issue outside of the environmental domain. We found that, on average, publicly traded retail firms subject to the law experienced significantly negative market reactions to the events culminating in the new legislation. We also document that factors associated with greater potential legitimacy threats—firm size and higher supply chain risk—were associated with more negative reactions, suggesting that investors negatively value such exposure in CSR areas outside of the environmental domain. We also found that, while compliance with the new legislation was relatively high, the disclosure response tended to be more symbolic than substantive. Further, because firms facing higher supply chain risks were both more likely to comply with the law and to include more substantive levels of disclosure, our results suggest that concerns with strategic legitimacy influenced the disclosure response.

Like all studies, ours is not without limitations. Our sample is limited to publicly traded retail companies, and these firms, particularly because of their exposure to consumer markets, undoubtedly differ in major ways from other types of companies. Accordingly, we are unable to infer how investors and other firms would respond to other mandates for CSR disclosure, should they arise. However, because the CTSCA also applies to large manufacturing firms with operations in California, extending our analyses to this sample, if feasible, could help shed at least some light on this issue. We also examine only the initial company response to the CTSCA. Whether the requirements to disclose companies’ efforts to eradicate slavery and human trafficking in their supply chains lead those firms to improve their performance in this area, and ultimately perhaps as well, their disclosure of those efforts, remains an unanswered question. In addition, our analysis of disclosure response examines primarily differences with respect to what we consider to be differing exposures to social exposure at a very general level. A more nuanced analysis of the narratives within the disclosures, and how those might vary with respect to what institutional theorists refer to as coercive, normative, and mimetic pressures (see, e.g., Higgins and Larrinaga 2014) could be enlightening. Similarly, richer assessment of the narratives might also be useful in uncovering differences in companies’ use of disclosures as attempts at establishing (or repairing) pragmatic as opposed to moral legitimacy (Suchman 1995). However, such qualitative analyses are beyond the scope of our examination and we leave these possibilities for future research.

Appendices

Appendix: Disclosure Content Analysis Scheme

The California Supply Chain Transparency Act of 2010 (CTSCA) requires companies to publicly disclose on their website the policies they have in place to ensure that their supply chains are free of slavery and trafficking. The law specifically mentions disclosure related to five major areas (identified below). We calculated two separate disclosure metrics, the first of which is un-weighted and involved awarding one point for each of the specific areas required under the CTSCA. For the second metric, we classified disclosures, where present, as either general (one point) or more extensive (two points). Below we identify the specific CTSCA disclosure requirements and, for each, provide examples of general and more extensive disclosure as based on our review.

Appendix 1: Evaluate and Address Risks in Supply Chains

• 1. Destination Maternity Company—DMC carefully consider selection of its vendors. In particular, DMC is risk averse to doing business with vendors in countries that do not have what we consider to be adequate human rights protections. Either an employee of DMC, or a third party directed by DMC, conducts periodic onsite audits on selected vendors to ensure material compliance with our Global Labor Practices, included to evaluate risk of human trafficking and slavery. Content Score of 1.

• 2. CVS Caremark—Respect for human rights is expressed in CVS Caremark’s Supplier Ethics Policy, which all vendors around the world must adhere to as a condition of doing business with the company. The policy conforms with the conventions of the International Labour Organization (ILO) and prohibits human trafficking and the use of child, forced or imprisoned labor, requires that working conditions are safe and fair; forbids any form of discrimination with regard to age, gender, minority status, and/or other protected classes; and upholds the right to freedom of organization. We monitor compliance with the Suppliers Ethics Policy through risk-based audits conducted by external third parties. Content Score of 2.

Appendix 2: Direct Supplier Certifies Compliance

• 1. Maidenform, Inc.—Prior to accepting any orders for Maidenform, Inc. product, our suppliers are required to sign our Sourcing Agreement and agree to be bound by our Code of Vendor Conduct, Maidenform, Inc.’s Code, which states that:

  Maidenform expects all Vendors to operate within full compliance of all applicable laws and regulations of the countries in which they operate… Content Score of 1.

  • 2. American Eagle Outfitters, Inc.—AEO, Inc.’s Vendor Code of Conduct (hyperlink to Code) is based on universally accepted human rights principles and sets forth our minimum standards and expectations for suppliers. Our Code expressly prohibits the use of child labor and forced or involuntary labor. These prohibitions include, but are not limited to, trafficked, prison, bonded, and indentured labor, as well as forced overtime.

    All suppliers must agree contractually and in writing to abide by the terms of our Vendor Code of Conduct and other applicable laws and regulations before we do business with them. As part of this agreement, AEO, Inc. suppliers also warrant that any subcontractors they may independently contract with to produce AEO, Inc. product will comply with the terms of our Code and other applicable laws and regulations. For more details on our Code of Conduct as well as associated guiding principles and governance, please see the Corporate Governance Section_(hyperlink) of AE Better World_(hyperlink). Content Score of 2.

Appendix 3: Supply Chain Audits

• 1. Jos. A. Banks Clothier, Inc.—The Company conducts, or directs that there shall be conducted, audits of most of its suppliers to evaluate compliance with Company standards regarding trafficking and slavery in supply chains. Most of the audits are performed by independent third parties; some are performed by Company Associates. Content Score of 1.

• 2. Talbots, Inc.—Auditing: Our factory monitoring partners audit factory compliance with The Talbots, Inc. Merchandise Supply Chain Code of Conduct (hyperlink), which prohibits human trafficking and forced labor. In fiscal 2010, 33 % of our apparel factory base was audited by an independent, third-party auditing firm. The remaining 67 % were audited by Li & Fung’s vendor compliance team. Approximately 4 % of all active apparel factories in fiscal 2010 received unannounced audits. In fiscal 2011, our goal is to increase the percentage of apparel factories that are audited by an independent, third-party auditing firm. We also plan to increase the percentage of factories that receive unannounced audits. Content Score of 2.

Appendix 4: Maintain Accountability Standards

• 1. Home Depot, Inc.—Supplier Certification: The Home Depot has a Supplier Buying Agreement in place with all direct supplies requiring them to comply with international standards and applicable laws and regulations, including those related to forced labor and child labor as specified in the Home Depot Social and Environmental Responsibility Standards. Content Score of 1.

• 2. Gap, Inc.—Prior to accepting any order for Gap, Inc. branded products, our suppliers are required to sign our Vendor Compliance Agreement and agree to be bound by our Code of Vendor Conduct (COVC). Gap Inc.’s COVC states that:

  Factories that produce goods for Gap, Inc. shall operate in full compliance with the laws of that respective countries and will all other applicable laws, rules, and regulations including those relating to labor, worker health and safety, and the environment.

  In signing Gap Inc.’s Vendor Compliance Agreement which also incorporates the COVC, Gap Inc. suppliers agree to comply with the following:

  All applicable laws, rule and regulations.. these laws include, but are not limited to, laws relating to the employment, conditions, of their respective employee such as (1) wage and hour, labor, child labor, and forced labor requirements, (2) health and safety, (3) immigration, (4) discrimination, (5) labor or workers’ rights in general and (6) environmental laws and regulations. Content Score of 2.

Appendix 5: Provide Employee and Manager Training

• 1. Men’s Wearhouse, Inc.—For training, managers with direct responsibility for supply chain management of our direct sourced products have attended training by our third-party consultant, Underwriters Laboratories, Inc. on human trafficking and slavery, particularly with respect to mitigated risks within the supply chains of products. Additionally, those managers attend regular calls with the third-party consultant to help ensure the safety, quality, and socially responsible manufacture of the Company’s direct sources of products. Content Score of 1.

• 2. Office Depot—Training: Office Depot continuously develops and enhances our training programs for our associates. We provide regional training to our associates and our associates are required to acknowledge and adhere to our Code of Ethical Behavior, which includes compliance with all applicable laws where Office Depot conducts business. Additionally, we are in the process of enhancing our associate training for our associates who are directly responsible for our supply chain management on mitigating risks of slavery and human trafficking and anticipate such training to commence in early 2012.

  Education and Training Awareness (Supplier Security Guidelines): A Security awareness program should be provided to employees including recognizing internal conspiracies, maintaining product integrity, and determining and addressing unauthorized access. These programs should encourage active employee participation in security controls. Content Score of 2.