28-06-2019
University of Amsterdam
Communication Science: Corporate Communication Master Thesis
Signals of Quality
Reliability of Voluntary vs. Mandatory CSR Reporting
in the European Fashion Industry
Theresa Roessler 12021857 theresa.rosler@student.uva.nl
Abstract
With European Directive 2013/34/EU in place, researchers are presented with the unexampled opportunity to compare voluntary and mandatory CSR reporting across European countries and industries. Expectations were high with regards to increasing
comparability, consistency, credibility, and overall quality of reports. This study investigates if some of them have been fulfilled within the European fashion industry. Only fairly recently has the industry experienced public scrutiny about its social and environmental impacts due to labor- and chemical-intensive production in emerging countries. Fashion companies’ CSR reports provide an interesting case as they serve as valuable communication tools in a
discussion about human rights and greenwashing. They function to reduce stakeholder skepticism about sustainability initiatives and to increase stakeholders’ level of information where direct observation is difficult if not impossible. Taking a signaling perspective, this study investigates the extent to which mandatory CSR reports published by European fashion companies are more reliable signals of a commitment to CSR than their voluntary reports, and further tests whether results are persistent when compared to a control group of voluntary reporters and when taking the firm’s quality into account. Results do not give clear indication of the Directive leading to more reliable CSR reports within the European fashion industry. Findings do show a significant effect of a firm’s quality: Across both years, firms with higher CSR ratings published more reliable reports than their lower-quality competitors. Because low-quality fashion companies are not able to achieve the same level of reporting reliability, a CSR report can serve as a useful signal for firms to convey their sustainability commitment to the public, and for stakeholders to distinguish between high- and low-quality companies.
Keywords: CSR reporting, reliability, signaling theory, European fashion industry,
Reliability of Voluntary vs. Mandatory CSR Reporting
in the European Fashion Industry
While firms in hazardous and controversial industries, such as the oil and gas, chemical or petroleum industry, have been condemned guilty for their harmful business conduct, long ago (Shabana, Buchholtz, & Carroll, 2016), the fashion industry has been experiencing public scrutiny only fairly recently. Fashion today involves critical social and environmental issues through labor- and chemical-intensive sourcing and production. Due to globalization, almost all larger textile companies have relocated their production to emerging and developing nations, such as India and Bangladesh. Often, a t-shirt’s path from raw material to finished product passes through various manufacturers, subcontractors and suppliers in several countries - each with their own laws and labor regulations concerning social standards, most of the time lower than those in industrialized nations (Garcia-Torres, Rey-Garcia, & Albareda-Vivo, 2017; GRI, 2008; Mammut Sports Group AG, 2016). With the fatal collapsing of Rana Plaza in Bangladesh in 2013, which killed 1,134 workers and left thousands more injured, topics like modern slavery, unethical working conditions in so-called sweatshops and human rights issues within the fashion industry were put on the political and public agenda (Mammut Sports Group AG, 2016). As a consequence, both large and small fashion companies are under constant surveillance and pressure to do better. Nevertheless, ‘good’ CSR activities are often met with severe stakeholder skepticism and buzzwords like greenwashing are common (Du, Bhattacharya, & Sen, 2010). The need for comprehensive information increases, as the fashion industry’s practices have considerable economic, social, and environmental impacts along the supply chain and moral conduct is to some extent questioned (Hahn & Kühnen, 2013; Lock & Schulz-Knappe, 2018; Pedersen, Neergaard, Pedersen, & Gwozdz, 2013; White, Nielsen, & Valentini, 2017). In an attempt to increase stakeholders’ level of information, a company’s report on corporate social responsibility
(CSR) can be a valuable communication tool to respond to such legitimacy tests (Patriotta, Gond, & Schultz, 2011; White et al., 2017). Signaling theory proposes that CSR reports are especially useful in reducing information asymmetry, where direct observation of a
commitment to sustainability is difficult for stakeholders (Connelly, Certo, Ireland, & Reutzel, 2011).
In the past, voluntarily issued CSR reports have been criticized to only “cover few stakeholders, cherry pick elements of news and generally ignore the major social issues that arise from corporate activity such as lobbying, advertising, increased consumption,
distributions of wealth and so on” (Milne & Gray, 2013, p. 17). Several studies have shown that standardization and disclosure regulations positively affect the quality, comparability and credibility of CSR reports (Ioannou & Serafeim, 2017; Lock & Seele, 2016; Tschopp & Huefner, 2015).
In order to ensure transparency and enhance the consistency, comparability and quality of non-financial reports across Europe, the European Parliament acknowledged the importance of businesses disclosing information on social and environmental issues along certain, transnational guidelines (EC, 2014). With the financial year of 2017, all large, listed European companies are, for the first time, obliged to report on their social and
environmental impacts within the new mandatory EU reporting regime. The recent shift from voluntary to mandatory reporting in Europe provides researchers with the unexampled
opportunity to compare the reports’ content and quality before and after, and across countries. This is especially interesting as the EU Directive provides the first European-wide law on sustainability reporting. While France, Spain, Denmark and Norway had already mandated annual reporting on environmental and social impacts, these national laws differ greatly in their scope, which makes international comparison difficult (Pedersen et al., 2013). In 2016, Lock and Seele concluded, that a “comprehensive regulation at a transnational level as
proposed by the EU could be a useful move to level the playing field in the market and raise the credibility of reporting” (194). If the EU’s regulations now provide such a framework for corporate CSR reporting across countries, to what extent are the mandatory reports published by European fashion companies more reliable signals of CSR commitment than their
previous, voluntary reports and to what extent are results persistent when compared to a control group of voluntary reporters and when taking the quality of the firm into account?
Results indicate that the EU Directive did not lead to more reliable mandatory CSR reports within the European fashion industry, but that across years, although generally scoring mediocrely, firms with higher CSR ratings published more reliable reports than their lower-quality competitors. A CSR report can therefore serve as a signal both for firms to convey a commitment to sustainability and for stakeholders to distinguish between high- and low-quality firms, because low-quality fashion companies are not able to achieve the same level of reporting reliability as higher-quality firms.
In the following, a short literature review will be given on the current state of research with regards to voluntary and mandatory CSR reporting, followed by the theoretical
framework of signaling theory and the development of the hypotheses. A methods section provides information on the quantitative content analysis that was conducted, the sample that was drawn and the operationalization of concepts central to signaling theory and this study. Afterwards the empirical results are reported and discussed in light of their implications as well as their limitations.
Theoretical Framework & Hypotheses
Voluntary and Mandatory CSR ReportingDisclosing information on CSR engagement has the potential to increase
often helps to form favorable impressions of the firm. Positive effects of comprehensive CSR reporting include image and reputation enhancement, meeting social expectations and
regulatory requirements, and building legitimacy and trusting relationships with stakeholders (Crane & Glozer, 2016; Hetze, 2016).
While increasingly recognized for its value to the development of CSR reporting, voluntary CSR disclosure allows reporting companies to choose their own reporting period and format, performance indicators and data analysis methods, making it difficult for
stakeholders to compare reports with each other (Habek, 2018). Tschopp and Huefner (2015) have argued that giving companies the freedom to report on their own indicators will increase the likelihood of them misrepresenting their actual environmental and social performance. Consequently, the authors emphasize the importance of standardization to achieve
comparability and consistency across reports.
Several European countries, namely France, Spain, Denmark and Norway, have already mandated annual reporting on environmental and social impacts for large
organizations and companies (Pedersen et al., 2013). These national laws differ greatly in their scope, which makes transnational and industry-wide comparison difficult. The EU Directive therefore presents a valuable, transnational tool to enhance the consistency,
comparability and quality across European countries. In general, mandatory CSR reporting is associated with an increase in the number of reporting companies and their reporting
quantity. Studies that have examined the before and after of regulated CSR reporting have generally confirmed an increase in reporting (Criado-Jiménez, Fernández-Chulián, Husillos-Carqués, & Larrinaga-González, 2008; Llena, Moneva, & Hernandez, 2007; Pedersen et al., 2013). Accordingly, Gulenko (2018) concludes that mandatory CSR reporting increases disclosure completeness, but not the relevance or credibility of reported information.
off the necessities, and consequently provide little useful information (Habek, 2018; Hoffmann, Dietsche & Hobelsberger, 2018). Generally, standardization seems to be
associated with higher reporting quality and there is a tendency towards increased quality for mandatory sustainability reports.
Chauvey and colleagues focus on quality measures based on the accounting principles of relevance, comparability, verifiability, clarity and neutrality. Their results show that French CSR reports’ informational quality increased slightly for mandatory reports, while still generally low. According to the authors, the increase in quality of reporting is due to more coverage of various topics, while the disclosure quality per area has remained largely the same (Chauvey, Giorano-Spring, Cho, & Patten, 2015). Habek and Wolniak’s (2016) analysis of reporting quality along the lines of relevance and credibility is characterized by similar findings: The authors conclude that, although reports generally score low on credibility, mandatory CSR reports are significantly more credible than voluntary reports. However, these findings were not consistent across countries, largely due to the varying mandated scope of reporting per country. Within Habermas’ communicative action theory Lock and Seele (2016) analyzed the credibility of reports with regards to how content-related and contextual factors, like national regulations on CSR reporting, influence reporting
credibility. Their findings show no significant differences between voluntary and mandatory reports, but credibility was significantly higher for standardized CSR reports. In a content analysis of reports subject to disclosure regulations, Ioannou and Serafeim (2017) too, found, that disclosure regulations increased organizations’ level of disclosure and efforts towards comparability and credibility.
With the EU Directive in place, certain guidelines and standards applied across reports and countries will make it easier for interested stakeholders to compare reports among each other. Untruthful information and qualitative differences will be identified more easily as
reports become more standardized. Dishonest reporting is costly and could result in negative stakeholder reactions and possible sanctions, increasing the incentive for companies to report honestly and comprehensively. It is therefore expected that firms will put greater effort into distinguishing themselves from competitors’ sustainability reports in terms of effective, reliable reporting.
Hypothesis 1a: Mandatory CSR reports of European fashion companies will be more
reliable than their previous, voluntary CSR reports.
While previous studies have typically only compared companies’ reports before and after mandatory disclosure regulations, this study additionally includes a control group of voluntary reporters that are not subject to the Directive but issue sustainability reports, nevertheless. This allows a comparison not only in absolute terms, i.e. between companies’ voluntary and mandatory reports before and after the Directive, but also in relative terms, that is between mandatory and voluntary reporters’ CSR disclosure. Including the control group of European fashion companies that voluntarily disclose sustainability information although not covered by the Directive allows a research design similar to that of an experiment with both treatment and control group.
Hypothesis 1b: Mandatory CSR reports published by European fashion companies
that are subject to the European Directive will be more reliable than CSR reports by European fashion companies that are not obliged to report under the Directive.
The following paragraphs build the theoretical framework around the additional hypotheses specific to signaling theory. As will be discussed, previous research has not
incorporated the theory’s central distinction between high- and low-quality signalers, although it offers a valuable, new perspective on CSR reporting practices.
Information Asymmetry & CSR Communication
Signaling theory is fundamentally concerned with the reduction of information asymmetry between two parties where a certain quality cannot be directly observed by outsiders. Michael Spence (1973) originated signaling theory in management literature when explaining the signaling function of education on the job market: Potential employers cannot easily observe the quality of applicants, which is why job candidates obtain education to signal their productive capabilities. Signals can only be effective in distinguishing applicants when the signaling costs are negatively correlated with productive capability, that is when less productive applicants cannot imitate the signal, i.e. obtain the same educational level (Connelly et al., 2011; Spence, 1973).
The theory’s primary elements are the two actors - the signaler and the receiver - and the signal itself. The signaler, an individual, or a firm, has an underlying quality that he wishes to signal to the receiver who then observes and interprets the signal (Connelly et al., 2011). As a signal, companies may choose any action that provides some indication of the company’s intentions. Often, signaling occurs in competitive environments where various companies try to influence important stakeholders by signaling information that is otherwise not easily observable (Dögl & Holtbrügge, 2014).
From a functional perspective, strategic CSR reporting serves to signal a commitment to sustainable business conduct to stakeholders, where direct observation is difficult and information asymmetry arises. Companies as signalers may use their CSR report as a response to stakeholder (signal receivers’) expectations, to close potential legitimacy gaps due to lack of information, and to distinguish themselves from their competitors (Bae, Masud, & Kim, 2018; Connelly et al., 2011; Simaens & Koster, 2013). CSR disclosure may
increase stakeholder knowledge about the company’s sustainability strategy and
achievements and decrease existing information asymmetry about abstract or intangible characteristics of the company (Michelon, Pilonato, & Ricceri, 2015). Furthermore, signaling social responsiveness to stakeholders may provide a competitive advantage as stakeholders may evaluate the company more favorably (Bae et al., 2018; Hetze, 2016).
Although widely applied to financial reporting in management literature and essentially concerned with the communication between a signaler and a signal receiver, signaling theory is not prevalent within communication science. Hahn and Kühnen (2013) have noted the potential for researchers to link signaling theory to CSR reporting. To date this has only been addressed by few scholars. Simaens and Koster (2013) showed that directors provide voluntary assurance in their sustainability reports to signal the report’s credibility to stakeholders, while Bae and colleagues (2018) investigated the correlation of corporate governance signals and the completeness of sustainability disclosure. Mahoney and colleagues analyzed whether firms use stand-alone CSR reports to signal their actual
commitment to sustainability (signaling function) or whether these reports are used primarily to convey an image of being a good corporate citizen while not actually committing to CSR (greenwashing). Results indicate that stand-alone CSR reports are used to signal a firm’s CSR commitment to stakeholders. The control group for this study consisted of firms that did not issue a report at all (Mahoney, Thorne, Cecil, & LaGore, 2013). With regards to different CSR initiatives, Dögl and Holtbrügge (2014) showed varying signaling effects on a
company’s perception by potential and actual employees: Less visible signals are perceived less meaningful, and thus have little effect on corporate environmental reputation. Hence the importance of signal observability. Hetze (2016), too, proposed signaling effects on corporate CSR reputation by linking signaling theory, CSR reporting, and stakeholder perceptions in a conceptual framework. She concludes that a CSR report as a communication signal does not
directly increase CSR reputation, but is mediated by several factors such as the signaling context and evaluations by stakeholders.
In sum these studies have focused on single signals, such as the mere existence of a report, its assurance or a firm’s board and shareholding structures, or focused on stakeholder perceptions of CSR reputation, rather than the content and quality of CSR reports themselves. Although issuing a CSR report can be considered a signal itself (Mahoney et al., 2013), the existence of a report alone says little about the company’s quality (Simaens & Koster, 2013). A content analysis by Ching and Gerab (2017) stands out from previous research as the authors not only combine the theoretical key aspects of legitimacy, stakeholder and signaling theory but also assess reporting quality according to a scoring system. The authors find that reporting quality in Brazil has consistently increased since 2008 and conclude “that the improvement in sustainability reporting quality acts as an important signal to gain legitimacy when information asymmetry happens during the legitimacy process” (p. 107).
Addressing reported content quality, and further incorporating signaling theory and its focal distinction between high- and low-quality firms, this study assesses the reliability of reported information as a signal of a firm’s underlying quality and commitment to
sustainability. Signal reliability consists of the four signal properties signal observability, cost, honesty, and fit, which will be explained in more detail hereinafter.
Signal Observability & Signal Cost
Signal observability and signal cost are two important characteristics of effective signals. Signal observability refers to the extent to which others, outside of the organization, can notice the signal. Dögl and Holtbrügge (2014), for example, have shown that the
visibility and perception by stakeholders: More visible signals were perceived as more meaningful and consequently influenced environmental reputation the most.
The costs associated with obtaining the signal are signal costs. This implies that some signalers can afford obtaining certain signals better than others, which makes cheating and false signaling difficult and unprofitable. To maintain effectiveness, the costs associated with a signal must be high enough so that it is unprofitable for dishonest signalers to imitate the signal (Cohen & Dean, 2005; Connelly et al., 2011).
Signal Honesty & Signal Fit
Signal fit and honesty are especially essential to signal reliability (Connelly et al., 2011). Signal honesty includes both the integrity of the signaler, that is the extent to which the signaler is honest about having the underlying quality associated with the signal, and the genuineness of the information, that is the truth of the information provided (Cohen & Dean, 2005; Connelly et al., 2011). Lock and Seele (2016) defined the former as sincerity, referring to the subjective truth of what the firm says in its CSR report. The latter part of honesty refers to the accuracy and correctness of the statements made.
Signal fit is defined as the extent to which the signal is correlated with the
unobservable quality of the signaler (Connelly et al., 2011). The signal needs to match the quality that a firm wants to convey to outsiders. Information provided in fashion companies’ CSR reports should therefore appropriately indicate industry-specific commitment to
sustainability.
Effective, reliable signals therefore contain observable, honest and accurate information, that is relevant and fit to signal a commitment to sustainability in the fashion industry and is costly and therefore unprofitable for lower-quality firms to imitate.
High- and Low-Quality Signalers
Signaling theory offers a new perspective on reliability outcomes, as it distinguishes between and low-quality firms signaling their true quality to stakeholders. For high-quality firms, sustainability reporting results in a better payoff than not reporting, because their true quality is high and informing stakeholders about this quality benefits corporate reputation. On the other hand, low-quality firms typically receive a lower payoff when they signal compared to when they do not, because their true quality is low and stakeholder perceptions might change for the worse if they learn about this quality. In a so-called
separating equilibrium, signaling is only profitable for high-quality firms (payoff A) and low-quality firms should not be motivated to signal (payoff C), which enables stakeholders to effectively judge each firm’s true quality. Conversely, when signaling is profitable both for high- (payoff B) and low-quality firms (payoff D), a pooling equilibrium results (Connelly et al., 2011). Table 1 provides an overview of the signaling outcomes (payoff) for both high- and low-quality firms in a separating and in a pooling equilibrium.
Table 1. Payoff for high- vs. low-quality firms in separating and pooling equilibrium.
Treatment Group Control Group
High-quality firm Low-quality firm High-quality firm Low-quality firm 2017 (t1) Pooling Equilibrium B1 D1 B2 D2 2018 (t2) Separating Equilibrium A1 C1 A2 C2
Voluntary CSR reports have been criticized for their incompleteness, omissions and so-called half-truths (Devin, 2016; Milne & Gray, 2013). In a voluntary setting, companies of low-quality may choose to emphasize positive information on certain sustainability aspects while downplaying or ignoring others. High-quality firms may signal primarily to distinguish themselves from their lower-quality competitors. In line with CSR literature, CSR
communication is regarded as critical and necessary as stakeholders become more vocal and demand more transparent information (Crane & Glozer, 2016). Compared to not issuing a report at all, voluntary CSR reporting in this case, seems profitable both for high- and low-quality companies because the estimated payoff outweighs the costs associated with
reporting. As a consequence, stakeholders have difficulties to effectively judge a firm’s true quality by its CSR reporting activities in a pooling equilibrium (Mahoney et al., 2013).
As the EU has mandated annual CSR reporting for all publicly listed large European companies according to its Directive, a separating equilibrium is expected to form. With European regulations in place, standardization increases comparability across reports and qualitative differences become more distinct. As a consequence, low quality can be judged more easily, while high quality, too, becomes more apparent. In this separating equilibrium, the reliability of a company’s CSR report should serve as a viable tool to distinguish between firms that are truly committed to sustainability and companies that aren’t. On a similar note, Hoffmann and colleagues (2018) expect two groups of mandatory CSR reporters in
Germany: firms that strive to adequately inform stakeholders about sustainability topics and firms that aim only at fulfilling what is required of them, providing little credible or useful information. This line of argumentation results in the following hypotheses:
Hypothesis 2: Mandatory CSR reports of high-quality European fashion companies
will be more reliable signals than mandatory CSR reports of low-quality European fashion companies. (payoff A1 > payoff C1)
Hypothesis 3a: Mandatory CSR reports of low-quality European fashion companies
will be less reliable signals than their voluntary reports. (payoff C1 < payoff D1)
Hypothesis 4a: Mandatory CSR reports of high-quality European fashion companies
will be more reliable signals than their voluntary reports. (payoff A1 > payoff B1)
Reports by firms in the control group not subject to mandatory reporting are expected to remain largely the same for both 2017 and 2018 as no regulations are in place for those companies (i.e. no treatment). Low-quality firms’ reports should therefore be more reliable compared to the reports of low-quality mandatory reporters, whereas the reports of high-quality firms in the control group should be less reliable than the reports of high-high-quality mandatory reporters.
Hypothesis 3b: Mandatory CSR reports of low-quality European fashion companies
will be less reliable signals than CSR reports by low-quality European fashion companies that are not obliged to report under the Directive (control group). (payoff C1 < payoff C2)
Hypothesis 4b: Mandatory CSR reports of high-quality European fashion companies
will be more reliable signals than CSR reports by high-quality European fashion companies that are not obliged to report under the Directive (control group). (payoff A1 > payoff A2)
Method
Research DesignIn order to compare voluntary to mandatory reporting, a quantitative content analysis of CSR reports of Europe’s fashion houses is the method of choice. A quantitative content analysis allows a good, flexible assessment of relevant data that can easily be replicated and is not susceptible to reactive effects (Bryman, 2012). Including not only a comparison between voluntary (2017) and mandatory (2018) reports, but also a control group of
European fashion companies that voluntarily disclose sustainability information although not covered by the Directive allows a research design similar to that of an experiment with both treatment and control group. The treatment in this case is the introduction of the European Directive for reports covering the financial year of 2017.
A single coder coded all sampled reports. 18 % of the sample were double coded by two independent coders, and two rounds of coder training were completed in order to ensure intercoder reliability (see Appendix D). The total amount of coding time adds up to
approximately 120 hours.
Sample
The 2014 Directive applies to public-interest entities and large undertakings in Europe as defined by Directive 2013/34/EU, which according to the European Commission
covers approximately 6,000 large companies and groups (EC, 2014). The Textil Wirtschaft’s 2017 list of Europe’s biggest fashion brands, consisting of 179 fashion houses, provides a valuable tool to identify the biggest fashion companies in Europe by revenue and growth (Ott, 2017). Because the objective of the proposed study is to compare voluntary to
to their previous national legislations on CSR reporting. Fashion houses from France, Spain, Denmark, and Norway can therefore not be considered for this study as they have already been publishing mandatory CSR reports for years. Potential changes and implications due to mandatory reporting will become more evident by examining only countries without previous national legislation (Gulenko, 2018).
The remaining 126 companies were grouped into those to which the Directive applies (n = 54) and those that are not subject to mandatory reporting because they are either from non-EU member states (n = 9), not publicly listed, or do not fulfill the criteria to be
considered a large undertaking according to the Directive 2013/34/EU (2013) (n = 63). A large, random sample was not feasible, as, contrary to the researcher’s expectations, most of the European fashion companies identified do not provide a CSR report on their corporate websites or otherwise publicly accessible online, and requests via email were seldomly answered. First time reporters were also not considered for the sample.
Consequently, a smaller, purposive sample of the companies that did publish CSR reports both in 2017 and 2018 was drawn for the treatment group (n = 12). Five voluntary reporters from Textil Wirtschaft’s list and five reporters not included on the list were sampled for the control group (n = 10). The final sample consists of 44 English corporate social responsibility reports, equally from 2017 (n = 22) and 2018 (n = 22). More than 80 % of these reports were stand-alone sustainability reports, whereas the rest was integrated annual reports that
combine financial and non-financial information in one report. CSR reports were downloaded from corporate websites, obtained directly from the companies via email or downloaded from online databases such as the Global Reporting Initiative’s database of CSR reports. A list of all sampled companies can be found in appendix A. More than half of sampled firms were German fashion companies, 18.2 % of companies were Swedish and 11.4 % of companies were Italian. The other firms were situated in Switzerland (9.1 %), the Netherlands and
Finland (each 4.5 %). Up until 2018, half of the companies had been reporting on
sustainability for less than seven years, with a minimum of only two years and a maximum of 18 years of previous reporting (M = 7.32, SD = 4.49). Firms varied greatly in size: The largest firm by revenue, Adidas, recorded more than 21 billion euros of revenue for the year 2017, whereas the smallest firm, Björn Borg, had an annual revenue of 65 million euros (2017) (M = 2,725.91, SD = 5,926.92). H&M recorded the most employees with 171
thousand people (2017), whereas Deuter only employed 105 people in 2017 (M = 14,865.91,
SD = 36,641.17).
Measures
High- and Low-Quality Firms
An important distinction in signaling theory is that between high- and low-quality signalers. For outsiders, this quality is typically difficult to observe which is why the firm engages in signaling (Connelly et al., 2011). Consequently, a firm’s quality, in the sense of an underlying, unobservable characteristic, is a difficult concept to operationalize. This study approaches quality of European fashion companies in two ways: Firstly, sustainability
rankings, primarily the ratings on the website rankabrand.com were used to classify high- and low-quality firms with regards to their sustainability in an explicit way. On rankabrand.com, brands are assessed with yes/no-questions about their sustainability performance and
assigned letter grades from A to E according to their score. Questions include, whether the brand has initiatives in place to reduce its energy consumption, carbon emissions and
chemicals use, whether and to what extent the brand sources more sustainable raw materials, and to what extent the contracted suppliers are socially compliant with certain, established labor standards. Brands that had not been rated by rankabrand.com were assessed using the same questions. Companies were then ranked by letter grade and the amount of questions that
were answered with yes and a median split was performed. Companies with a D+ or above
belong to the high-quality group, whereas brands with a D or lower were assigned to the low-quality group.
Secondly, sampled companies were categorized into high and low quality according to the quality of the clothes they sell, to see whether, based on the quality and price of the sold product of a firm, consumers can imply a certain quality of the company. The price of a simple t-shirt for each brand was used to perform a median split. A list of sampled companies and their classification into high- and low-quality can be found in appendix A.
This twofold approach will allow insight into whether the quality of the clothes sold by a fashion company implies a certain underlying quality with regards to the firm’s CSR practices, or whether a sustainability ranking can be used to link corporate sustainability and quality.
Signal Reliability
The dependent variable of interest for this study is signal reliability which consists of the four signal properties observability, cost, honesty, and fit.
Signal observability, the extent to which others, outside of the organization can notice
the signal, was operationalized as the combination of availability (Dai, Du, Young, & Tang, 2018; Leitoniene & Sapkauskiene, 2015), accessibility (Habek & Wolniak, 2016; Lock & Seele, 2016) and readability (Lock & Seele, 2016). Variables include, for example, whether the reports can be easily downloaded and accessed, whether they are available in several languages and to what extent information within the report can easily be found via a table of contents or electronic navigation helps. Readability was measured using the Flesch-Kincaid Reading Ease Index. Overall, signal observability of reports is mediocre, with an average score of 11.2 on a scale from 0 to 20 (SD = 2.38).
Signal costs, the financial costs and the effort associated with CSR reporting, consist
of assurance costs, because obtaining external assurance is typically very expensive (Habek & Wolniak, 2016; Lock & Seele, 2016) and CSR governance costs, such as the size of the CSR department of a firm (Dai et al., 2018). Half of the reports scored below 2.45 (on a scale from 0 to 20), overall, signal costs are therefore low (M = 3.53, SD = 3.56).
The third dimension, signal honesty refers to the integrity of the signaler and the genuineness of the reported information. Variables assessing stakeholder inclusiveness,
dialogue and feedback, as well as the extent to which the methodology and procedure for data
gathering, analysis and measurement are explained and original sources are provided, constitute the integrity of the signaler (Dai et al., Habek & Wolniak, 2016; Leitoniene & Sapkauskiene, 2015; Lock & Seele, 2016). Furthermore, the firm’s commitment to CSR and the accuracy of reported information are important aspects that constitute the genuineness of information. External acknowledgements and awards as well as memberships in non-profit associations with sustainability objectives convey a credible commitment to CSR topics (Daub, 2007). Instead of merely assessing the completeness of information with regard to a number of specific performance indicators, qualitative and quantitative data were
differentiated to assess the accuracy of reported information. Several scholars have argued that in many cases, a combination of quantitative data and qualitative information is more objective and informative than mere qualitative disclosure (Dai et al., 2018; Daub, 2007; Dumitru, Dyduch, Guse, & Krasodomska, 2017; Fatima, Abdulah, & Sulaiman, 2015). Because not all performance indicators can be translated into numerical data, detailed and sufficient qualitative information was also taken into account. Consequently, performance indicators that were reported using both qualitative and quantitative data or were described in a detailed, qualitative manner, were coded more accurate than indicators presented with only
quantitative data or little qualitative information. The mean for signal honesty shows that sampled reports on average achieved around 9 out of 20 points (M = 9.13, SD = 3.55).
Lastly, signal fit was operationalized as the timeliness and relevance of provided information to appropriately signal a commitment to sustainable business conduct in the fashion industry. Accordingly, the Global Reporting Initiative’s sector supplement for the apparel and footwear sector (2008) was used to define the 34 most relevant, material
performance indicators essential to capture sustainability along the industry’s complex supply chains. Signal fit was overall mediocre (M = 11.46, SD = 2.42 ), with a minimum of 7.06 and a maximum of 16.08 (scale 0 to 20).
All four dimensions were weighted equally (25 %) and added to represent the signal
reliability of a report with a maximum potential score of 80. Within the sample, the variable
is approximately normally distributed, with a minimum score of 17.82 (Calida, 2018), an average score of 35.32 (SD = 8.76), and a maximum score of 52.91 (Hugo Boss, 2018). Overall, reports are therefore only moderately reliable. The top five most reliable reports were issued by Benetton in 2018, Puma (both years), and Hugo Boss (both years), the only firm with reliability scores above 50 is Hugo Boss. The bottom two reports with reliability scores below 20 were published by the Swiss company Calida.
Table 2 provides an overview of the four signal properties, their sub-dimensions and the respective coded variables. The complete codebook as well as the values for
Krippendorff’s Alpha can be found in the appendix.
In line with previous studies (Bae et al., 2018; Daub, 2007; Lock & Seele, 2016; Michelon et al., 2015), the most recent Global Reporting Initiative’s guidelines (2016) were used as a framework to develop the codebook and assess the quality of reports. The GRI’s reporting guidelines are the first and most widely used standards for CSR reporting: In 2017,
75 % of the largest 250 companies in the world applied the GRI standards to report on sustainability (KPMG, 2017).
Table 2. Operationalization of the four signal properties and respective coded variables.
Observability Cost
Sub-dimension Variables Sub-dimension Variables
Availability SO1, SO2, SO3 Assurance Cost SC1, SC2, SC3, SC4
Accessibility SO4.1, SO4.2, SO4.3, SO4.4, SO4.5, SO4.6 CSR Governance Cost SC5, SC6 Readability SO5
Results
Firstly, to confirm that all four signal properties contribute to signal reliability, a principle component factor analysis was conducted. The PCA shows that the four dimensions form a single uni-dimensional scale to measure signal reliability: only one factor has an eigenvalue above one (eigenvalue 2.19) and there is a clear point of inflexion after this component in the scree plot. Together, these factors explain 54.68 % of the variance in signal
reliability. All items correlate positively with the first component, the variable signal honesty
has the strongest association (factor loading is .90). The reliability of the scale is reasonable, Cronbach’s alpha = .69.
Honesty Fit
Sub-dimension Variables Sub-dimension Variables
Stakeholder Inclusiveness SH1.1, SH1.2, SH1.3 Timeliness SF1, SF2 Feedback SH2 Sector-specific Completeness SF3 Methodology SH3.1, SH3.2, SH3.3, SH3.4 CSR Commitment SH4, SH5.1, SH5.2, SH5.3, SH5.4, SH5.5 Accuracy SH6.1 – SH6.34
All of the following results need to be interpreted with caution as normal distribution of signal reliability per group is not given and groups are small in size, the statistical power is therefore low.
H1a predicted an increase in reliability for mandatory (2018) CSR reports compared
to previous voluntary (2017) reports in the treatment group. An independent samples t-test revealed no significant differences between mandatory CSR reports of European fashion companies (c) and their previous, voluntary reports (M = 37.47, SD = 9.58), t(22) = - 0.28, p = .392. H1a is therefore not supported.
Hypothesis 1b stated that mandatory CSR reports published by companies in the
treatment group will be more reliable signals than those by companies in the control group. Firstly, an independent samples t-test was performed to check whether any qualitative differences between reporting years within the control group exist. Results show no significant findings: 2017 reports and 2018 reports by voluntary reporters were not significantly different with regards to reliability, t(18) = - 1.09, p = .291. An independent samples t-test was then conducted to compare the group means between voluntary and mandatory reporters for the year 2018. No significant differences were found between the reliability of mandatory CSR reports by European fashion companies in the treatment group (M = 38.57, SD = 9.90) and that of CSR reports by companies in the control group (M = 33.66, SD = 7.02), t(20) = 1.31, p = .102. H1b is therefore not supported.
Hypothesis 2 predicted that mandatory CSR reports of high-quality companies will be
more reliable than those of low-quality firms. An independent samples t-test was performed to compare the group means for signal reliability. Group means were not significantly
different when quality of clothing was used as an indicator of a firm’s quality, t(10) = 0.38, p = .355, whereas when using the firm’s CSR rating as a sign of quality, results came close to statistical significance: High-quality firms in the treatment group achieved higher reliability
scores for their mandatory reports (M = 43.15, SD = 7.15), than low-quality companies in the treatment group (M = 33.99, SD = 10.70), t(10) = 1.74, p = .056, 95 % CI [-2.55, 20.87], d = 1.10. Although this effect is large, due to lack of significance, H2 cannot be supported.
For H3a a decrease in reliability was expected for mandatory reports of low-quality European fashion companies when compared to their previous voluntary reports. Results of an independent samples t-test show no significant differences, neither between the mandatory and voluntary reports for firms with low-quality clothing, t(10) = 0.20, p = .425, nor between the mandatory and voluntary reports for firms with low CSR ratings, t(10) = 0.12, p = .454. Hence, H3a is not supported.
Hypothesis 3b tested whether mandatory CSR reports of low-quality European
fashion companies are less reliable than reports by low-quality firms in the control group. Similar to the findings for H3a, the independent samples t-test revealed no significant
differences when comparing firms with low-quality clothing in the treatment with those in the control group, t(8) = 0.01, p = .498, nor when comparing firms with low CSR ratings in the treatment with those in the control group, t(9) = 0.52, p = .310. Results do therefore not support H3b.
For H4a, it was hypothesized that mandatory CSR reports of high-quality European fashion companies will be more reliable signals than their voluntary reports. To test this hypothesis, an independent samples t-test was performed. No significant differences were found between the mandatory and voluntary reports for firms with high-quality clothing,
t(10) = 0.19, p = .425, or between the mandatory and voluntary reports for firms with high
CSR ratings, t(10) = 0.35, p = .369. These findings do not support H4a.
Lastly, H4b predicted a higher reliability for mandatory CSR reports of high-quality firms when compared to the reports by high-quality firms in the control group. An
Again, group means were not significantly different when quality of clothing was used as an indicator of a firm’s quality, t(10) = 1.44, p = .09, whereas when using the firm’s CSR rating as a sign of quality, results came close to statistical significance: Highly rated firms in the treatment group achieved better reliability scores for their mandatory reports (M = 43.15, SD = 7.15), than high-quality fashion companies in the control group (M = 36.32, SD = 5.39),
t(9) = 1.75, p = .057, 95 % CI [-1.98, 15.63], d = 1.17. Although this effect is large, due to
lack of significance, H4b cannot be supported.
Although no significant differences were found and findings cannot be generalized, the tendencies for the sample means are presented in an equivalent to table 2 in appendix B.
In order to test for third variables that might influence the reliability of reports, a linear regression analysis with the control variables size of the company, experience in CSR
reporting, and a dummy variable reporting format was performed. The size of the company,
both measured in revenue and the number of employees, as well as the amount of years that the company had been reporting on CSR already, and whether the report was integrated or standalone did not have significant effects on signal reliability.
For the reason that one of the central aspects in signaling theory is the distinction between high- and low-quality firms, further analyses were performed, not taking mandatory and voluntary CSR reporting into account as the analyses above did not result in significant findings.
An independent samples t-test shows that, overall, the reports issued by firms with higher CSR ratings (high-quality) were significantly more reliable (M = 39.07, SD = 7.07), than reports issued by firms with lower CSR ratings (low-quality) (M = 31.56 , SD = 8.80),
t(42) = 3.12, p = .003, 95 % CI [2.65, 12.37], d = 0.96, independent of whether they belong to
firm’s sold clothing, no significant differences with regards to the reliability of reports were found, t(42) = - 0.67, p = .508.
Furthermore, a one-way ANOVA was conducted to compare the group means for high- and low-quality firms in the treatment and control group respectively. The analysis of variance showed a large significant effect, F(3) = 6.03, p = .002, η² = .31. The Bonferroni post hoc test revealed significant differences between high-quality firms in the treatment group that issued significantly more reliable reports (M = 42.40, SD = 7.18) than low-quality firms in the treatment group (M = 33.64, SD = 9.86), p = .041, 95 % CI [0.23, 17.31]. High-quality firms in the treatment group also issued significantly more reliable reports than low-quality firms in the control group (M = 29.07, SD = 7.03), p = .001, 95 % CI [4.37, 22.29]. No significant differences were found between the other groups. These results remain significant only when using the firm’s high or low CSR rating as an implication of quality and when combining both years of reporting.
Discussion
SummaryThis study set out to examine the extent to which mandatory reports published by European fashion companies are more reliable signals of CSR commitment than their previous, voluntary reports and to what extent these results prove to be persistent when compared to a control group of voluntary reporters and when taking the quality of the firm into account. Overall, signal reliability was mediocre across both years of reporting and findings do not indicate that, within the European fashion industry, mandatory reports are more reliable than voluntary reports or the reports published by companies in the control group. Furthermore, no significant differences were found between the voluntary and mandatory reports of high- and low-quality fashion firms, although findings tend towards
more reliable, mandatory reports for firms with higher CSR ratings compared to lower-quality firms in the treatment group and similar, high-lower-quality firms in the control group.
When ignoring the year of reporting, i.e. the differentiation between voluntary and mandatory reporting, noticeable, significant differences were found between firms with higher CSR ratings and firms with lower CSR ratings. It seems that European disclosure regulations did not change the signaling setting from a pooling to a separating equilibrium as expected, but that CSR reporting within the European fashion industry in general serves as a signal to distinguish between high- and low-quality firms, as low-quality fashion companies are not able to achieve the same level of reporting reliability as high-quality firms. A CSR report’s reliability does seem to be linked to a firm’s commitment to sustainability and therefore can be a useful signal for a high-quality firm to reduce information asymmetry between the firm and its stakeholders, communicate its commitment to CSR and to
distinguish itself from lower-quality competitors. Additionally, stakeholders can judge the firm’s true quality according to its CSR report. Furthermore, CSR ratings seem to be good, explicit indicators of a firm’s quality when it comes to sustainable business practices,
whereas the quality of a firm’s sold products does not seem to indicate this underlying quality well.
Implications
Firstly, the findings of this study can, to some extent, explain why there are some inconsistent findings with regards to the credibility and quality of mandatory CSR reporting (e.g. Habek & Wolniak, 2016; Lock & Seele, 2016). Within the European fashion industry, the reliability of a CSR report seems to be largely related to the quality of the firm rather than the extent to which reporting is mandated by European law.
Furthermore, an important indication of these findings is that the EU might have to increase its regulations when it comes to what firms are required to report, as the reliability of reports is generally mediocre, and the current provisions do not seem to increase reporting reliability, at least for the fashion industry in Europe. What came to the researcher’s attention when drawing the sample was that most companies in the population, although by European law mandated to, did not publish a sustainability report at all, or none that was publicly accessible or obtainable via e-mail contact. A missing body exercising control over the extent to which large European companies do comply with the EU Directive and a lack of
consequent sanctions seem to be resulting in little compliance by European fashion companies.
Limitations & Future Research
Of course, this study is not without its limitations. Due to the fact that many listed fashion companies did not provide a CSR report, a purposive sample had to be drawn. Hence, no inferences can be made about first time reporters and companies that, if they publish a report at all, do not have it publicly available. Lack of observability in this case may indicate a low quality of reporting, which makes the sample more selective, as some of the most unreliable reports might just not be accessible to stakeholders. Consequently, the groups for both the treatment and control group, as well as the groups of high- and low-quality are small in size. Findings therefore need to be interpreted with caution, as signal reliability is not normally distributed for the various groups, which decreases the meaningfulness of group means and the significance of findings. Furthermore, the sampling method resulted in more than half of the sample being comprised of German fashion companies. Statistically, country comparisons were therefore not sensible. Additionally, fashion companies from countries with previous national legislation on mandatory CSR reporting were intentionally excluded
from the sample. The findings of this study are therefore not applicable to countries other than those without previous national regulations on CSR reporting. Future studies with a larger scope might want to include these countries, too, and create sub-samples, as France and Spain do form a large proportion of the European fashion industry.
With regards to the codebook, it needs to be noted, that some of the coded variables did not achieve acceptable reliability scores (see Appendix D). Another round of coder training could have improved these values. Nevertheless, the codebook was used as is, as these variables are few, and have only minor influence on the total score for signal reliability.
One last noteworthy aspect with regards to the study’s limitations is the focus on CSR reports as means of communication between a firm and its stakeholders. Regulations on mandatory sustainability reporting might also spur a firms’ other communication channels, as well as increase media coverage of corporate CSR initiatives. Including other information channels that companies might use to communicate about their CSR activities might influence the overall effect of mandated CSR reporting (Gulenko, 2018).
Notwithstanding these limitations, this study showed that the high or low quality of a fashion company matters largely when it comes to the reliability of its CSR reporting. Future research could apply this differentiation to other industries to see whether these findings are consistent across various, socially and environmentally impactful business sectors, as well as operationalize a firm’s quality in different, industry-specific ways. Other studies, focusing on voluntary versus mandatory CSR reporting should take the quality of a firm into account, especially when faced with inconsistent or inconclusive findings.
References
Bae, S., Masud, A., & Kim, J. (2018). A cross-country investigation of corporate governance and corporate sustainability disclosure: A signaling theory perspective. Sustainability,
Bryman, A. (2012). Social research methods. Oxford: Oxford University Press. Chauvey, J-N., Giorano-Spring, S., Cho, C., & Patten, D. (2015). The normativity and
legitimacy of CSR disclosure: Evidence from France. Journal of Business Ethics, 130, 789-803.
Ching, H.Y., & Gerab, F. (2017). Sustainability reports in Brazil through the lens of
signaling, legitimacy and stakeholder theories. Social Responsibility Journal, 13(1), 95-110.
Cohen, B. D., & Dean, T. J. (2005). Information asymmetry and investor valuation of IPOs: Top management team legitimacy as a capital market signal. Strategic Management
Journal, 26(7), 683-690.
Connelly, B.L., Certo, S.T., Ireland, R.D., & Reutzel, C.R. (2011). Signaling Theory: A review and assessment. Journal of Management 37(1), 39-67.
Crane, A., & Glozer, S. (2016). Researching corporate social responsibility communication: themes, opportunities and challenges. Journal of Management Studies, 53(7), 1223-1252.
Criado-Jiménez I., Fernández-Chulián M., Husillos-Carqués F., & Larrinaga-González C., (2008). Compliance with mandatory environmental reporting in financial statements: The case of Spain (2001–2003). Journal of Business Ethics, 79(3), 245–262.
Dai, N.T., Du, F., Young, S.M., & Tang, G. (2018). Seeking legitimacy through CSR
reporting: Evidence from China. Journal of Management Accounting Research, 30(1), 1-29.
Daub, C. (2007). Assessing the quality of sustainability reporting: an alternative methodological approach. Journal of Cleaner Production, 15, 75-85.
Devin, B. (2016). Half-truths and dirty secrets: Omissions in CSR communication. Public
Relations Review, 42, 226-228.
Dögl, C., & Holtbrügge, D. (2014). Corporate environmental responsibility, employer reputation and employee commitment: an empirical study in developed and emerging economies. The International Journal of Human Resource Management, 25(12), 1739-1762.
Du, S.; Bhattacharya, C.B. & Sen, S. (2010). Maximizing business returns to corporate social responsibility (CSR): The role of CSR communication. International Journal of
Dumitru, M., Dyduch, J., Guse, R., & Krasodomska, J. (2017). Corporate reporting practices in Poland and Romania – An ex-ante study to the new non-financial reporting
European Directive. Accounting in Europe, 14(3), 279-304.
European Commission – EC (2014). DIRECTIVE 2014/95/EU OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 22 October 2014 amending
Directive 2013/34/EU as regards disclosure of non-financial and diversity information by certain large undertakings and groups. Retrieved from
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32014L0095
Fatima, A., Abdullah, N, & Sulaiman, M. (2015). Environmental disclosure quality:
examining the impact of the stock exchange of Malaysia’s listing requirements. Social
Responsibility Journal, 11(4), 904-922.
Garcia-Torres, S., Rey-Garcia, M., & Albareda-Vivo, L. (2017). Effective disclosure in the fast-fashion industry: from sustainability reporting to action. Sustainability, 9(12), 2256.
Global Reporting Initiative (GRI) (2008). Sustainability guidelines & apparel and footwear sector supplement, pilot version.
Global Reporting Initiative (GRI) (2016). GRI 102 General Disclosures. Retrieved from
https://www.globalreporting.org/standards/gri-standards-download-center
Gulenko, M. (2018). Mandatory CSR reporting – literature review and future developments in Germany. Sustainability Management Forum, 26(3), 3-17.
Habek, P. (2018). Does reporting obligation affect credibility of information disclosed in corporate social responsibility reports? at Enterprise and Competitive Environment, March, 2018.
Habek, P., & Wolniak, R. (2016). Assessing the quality of corporate social responsibility reports: The case of reporting practices in selected European Union member states.
Qual Quant, 50, 399-420.
Hahn, R. & Kühnen, M. (2013). Determinants of sustainability reporting: a review of results, trends, theory, and opportunities in an expanding field of research. Journal of Cleaner
Production, 59, 5-21.
Hetze, K. (2016). Effects on the (CSR) reputation: CSR reporting discussed in the light of signaling and stakeholder perception theories. Corporate Reputation Review, 19(3), 181-296.
Hoffmann, E., Dietsche, C., & Hobelsberger, C. (2018). Between mandatory and voluntary: non-financial reporting by German companies. Sustainability Management Forum,
26(1), 47-63.
Ioannou, I., & Serafeim, G. (2017). The consequences of mandatory corporate sustainability reporting. Retrieved from
https://papers.ssrn.com/soL3/papers.cfm?abstract_id=1799589
KPMG (2017). The road ahead. The KPMG survey of corporate responsibility reporting 2017. Retrieved from https://assets.kpmg/content/dam/kpmg/xx/pdf/2017/10/kpmg-survey-of-corporate-responsibility-reporting-2017.pdf
Leitoniene, S., & Sapkauskiene, A. (2015). Quality of corporate social responsibility information. Social and Behavioral Sciences, 213, 334-339.
Llena F., Moneva J., & Hernandez B. (2007). Environmental disclosures and compulsory accounting standards: The case of Spanish annual reports. Business Strategy and the
Environment, 16(1), 50–63.
Lock, I., & Schulz-Knappe, C. (2018). Credible corporate social responsibility (CSR)
communication predicts legitimacy. Evidence from an experimental study. Corporate
Communications: An International Journal, 24(1), 2-20.
Lock, I., & Seele, P. (2016). The credibility of CSR (corporate social responsibility) reports in Europe. Evidence from a quantitative content analysis in 11 countries. Journal of
Cleaner Production, 122, 186-200.
Mahoney, L., Thorne, L., Cecil, L., & LaGore, W. (2013). A research note on standalone corporate social responsibility reports: Signaling or greenwashing? Critical
Perspectives on Accounting, 24, 350-359.
Mammut Sports Group AG (2017). Social Report 2016.
Michelon, G., Pilonato, S., & Ricceri, F. (2015). CSR reporting practices and the quality of disclosure: An empirical analysis. Critical Perspectives on Accounting, 33, 59-78. Milne, J., & Gray, R. (2013). W(h)ither Ecology? The Triple Bottom Line, the Global
Reporting Initiative, and Corporate Sustainability Reporting. Journal of Business
Ethics, 118(1), 13-29.
Ott, M. (2017, September 27). TW-Ranking Modemarken-Anbieter: Inditex, H&M und Adidas sind die Größten. TextilWirtschaft. Retrieved from
https://www.textilwirtschaft.de/business/markt/exklusiv-tw-ranking-inditex-hm-und-adidas-sind-die-groessten-206663
Patriotta, G., Gond, J. P., & Schultz, F. (2011). Maintaining legitimacy: Controversies, orders of worth, and public justifications. Journal of Management Studies, 48(8), 1804-1836. Pedersen, Neergaard, Pedersen, & Gwozdz, (2013). Conformance and deviance: Company
responses to institutional pressures for corporate social responsibility reporting.
Business Strategy and the Environment, 22, 357-373.
Shabana, K.M., Buchholtz, A.K., & Carroll, A.B. (2016). The institutionalization of corporate social responsibility reporting. Business & Society, 56(8), 1107-1135. Simaens, A., & Koster, M. (2013). Reporting on sustainable operations by third sector
organizations: A signalling approach. Public Management Review, 15(7), 1–23. Spence A. 1973. Job market signaling. Quarterly Journal of Economics, 87(3), 355–379. Tschopp, D. & Huefner, R. (2015). Comparing the Evolution of CSR Reporting to that of
Financial Reporting. Journal of Business Ethics, 127(3), 565-577.
White, C.L., Nielsen, A. E., & Valentini, C. (2017). CSR research in the apparel industry: A quantitative and qualitative review of existing literature. Corporate Social
Appendix A: List of Sampled Companies by Turnover 2017
Company Country Turnover
in million € 2017 No. of employees 2017 Sustainability rating High- vs. low quality
Adidas Group Germany 21,218 56,888 D+ High
H&M Sweden 20,747 171,000 C Low
Puma Germany 4,136 11,389 C High
Hugo Boss Germany 2,733 13,966 D+ High
*KiK Germany 2,004 25,313 D- Low
Esprit Germany 1,761 7,304 D+ Low
OVS Italy 1,526 6,821 D- Low
Benetton Group Italy 1,280 7,505 D- High *Engelbert Strauss Germany 1,050 1,200 D+ Low
Stockmann Finland 606 7,325 D Low
*Zeeman Netherlands 585 7,593 C Low
Marc O’Polo Germany 387 1,842 D- High
*Calida Switzerland 343 1,063 E High
*Brax Leineweber Germany 314 1,251 E High *Olymp Bezner Germany 286 850 D- Low
Ahlers Group Germany 236 2,062 D- High
Gina Tricot Sweden 210 1,892 D+ Low
*Mammut Switzerland 206 752 D+ High
*Schöffel Germany 100 217 C High
*Deuter Germany 100 105 C High
*Vagabond Shoemakers
Sweden 77 500 D- High
Björn Borg Sweden 65 212 D- Low
Appendix B
Table 3. Group means for high- vs. low-quality firms in treatment and control group.
2017 (t1) Treatment Group M = 37.47, SD = 9.58 Control Group M = 30.49, SD = 5.98 B1 High-quality clothing M = 38.31, SD = 12.22 D1 Low-quality clothing M = 36.63, SD = 7.14 B2 High-quality clothing M = 28.85, SD = 5.84 D2 Low-quality clothing M = 32.96, SD = 6.08 High-quality CSR ranking M = 41.66, SD = 7.82 Low-quality CSR ranking M = 33.28, SD = 9.94 High-quality CSR ranking M = 33.84, SD = 3.84 Low-quality CSR ranking M = 27,15, SD = 6.15 2018 (t2) Treatment Group M = 38.57, SD = 9.90 Control Group M = 33.66, SD = 7.02 A1 High-quality clothing M = 39.71, SD = 12.67 C1 Low-quality clothing M = 37.43, SD = 7.24 A2 High-quality clothing M = 31.16, SD = 7.12 C2 Low-quality clothing M = 37.41, SD = 5.67 High-quality CSR ranking M = 43.15, SD = 7.15 Low-quality CSR ranking M = 33.99, SD = 10.70 High-quality CSR ranking M = 36.32, SD = 5.39 Low-quality CSR ranking M = 31.00, SD = 8.01
Appendix C: Codebook
Administrative Items
AI1 Name of the coder.
• (1.00) Theresa (Coder1) • (2.00) Vincent (Coder2)
AI2 Date of coding.
dd.mm.yy
AI3 Report specific ID.
(year covered by the report/company name). Open code _____________
AI4 Company name.
Open code _____________
AI5 Country of origin of the firm.
Check where the company’s head office is located or consult Appendix 1. • (1.00) Italy • (2.00) Germany • (3.00) Sweden • (4.00) Netherlands • (5.00) Poland • (6.00) Belgium • (7.00) Finland • (8.00) Austria • (9.00) Switzerland
AI6 Year of report. The year in which the report was published.
• (1.00) 2017 • (2.00) 2018
AI7 Reporting period. The year the report refers to. In case it does not refer to the calendar
year, code the most recent year (e.g., reporting period: 09/2016-09/2017, code 2017). Year: __________
AI8 Type of report.
• (1.00) Stand-alone • (2.00) Integrated,
AI9 First report issued. Year when the first CSR report was issued by the firm. If no
information on this is reported, check the company’s website, their archive for CSR reports, or email the company to find out.
Year: _____________
AI10 Company size.
AI10.1 Size by revenue. Enter the company’s revenue for the financial year of
2017 in million euros according to the Textil Wirtschaft’s list of the biggest European fashion houses. (consult Appendix 1)
Open code _______________ million €
AI10.2 Size by number of employees. Enter the number of employees for the
year 2017. This number is usually given in the CSR report in chapters like “xx company in numbers” “our employees”, “our personnel” etc.
Variables: Signal Observability
AvailabilityClarity: The reporting organization shall make information available in a manner that is understandable and accessible to stakeholders using that information.
SO1 Online availability. Information is available to stakeholders on the company’s website,
can easily be downloaded without giving credentials, logging in, or payment. Start looking for the report on the corporate website www.company.com
• (1.00) Yes (1) • (0.00) No (0)
SO2 Online availability (2). Where is the report located on the corporate website?
Please enter the number of tiers. Use the accessed URL (used to download the report) and count the number of slashes in that URL after the global website (start with: www.company.com), then add one to that number. Don’t count / that are part of the document title of the report.
Example: Enter 5 for the following link as there are 4 slashes after www.inditex.com, and the pdf for the report is located on the 5th tier
1 2 3 4 5
https://www.inditex.com/documents/10279/563475/2017+Inditex+Annual+Report.pdf/f5bebf a4-edd2-ed6d-248a-8afb85c731d0
Open code: tier report is located on (enter a number): _______________ • Report not available online (0)
SO3 Languages available. The report is available in several languages. Check if the report
can be downloaded in several languages or try switching the language of the website to another language. Code only one answer here.
• (1.00) Report available in English. (1)
• (2.00) Report available in English and a second language, i.e. the language of the country of the company’s location (e.g. German for a German company). (2)
• (3.00) Report available in more than two languages (e.g. English, German and additional languages for a German company). (3)
Accessibility
Stakeholders can find the specific information they want without unreasonable effort through tables of contents, maps, links, or other aids.
SO4 Accessibility. Stakeholders can find the specific information they want without
unreasonable effort through tables of content, maps, links, or other aids.
SO4.1 Table of contents/map of contents (1) • (1.00) Yes (1)
• (0.00) No (0)
SO4.2 Electronic navigation helps in PDF (1)
i.e. interactive PDF, clicking on the headings in the table of contents directs you to the corresponding chapter
• (1.00) Yes (1) • (0.00) No (0)
SO4.3 Chapter headings close to page number (1)
i.e. furthest apart: chapter heading on the top left of the page and page number at the bottom right of the page; instead: page number and chapter heading are both at the bottom/top of the page forming one line
• (1.00) Yes (1) • (0.00) No (0)
SO4.4 Impressum/Imprint (1)
The impressum should include the publisher of the report, which in most cases would be the company. The company name and address should be included. If it’s only a contact or feedback option please code No and code Yes for SH2 Feedback.
• (1.00) Yes (1) • (0.00) No (0)
SO4.5 PDF open (1)
copy/paste of the data possible • (1.00) Yes (1)
