A PPAREL I NDUSTRY R ESPONSIBILITY IN THE C ORPORATE S OCIAL T HE C OMMUNICATION OF

(1)

T

HE

C

OMMUNICATION OF

C

ORPORATE

S

OCIAL

R

ESPONSIBILITY IN THE

A

PPAREL

I

NDUSTRY

–

A

M

ULTI

-C

HANNEL

I

NVESTIGATION

Thesis number: 2020.18.05 Thesis for One-Year Master, 15 ECTS

(2)

I

Title: The Communication of Corporate Social Responsibility in the Apparel Industry: A Multi-Channel Investigation

Publication year: 2020

Author: Lidy van Deursen

Supervisor: Jonas Larsson Abstract:

While corporate social responsibility (CSR) disclosure has proliferated in the apparel industry since the 1990s, CSR communications research remains underdeveloped. While the extant literature investigates how brands utilise CSR reports and websites to communicate CSR, little attention has been given to other communication channels. In response, this thesis explores how apparel brands use different corporate communication channels to communicate CSR to stakeholders. Using ethnographic content analysis, the CSR reports, corporate websites and Facebook posts of thirty apparel brands were analysed. Findings indicate that brands approach CSR communication strategically with approximately two-thirds of brands consistently using the same channel combination for each CSR topic, while the remaining third employ a dynamic approach, using different channel combinations to discuss different CSR topics. While CSR reports and websites show a typically moderate-high level of utilisation for the discussion of CSR topics, the same CSR topics are considerably less likely to be discussed over Facebook. Facebook CSR communication is generally selective, with contentious issues seldom addressed via this channel. There is little consistency between brands in how they approached CSR communication, with each brand employing a combination of strategies to legitimise their business practices and demonstrate their good corporate citizenship. This study empirically supports and extends research conducted on single channel CSR communications of apparel brands and suggests that brands are cognizant of the business case for CSR as well as the repercussions of poorly executed CSR communication.

Keywords: Corporate social responsibility; CSR Communication; Apparel brands;

(3)

II

2.9 Framing CSR ... 15 2.10 Summary of CSR Literature ... 16 3 Theoretical Framework ... 16 3.1 Legitimacy Theory ... 16 3.1.1 Institutional Legitimacy ... 17 3.1.2 Organisational Legitimacy ... 17 3.2 Stakeholder Theory ... 17 3.3 Institutional Theory ... 18 3.4 Signalling Theory ... 18 3.5 CSR Communication Framework... 19 3.6 Synthesis of Theories ... 20 4 Methodological Framework ... 22 4.1 Approach ... 22

4.1.1 Ethnographic Content Analysis ... 22

4.2 Sampling ... 23

4.2.1 Sample Size and Description ... 24

4.2.2 Sampling Limitations ... 25

4.2.3 Unit of Analysis... 25

4.3 Data Collection ... 26

4.3.1 CSR Reports ... 26

4.3.2 CSR Section of the Company Website... 26

4.3.3 Facebook Data ... 26

4.4 Software ... 26

4.4.1 Facepager... 27

4.4.2 ATLAS.ti ... 27

4.5 Qualitative Document Analysis Framework ... 28

4.6 Coding... 29

4.6.1 Deductive Coding ... 29

4.6.2 Inductive Coding ... 29

4.6.3 Coding Schedule... 30

4.7 Research Quality ... 31

5 Results and Analysis ... 32

(4)

III

5.2 CSR Communication Channel Utilisation ... 34

5.3 CSR Communication Channel Consistency ... 37

5.4 CSR Communication Strategies ... 40

5.4.1 Withholding CSR Communication Strategy ... 42

5.4.2 Calculated CSR Communication Strategy ... 42

5.4.3 Dynamic CSR Communication Strategy ... 44

5.4.4 Strategic CSR Communication Strategy ... 46

5.5 CSR Topic Engagement ... 47

5.6 CSR Communication Channel Content ... 49

5.6.1 CSR Issues with No Social Media Disclosure ... 50

5.6.2 Consistency in Messaging across CSR Communication Channels ... 51

5.6.3 Variance in Messaging across CSR Communication Channels ... 52

6 Discussion ... 54

6.1 The Strategic Communication of CSR ... 54

6.2 CSR Communication as a Pathway to Sustainability ... 56

7 Conclusion ... 59 7.1 Recommendations of Research ... 59 7.2 Contribution ... 60 7.3 Limitations ... 60 7.4 Future Research ... 61 References ... 63

Appendix A: Channel Utilisation by CSR Topic ... 70

Appendix B: CSR Communication Channel Utilisation, by Brand ... 71

List of Figures

Figure 1: Framework for CSR Communication ... 20

Figure 2: Relationship Between Theories that Concern Legitimacy ... 21

Figure 3: CSR Topic Engagement, by Channel ... 35

Figure 4: Communication Channel Combination Approaches, by CSR Topic ... 38

Figure 5: Communication Channel Combination Approaches, by Brand ... 39

Figure 6: CSR Communication Strategy Matrix ... 41

Figure 7: Percentage of Companies Addressing CSR Topics, by Communications Strategy ... 47

Figure 8: Engagement with CSR Topics, by Communications Strategy ... 49

List of Tables

Table 1: Descriptive Statistics of Sampled Companies ... 24

Table 2: Process of Qualitative Document Analysis ... 29

Table 3: ATLAS.ti Codebook ... 30

Table 4: CSR Communication Channel Utilised, by Company ... 33

Table 5: Dominant CSR Channel Combination, by Company... 40

Table 6: Faherty Brand Engage with CSR Topics Closely Linked to Brand Identity ... 43

Table 7: ASOS Uses a Dynamic CSR Communication Strategy... 45

Table 8: CSR Topic: Activism ... 48

Table 9: Esprit Demonstrates Consistent Communication about Transparency across All Channels ... 52

Table 10: Kathmandu Discuss Health and Safety ... 53

Table 11: Channel Utilisation by CSR Topic ... 70

(5)

IV

Abbreviations and Acronyms

AEO American Eagle Outfitters

API Application Programming Interface

B2B Business to Business

B2C Business to Consumer

CAQDAS Computer-aided qualitative data analysis software

CEO Chief Executive Officer

CSR Corporate Social Responsibility

CSiR Corporate Social Irresponsibility

CTSCA California Transparency in Supply Chains Act of 2010

ECA Ethnographic Content Analysis

FB Facebook

GHG Greenhouse Gas

GRI Global Reporting Initiative

LCA Life-cycle Assessment

MNC Multi-national Corporation

NGO Non-governmental Organisation

PDF Portable Document Format

RQ Research Question

SAC Sustainable Apparel Coalition

SDGs United Nations Sustainable Development Goals

SME Small and Medium Enterprises

UK The United Kingdom

UN United Nations

(6)

1 Introduction

1.1 Background

Within the context of globalisation1, the trend of outsourcing textile and garment production

has resulted in corruption, discrimination and human rights violations (Perry and Towers 2009) and subsequent criticism from stakeholders. In this context, corporate social responsibility (CSR) and CSR communication have been frequently promoted as a method capable of managing the complex, negative aspects of this supply chain configuration as well as public perception. Initially concerned with social issues, CSR practice has developed over time, and the definition now frequently includes environmental preservation principles in addition to social welfare considerations. In line with this development, modern CSR reporting standards typically advocate for reporting in line with triple bottom line principles, stipulating the importance of social and environmental performance alongside traditional economic measures (Da Giau et al. 2016; Reilly and Larya 2018). Moreover, CSR reporting is becoming common practice with KPMG (2017) reporting that of the world’s 250 largest corporations, 93% are now producing an annual CSR report, in comparison to 35% in 1999.

CSR can be considered a method of addressing the social and environmental harm resulting from corporate activities (Perry and Towers 2009); therefore, CSR communication is how companies demonstrate their corporate citizenship, legitimise their practices and gain both social acceptance and the ongoing right to operate (Alam and Akter 2019). Colucci, Tuan and Visentin (2020) discuss CSR engagement, designating the activities that companies engage in as either CSR walk or CSR talk. CSR talk encompasses “the external communication tools instrumentally used by a company to inform stakeholders about environmental and social initiatives;” (p.2) whereas CSR walk refers to the actions taken by the company to reduce environmental impact or improve social conditions both within the firm and across the firms greater value chain (Colucci, Tuan and Visentin 2020; Wickert, Scherer and Spence 2016). Both CSR walk and CSR talk are crucial as one without the other will inevitably achieve sub-optimal results (Du, Bhattacharya and Sen 2010) and invite stakeholder scepticism (Childs, Woo and Kim 2019). While successful CSR implementation requires both implementation and communication to be addressed, this thesis will focus primarily on CSR communication. This decision is supported by Ban (2019) who argues that in an academic context, it is appropriate to treat CSR discourse as separate from CSR practice; and furthermore, by White, Neilsen and Valentini who encourage further research in the area, stating, “CSR communication in research about the apparel industry is understudied” (2017, p. 387).

1 “Globalisation describes the large-scale shift of labour-intensive manufacturing operations of Western apparel

(7)

2

CSR communication can take many forms. Formal CSR communication will often take the form of CSR reports, codes of conduct, a dedicated section in an annual report, a dedicated section on a company website (Colucci, Tuan and Visentin 2020), or formal communication from company executives (Kim and Kim 2017). Informal CSR communication usually takes the form of marketing and can be in the form of website information, social media or instore marketing material (Reilly and Larya 2018).

1.2 Importance of CSR Communication Research

White, Nielsen and Valentini (2017) who extensively review the academic literature, find that CSR research in the apparel industry has proliferated since 2012; stating however, that it continues to lag behind the CSR research in other industries. In addition, they consider the approaches to studying CSR underdeveloped and fragmented, stating that the “communicative dimension of engaging stakeholders in dialog ... [has] not yet [been] specifically addressed in the fashion and apparel industry” (White, Nielsen and Valentini 2017, p. 390). As such, the study of CSR communication in the apparel industry is important from a theoretical perspective as there is minimal research on how brands use different platforms to communicate CSR initiatives.

The study of CSR communication also has implications for corporations, stakeholders and the greater field of textile management. Firstly, by reviewing the cross-channel CSR communications of apparel brands, different approaches to CSR communication can be identified, presenting brands with the opportunity to refine their communication strategy to better communicate with their stakeholders. Secondly, this research will act as a guide for stakeholders who are wanting to learn more about how to seek out CSR information relevant to their informational needs. Finally, the study of CSR communication can be used to further the United Nations (UN) Sustainable Development Goals (SDGs), which make “sustainability reporting a clear expectation” (Global Reporting Initiative and United Nations Global Compact Office 2015). The most relevant contribution of this research would be towards SDG goals 12.6 and 12.8:

SDG Target 12.6: Encourage companies, especially large and transnational

companies, to adopt sustainable practices and to integrate sustainability information into their reporting cycle

SDG Target 12.8: By 2030, ensure that people everywhere have the relevant information and awareness for sustainable development and lifestyles in harmony with nature

(8)

3

Both of the aforementioned targets are part of SDG 12: Responsible Consumption and

Production and can be addressed by examining the CSR communication practices of apparel

brands. Finally, the insights in this thesis can be used to improve industry practice, thereby contributing to the field of textile management.

1.3 Purpose

As highlighted by White, Neilsen & Valentini “CSR communication in research about the apparel industry is understudied” (2017, p. 387); an opinion I share after reviewing the available literature on the topic. The study of CSR communication is important for all company stakeholders (Du, Bhattacharya and Sen 2010) as effective communication should allow for interested parties to accurately assess the CSR performance of a company relative to its peers. As such, the purpose of this thesis is to expand the literature on CSR communication by investigating how apparel brands use different corporate channels to communicate CSR to stakeholders.

1.4 Aim

The aim of this thesis is to compare how apparel brands use different communication channels (both alone and in combination) to communicate CSR to stakeholders. To fulfil the purpose, this thesis will investigate which channels are used for the communication of CSR, which CSR topics apparel brands are engaging with, and how consistently CSR communication is approached. Furthermore, the ways in which brands communicate via each channel will be compared and contrasted, looking for strategic differences in approach.

To guide the research process, the following research question has been formulated:

How are apparel brands using corporate communication channels to demonstrate engagement with CSR?

1.5 Scope

1.5.1 Unit of Analysis

(9)

4

websites will be investigated. The web has become the foremost channel for CSR communication (Fulton and Lee 2013) owing to the fact that corporate websites are a fast and economical way to publish dynamic information, can be updated as often as desired, and targeted towards specific stakeholder groups (Da Giau et al. 2016). Finally, Facebook posts2

will serve as a further unit of analysis. Facebook has been selected for analysis as it is a prominent social media channel and data can be automatically harvested from public accounts, facilitating the research process. Social media has been largely overlooked in the study of CSR communication in the apparel industry; however, Kent and Taylor (2016) argue that social media, used properly, has enormous potential for the communication of CSR. By cross-comparing social media with more tradition channels of CSR communication, new insights into the field of CSR communication are expected. The aforementioned channels of corporate CSR communication will be compared using ethnographic content analysis; the process will be discussed at length in the methodology chapter.

1.5.2 Delimitations

The research will be limited to three channels of corporate communication and through cluster sampling will target companies who are likely to be engaging in CSR to ensure a rich source of content for analysis. Furthermore, the research will be limited to companies who are either headquartered in a Western country, or have significant sales in Western markets as research suggests that there are significant differences in CSR practice and disclosure between developed, and developing economies (Hishan and Ramakrishnan 2015; Islam and Deegan 2010; Kamal and Deegan 2013; Nurhayati et al. 2016; Ramakrishnan, Hishan and Kanjanapathy 2016). The research will also be limited to business to consumer brands (B2C) as well as companies whose full range of communications are available in the English language. Finally, the study will be cross-sectional as opposed to longitudinal; as such, the changes in communication methods over time will be outside the scope of this particular study.

2 CSR and Communication

As the application of CSR has broadened over time, the definition has resultantly become more complex. The origin of CSR can be traced to a philosophical debate on whether businesses are obligated to contribute to the betterment of society (Woo and Jin 2015). While CSR is by no means a new concept, with origins as early as the 1930s, debates on the subject increased in the 1950s. In general, the concept of CSR opposes the view that a company’s sole responsibility is to their shareholders; however, the bounds of CSR have been hotly debated, resulting in significant ambiguity around the term (Carroll 1979). In response, Carroll published what

2 A Facebook post is imagery and/or text posted by a Facebook user that appears on the timeline of followers

(10)

5

would become the most well-known definition of CSR stating: “The social responsibility of business encompasses the economic, legal, ethical, and discretionary expectations that society has of organizations at a given point in time” (1979, p. 500); highlighting both the scope of CSR and the propensity for practice to change alongside societal expectations. CSR has become increasingly relevant in recent decades as a method for companies to address the inherent challenges of globalised supply chains (Woo and Jin 2015).

In the apparel industry, the practice of disclosing CSR became more prominent in the mid-1990s as a result of negative media attention directed at brands (Islam and Deegan 2010). It was at this time that western buyers began to place pressure on their overseas suppliers, to deliver not only product, but also additional compliance measures that had not been stipulated previously (Islam and Deegan 2008). Simultaneously, companies began communicating with stakeholders about their efforts to prevent or remedy the violations in their supply chains (Islam and Deegan 2010). Woo and Jin (2015) view CSR communication as an effort to create a favourable corporate image, especially amongst stakeholders who are sensitive to social issues. This view is supported by Da Giau et al. (2016, p. 77) who define CSR communication as:

... the set of strategies and subsequent practices that have a relevant role in disseminating information about an organization’s environmental and social

behaviours to influence, support and enhance the company’s corporate image in the eyes of its stakeholders and final customers and to indirectly promote its products.

In a modern context, CSR retains a level of ambiguity with businesses and stakeholders developing their own understanding of the concept in the absence of a universally accepted definition. As such, CSR in the context of this thesis should be understood in the broadest possible sense, encompassing all information that is considered material to businesses or stakeholders in the demonstration of corporate responsibility.

2.1 Likelihood of Engaging in CSR

(11)

6

emphasising that they are the most significant influencing factor in engaging businesses in social and environmental reporting in Bangladesh.

Despite a slower initial uptake, the concept of CSR is beginning to permeate developing nations. Much of the literature in this area focuses on textile and apparel manufacturing firms, many of whom service large, western retailers; as such, the CSR requirements of these companies become prerequisites for trading between nations (Islam and Deegan 2010; Wong and Dhanesh 2017). Conforming to the CSR requirements of multinational brands can be problematic as manufacturers may not have the available resources to implement the changes requested by these large, powerful retailers (Perry and Towers 2009), as such, risking the continuation of the trading relationship (Mann et al. 2014). Despite the issues that may exist in implementing CSR practices, Perry and Towers (2009) see the purpose of CSR as a method of transforming the supply chain into a demand chain where upstream parties become aware of the requirements of the end consumer, evidenced by consumer demand for ethical products resulting in additional CSR requirements for supply chain actors.

Aside from addressing regional differences in approach to CSR, much of the literature has focused on which factors affect the likelihood of a given business to report CSR. Firm size is the factor most commonly reported to affect CSR disclosure with a consensus that larger brands are more likely to report on CSR than their smaller counterparts (Birkey et al. 2018; Colucci, Tuan and Visentin 2020; Islam and Deegan 2010; Kamal and Deegan 2013; Nurhayati et al. 2016; Perry and Towers 2009). The suggested reasons for increased reporting by larger firms include that mass-market retailers are more visible (Colucci, Tuan and Visentin 2020) and as such, draw more scrutiny (Islam and Deegan 2010). Furthermore, Birkey et al. (2018) points out that larger firms are more likely to have diverse and high-risk supply chains and as such, companies are motivated to participate in CSR communication to lower their risk. Large companies are also seen as responsible for setting the CSR agenda (Perry and Towers 2009), which companies of all sizes are subsequently expected to comply with; this can cause complications for smaller brands who struggle to comply with CSR requirements due to inferior financial and human resources (Perry and Towers 2009).

(12)

7

way firms interact with external actors (Colucci, Tuan and Visentin 2020). The suggested reasons for the marked reduction in CSR reporting in small firms include that smaller firms receive proportionally much less scrutiny than their larger competitors, and have thus far managed to fly under the radar (Colucci, Tuan and Visentin 2020, p. 3). Additionally, the relative costs of implementing CSR are disproportionately higher for smaller businesses (Colucci, Tuan and Visentin 2020), who may find themselves with insufficient resources to engage in CSR in any significant way.

2.2 CSR and Power Dynamics

Another key theme that emerged from the literature is the power imbalance that must be considered when the implications of CSR on stakeholders is concerned. Perry and Towers (2009, p. 383) emphasise that the apparel industry is typically “characterised by powerful global retail buyers and relatively weak suppliers” forming “flexible subcontracting relationships” with suppliers in “lower labour cost countries” resulting in an “unequal power distribution”. Luque and Herrero-García (2019, p. 1286) are particularly critical of corporations and their significant power to sway CSR legislation in their favour, stating:

[Transnational companies] have been able to develop made‐to‐order labour

regulations and social and human rights arrangements, leading to a corporate ethic which is far removed from that which most ordinary people would recognise.

Conversely, Fulton and Lee (2013) discuss the inherent limitations encountered by SME brands working with larger suppliers, questioning the power that smaller brands have over their supply chains. The authors recommend that when deciding what CSR information to communicate, an organisation should...

consider how much influence or control they have over their entities upstream (e.g. supply chain) as well as their entities downstream (e.g. distributors and consumers), as this is directly relevant to [their] ability to be sustainable (2013, p. 354).

Debates on power differentials are common in the CSR literature, with much discussion centred around the underlying motivations for companies to engage in CSR (Colucci, Tuan and Visentin 2020). Ban (2019, p. 2) sums up the central debate, stating that “the two approaches depart on whether to treat managerial CSR discourse as steps toward improvement, or a smokescreen of

corporate agenda.”

2.3 CSR Walk and CSR Talk

CSR walk and CSR talk (Wickert, Scherer and Spence 2016) are considered to be the two,

(13)

8

must be present, and balanced, to achieve desired legitimacy gains (Birkey et al. 2018) and as a result, competitive advantage (Mann et al. 2014).

CSR talk encompasses the external communication tools instrumentally used by a company to inform stakeholders about environmental and social initiatives such as CSR reports, corporate websites, advertising and product labelling ...

CSR walk, differently, encompasses substantive actions within the firm, such as changing methods of production to reduce environmental impacts or changing labor relationships both within the firm and across the firm’s value chain

(Colucci, Tuan and Visentin 2020, p. 2)

While academics consider it essential for companies to engage simultaneously in CSR walk and CSR talk, they report that the two are frequently misaligned (Colucci, Tuan and Visentin 2020). Hishan and Ramakrishnan (2015, p. 53) consider communication the missing link in the practice of CSR, while Garcia-Torres, Rey-Garcia and Albareda-Vivo (2017, p. 4) point out that the talk-action gap hampers sustainability and causes CSR to be perceived as merely

performative. Further studies question the usefulness of CSR communication, with Mann et al.

(2014) noting a lack of disclosure and specificity in what brands communicate, while Wu, Shao and Chen (2018) conclude that the current methods employed in CSR reporting undermine the consistency and efficacy of the communication channel.

(14)

9

2.4 CSR and Marketing

Mass-market brands often pursue CSR communication as an attempt to create positive brand associations and compensate for other irresponsible business practices (Colucci, Tuan and Visentin 2020); however, it must be noted that standalone CSR communication is a risky practice as it is purely marketing (Wong and Dhanesh 2017). An approach to CSR that is primarily communication-focused is discouraged as it is disingenuous and invites negative stakeholder responses. Childs, Woo and Kim (2019) warn that consumers are capable of judging the source of information when inferring credibility, and if CSR communication is judged to be insincere, it can be detrimental to the company's image. Luque and Herrero-García (2019, p. 1287) agree, stating:

The processes of CSR encourage certain negative approaches to marketing, such as the washing of the corporate image, the masking of any information that is negative for the company, and the exaggerating of positive aspects of business activity at the expense of any information that could harm the corporate image and dividend.

In support, Utgård (2018) cautions that stakeholders are vigilant for signs of CSR-washing3,

and rational observers will not interpret general statements and other half-hearted attempts at CSR communication as credible. This view is further supported by Pope and Wæraas (2016) who find that effective CSR-washing is both exceedingly difficult to accomplish and resultantly, far less common than presumed by consumers, media, NGOs and academics.

In addition to disingenuous marketing, Fulton and Lee (2013, p. 354) discover that existing marketing strategies for environmentally-friendly products are considered by consumers to be neither relevant nor engaging, indicating that new ways of communicating about CSR are necessary to reengage consumers with sustainability messaging.

2.5 Mandatory CSR Disclosure

There are few examples of mandatory CSR disclosure in the apparel industry; however, the California Transparency in Supply Chains Act of 2010 (CTSCA) serves as a relevant industry example. The purpose of the act is to target human trafficking and modern slavery and is designed to enable consumers to avoid inadvertently supporting unethical labour practices (Birkey et al. 2018). In their longitudinal study of CSR disclosures on company websites, Mann et al. (2014) noted a significant increase in disclosure in the year succeeding the CTSCA; however, the authors noted that not all companies featured a prominent link to the content as prescribed in the legislation. After analysing the effects of the CTSCA on CSR reporting,

3 CSR-washing is defined as “the successful use of a false CSR claim to improve a company’s competitive

(15)

10

Birkey et al. (2018) conclude that government mandates can increase normativity of CSR disclosure and can be an effective measure to increase CSR engagement. Similarly, Hishan and Ramakrishnan (2015) recognise the increased CSR uptake in India preceded by the New

Companies Act 2013, which made CSR mandatory. Finally, Alam and Akter (2019) recognising

the influence of legislation, recommend the implementation of governmental policies to increase CSR activities in Bangladesh.

2.6 Voluntary CSR Disclosure

There are many suggestions put forward as to why brands choose to engage in CSR communication. In some cases, it is mandatory, as in the CTSCA (Birkey et al. 2018; Mann et al. 2014), but in most cases CSR communication is voluntary, causing much speculation as to why brands choose to publicly disclose what was previously treated as privileged information. A popular explanation for CSR disclosure is that of legitimacy (Alam and Akter 2019; Birkey et al. 2018; Islam and Deegan 2010; Kamal and Deegan 2013; Nurhayati et al. 2016). Disclosure of CSR information is believed to make a company more credible and to build legitimacy in the eyes of company stakeholders (Wong and Dhanesh 2017). The incentives for CSR disclosure differ across the supply chain with Hishan and Ramakrishnan (2015) claiming that manufacturers in India disclose CSR information to their business-to-business (B2B) customers to signal that they are trustworthy. Childs, Woo and Kim (2019) add that the practice of CSR can build corporate credibility and that consumers tend to judge the advertising claims of credible brands as more trustworthy.

(16)

11

There is also the suggestion that some companies participate in CSR because they perceive it to present an opportunity. Perry and Towers (2009) highlight that certain supply chain improvements implemented under CSR can increase the profitability of a firm, providing a business case for CSR (Du, Bhattacharya and Sen 2010; Perry and Towers 2009). Furthermore, organisations are aware that a good reputation can have a strong positive impact on consumer behaviour (Luque and Herrero-García 2019); as such, Da Giau et al. (2016) recommend that brands should use their web channel to demonstrate their commitment to CSR, as making this data easily available can influence consumer purchase intent. The rise of the conscious consumer also presents opportunities for brands to capture the business of this emerging consumer segment. Changing customer demographics are seeing millennials who value social and environmental sustainability accumulate more purchasing power (Wong and Dhanesh 2017); resultantly, there is a rising demand for ethical and sustainable products (Perry and Towers 2009). Fulton and Lee (2013, p. 354) note the difficulties encountered by this consumer type, stating:

Many consumers want to make environmentally friendly choices, but they cannot complete all of the background research necessary to determine a product’s sustainability on their own. Most consumers will not know about companies’ green initiatives unless the retailer informs them of these actions. Therefore, it is important for retailers to allow consumers to make an informed decision about a sustainable purchase that is possible through their company.

Finally, Perry and Towers (2009, pp. 383, 384) note that businesses may act in enlightened

self-interest, with priorities other than profit-maximisation; this is the case when business owners

perceive a moral obligation to act responsibly and use the business as a tool, from a purely altruistic desire, to do good (Perry and Towers 2009).

2.7 CSR Communication Channels

(17)

12

(2014) performed a comprehensive analysis of channels by also analysing company blogs and searching individual product information on brand websites.

It makes sense for companies to employ a multi-channel communication approach when trying to address multiple stakeholder groups (Du, Bhattacharya and Sen 2010). Hishan and Ramakrishnan (2015) suggest that an appropriate CSR communication strategy will consider which stakeholders are engaging with the brand via each channel and tailor their message accordingly. Furthermore, Colucci, Tuan and Visentin (2020) add that the internet is a democratic platform, and as such, smaller brands can communicate on their website as effectively as their larger competitors. Despite all of this, there has been no literature produced investigating how each channel is used differently by apparel companies to communicate CSR.

2.7.1 CSR Communication and Annual Reports

Formalised sustainability reporting is “expected to provide comparability, accuracy, clarity and reliability to stakeholders and readers in a timely manner” (Fulton and Lee 2013, p. 354). As such, CSR reporting is considered to be the most comprehensive and important accountability tool employed by businesses (Garcia-Torres, Rey-Garcia and Albareda-Vivo 2017). Within the field of CSR reporting, the Global Reporting Initiative (GRI) is the most common voluntary framework and is used globally (Colucci, Tuan and Visentin 2020; Garcia-Torres, Rey-Garcia and Albareda-Vivo 2017; Wu, Shao and Chen 2018). Proponents of sustainability reporting acknowledge the scepticism related to the talk-action gap but subscribe to the belief that talk can precede action (Garcia-Torres, Rey-Garcia and Albareda-Vivo 2017). Colucci, Tuan and Visentin (2020, p. 2) add that "voluntary adopt[ion] the GRI framework demonstrate[s] to external stakeholders [adherence] to social norms and expectations", having a legitimising effect on brand perception. Companies using the GRI framework are given free rein to decide which GRI issues covered by the standard are material to their business and should be reported on (Wu, Shao and Chen 2018). However, while companies are expected to balance favourable and unfavourable information, reporting will frequently be used to highlight a firm's goals and priorities (Fulton and Lee 2013). The fact that companies get to choose what they disclose is seen as problematic by some (Garcia-Torres, Rey-Garcia and Albareda-Vivo 2017) as the voluntary nature of what to include can make CSR reports misleading by omission.

2.7.2 CSR Communications on the Company Website

(18)

13

updated as often as desired, and targeted towards specific stakeholder groups (Da Giau et al. 2016). Websites can be employed effectively by both small and large brands (Colucci, Tuan and Visentin 2020) and allow for companies to make information readily available to stakeholders (Da Giau et al. 2016). Although the web has made CSR information easier to distribute, there are still criticisms regarding how it is used. Mann et al. (2014) note that online information is still often difficult to access and interpret. Wong and Dhanesh (2017) find that retailers are using vastly different methods to communicate their CSR ranging from highly emotive, yet, uninformative statements to highly quantitative data that may be difficult for a layperson to interpret. Finally, White, Neilsen and Valentini note that despite the benefits of publishing CSR information online, it is flawed to "assume that all important stakeholders are actually reading websites" (2017, p. 388) to gain CSR insight.

2.7.3 CSR Communication on Social Media

The use of social media as a tool for apparel brands to communicate with stakeholders has proliferated in the past couple of decades. While it is thought that many brands utilise social media as a channel to communicate aspects of CSR, little research has explored the topic. Reilly and Larya (2018) recognising the increased use of the internet for CSR communication, differentiate between formal CSR communication (CSR reports and related documents) and

informal CSR communication (social media). In investigating the use of Twitter and Facebook

to communicate CSR, they found “no observable patterns differentiating luxury brands from lower end brands in their CSR social media engagement” (Reilly and Larya 2018, p. 628). By its very nature, social media has a significantly higher potential for stakeholder interaction; however, brands are typically reluctant to engage in online dialogue with stakeholders about CSR issues (Etter 2013). Despite the typical lack of engagement, Kent and Taylor (2016) argue that social media, used properly, has enormous potential in reaping maximum benefits from CSR communication. Du, Bhattacharya and Sen (2010) agree, adding social media is primarily the domain of consumers, a powerful stakeholder group, and while corporate-controlled CSR messaging can be beneficial, successfully engaging consumers to act as CSR advocates inherently result in higher reach and credibility. Despite the dialogic potential of the medium, Kent and Taylor (2016) find that companies instead use social media as another one-way communication channel due to fear of negative publicity and receiving public CSR-related critique. While Etter (2013, p. 608) highlights that many brands take a cautious communication

approach to the discussion of CSR on social media, the findings of his Twitter study indicate

(19)

14

2.8 Corporate Social (Ir)responsibility

There is significant scepticism amongst academics about the validity of CSR, as the practice frequently fails to recognise or address many of the underlying issues central to the harm caused by the fashion industry. White, Nielsen and Valentini (2017, p. 390) state:

The reactive approach of the industry to adopt CSR from a more comprehensive perspective may be explained by the tacit hesitation of businesses, politicians, and consumers to address the urge to consume, which can be a barrier to promoting more sustainable living patterns in everyday life and adjusting the world economy in this direction.

Furthermore, the authors add that as there is minimal political pressure placed on the apparel industry, as the “willingness to restrict over-consumption and encourage more sustainable practices is not necessarily a priority” (White, Nielsen and Valentini 2017, p. 383).

A further issue is the outsourcing of responsibility; the predominant business model employed by large businesses who manufacture their products in developing nations (Da Giau et al. 2016). In exporting and outsourcing production, businesses also seek to outsource their liability for any element within their supply chain, “enjoy[ing] the profits of this situation, while externalizing risks and responsibilities” (Luque and Herrero-García 2019, p. 1286). Ban (2019) explores the discourse around the organisation of labour in apparel supply chains and finds that discourse primarily focuses on the issue of the massive power differential between multinational retailers and their contractors, and additionally, between management and workers in these contracted firms. The authors go on to state that "labor malpractice is not merely about suppliers' compliance to codes of ethics, but often the result of the influence of powerful buyers" (Ban 2019, p. 6). Scepticism of CSR appears in the discourse as businesses manage to fulfil voluntary CSR standards with a code of conduct while leaving in place the exploitative model of supply chain production (Ban 2019). Luque and Herrero-García (2019, p. 1288) go further, stating:

(20)

15

The term corporate social irresponsibility (CSiR) challenges the notion of CSR, suggesting that CSiR as a concept, more accurately reflects the true conduct of many companies who use CSR communication strategies; seeking to appear respectable while continuing to participate in exploitative practices (Luque and Herrero-García 2019). The CSiR perspective emphasises that CSR is impossible when elective and does not provide an adequate solution to the systematic exploitation occurring in modern supply chains (Luque and Herrero-García 2019).

2.9 Framing CSR

Ban (2019) and Wong and Dhanesh (2017) take a different approach to the investigation of CSR communication by using framing theory to investigate the rhetorical devices employed by brands. Framing is suited to investigating CSR as brands often employ purposeful and controlled language in addressing CSR topics (Wong and Dhanesh 2017) which can be dissected by academic researchers.

In their analysis of luxury brand websites, Wong and Dhanesh (2017) find that most companies frame their participation in CSR as discretionary and driven by altruistic motives. As such, brands approach the topic of their CSR participation as being mostly about the betterment of society, their willingness to help stakeholders solve social issues, and importantly, frame the role of the company in addressing these issues as being purely voluntary (Wong and Dhanesh 2017). Similarly, Ban (2019) explores the way that companies talk about the workers in their supply chains, acknowledging the large population of disenfranchised, marginalised supply chain workers, without reflecting on the role that they have played in their marginalisation. Instead, companies exhibit strategies to frame the issue as a result of neglect from the governmental and social structure of developing nations, taking the position that “if Third World workers are treated badly, it is because they belong to socially disenfranchised groups in the first place” (Ban 2019, p. 10). Furthermore, western brands use discourse to position suppliers as the perpetrators of incompliance with brands serving as "watchdogs for instances of incompliance" (Ban 2019, p. 9).

(21)

16

Findings from Wong and Dhanesh suggest that brands curate their CSR response to fulfil the higher-order needs of their stakeholder groups including: “self-definition (the need to know oneself), self-enhancement (the need to feel good about oneself), and self-distinctiveness (the need to feel special)” (2017, p. 424). As such, many of the frames employed by brands seek to align brand identity with CSR cause and communicate in a feel-good manner. Strategies such as foregrounding the brand, and their involvement in remedying the social issue, ahead of the issue itself (Du, Bhattacharya and Sen 2010; Wong and Dhanesh 2017) are employed to increase the credibility of the claim and minimise the risk of consumers sensing ulterior motives (Childs, Woo and Kim 2019). Furthermore, brands typically chose to use their websites to frame CSR issues abstractly, using holistic, qualitative and emotive methods to communicate their CSR activities (Wong and Dhanesh 2017), in effect, creating a feel-good overview of brand CSR achievement.

2.10 Summary of CSR Literature

In total, 31 academic articles have been reviewed to establish the state of current research on CSR communication. In the literature, a variety of different techniques have been used to conduct primary research. Most commonly content analysis was employed (Fulton and Lee 2013; Garcia-Torres, Rey-Garcia and Albareda-Vivo 2017; Kamal and Deegan 2013; Kozlowski, Searcy and Bardecki 2014; Nurhayati et al. 2016), followed by a case study methodology (Childs, Woo and Kim 2019; Da Giau et al. 2016; Woo and Jin 2015), literature review (Perry and Towers 2009; White, Nielsen and Valentini 2017), frame analysis (Ban 2019; Wong and Dhanesh 2017) and finally Alam and Akter (2019) employed a self-reporting questionnaire. In addition to providing information on the state of CSR communications, the literature served as inspiration for an appropriate theoretical and methodological framework. In the subsequent chapter, the theoretical framework to be used for the primary research component of this thesis will be explained in detail.

3 Theoretical Framework

3.1 Legitimacy Theory

(22)

17

right to continue operations through [suspension of] support and allocation of resources" (Alam and Akter 2019, p. 2658) essential to their survival.

Legitimacy theory recognises that businesses attempt to align their business systems with social systems through the process of legitimation. It does not, however, describe how "congruency between the two systems can be achieved" (Chen and Roberts 2010, pp. 654-655). As such, there are two significant branches of legitimacy theory that go further in their attempt to reconcile this, namely institutional legitimacy and organisational legitimacy (Chen and Roberts 2010).

3.1.1 Institutional Legitimacy

Institutional legitimacy investigates socially accepted institutional structures and activities and then uses these as a benchmark with which to compare legitimacy seeking organisations (Chen and Roberts 2010). By organisations following the same path to seek legitimacy, this in turn “reinforces the legitimacy of the already institutionalized system – a capitalistic society in which our analysis of the social and organizational relationship is situated” (Chen and Roberts 2010, p. 655).

3.1.2 Organisational Legitimacy

Much of the literature on CSR disclosure borrows from organisational legitimacy theory (Islam and Deegan 2010; Islam and Deegan 2008; Nurhayati et al. 2016). Organisational legitimacy is concerned with the strategic measures that organisations adopt in seeking legitimacy, specifically the suggestive symbols employed in order to elicit support and social acceptance (Chen and Roberts 2010). Conversely, when potential or actual disparity exists between societal and firm values, legitimacy is threatened (Nurhayati et al. 2016). Even though a firm may be acting in line with societal expectation, it is still considered vitally important to demonstrate this conformance. As such, CSR communication and formalised CSR reporting play a significant part in conveying adherence to social expectations (Nurhayati et al. 2016).

3.2 Stakeholder Theory

Stakeholder theory recognises that firms have responsibilities that extend beyond their shareholders. Instead, stakeholder theory is concerned with the relationships between organisations and the stakeholders that make up their greater business environment (Chen and Roberts 2010). As stated by Perry and Towers (2009, p. 381):

Stakeholder theory assumes acceptance of embeddedness, which describes the idea that business does not exist in a vacuum but is embedded in society, and

(23)

18

Furthermore, stakeholder theory recognises stakeholders aspirations and concerns can be diverse and conflicting and that different stakeholder groups exhibit different levels of ability to exert pressure on an organisation (Chen and Roberts 2010).

Stakeholder theory is useful in the analysis of CSR discourse as it suggests that companies need to understand the concerns and aspirations of their stakeholders and that corporate communication strategy should reflect the informational requirements of stakeholders (Nurhayati et al. 2016; White, Nielsen and Valentini 2017). Perry and Towers (2009) suggest a link between the prevalence of stakeholder theory and greater awareness of the adverse effects of globalism and the resultant environmental and social impacts. They assert that CSR is built on the similar notion, "that firms should assume social and environmental responsibilities to stakeholders ... in addition to their financial responsibilities to shareholders" (Perry and Towers 2009, p. 381).

3.3 Institutional Theory

Institutional theory is similar to legitimacy theory but is more concerned with the stability and survival of organisations, concentrating heavily on the relationship between companies and their operating environment (Chen and Roberts 2010). Institutional theory provides insight into how businesses respond to societal rules, values (Woo and Jin 2015) and respond to various forms of pressure. The concept of institutional isomorphism is particularly relevant to this thesis and "suggests that mimetic, normative, and coercive pressures may lead indicator disclosure practices in this sector to converge over time" (Kozlowski, Searcy and Bardecki 2014, p. 391). Coercive pressure arises from governmental regulations; while mimetic pressure is at work when new organisations embrace the existing systems of used by other institutions in their industry; finally, normative pressure causes organisations to intuitively recognise and subsequently follow societal norms (Chen and Roberts 2010).

Applied to CSR disclosure, institutional theory suggests that different forms of pressure may lead to disclosure practices within a given industry converging over time (Kozlowski, Searcy and Bardecki 2014). However, Woo and Jin (2015) add that factors such as country-of-origin and retail markets can affect the institutional environment; thus, different market environments are likely to affect companies’ different CSR approaches. Finally, Chen and Roberts (2010, p. 653) describe conformity to “established institutional patterns [as a] pathway to legitimacy”, serving to enhance a corporations survival prospects, and stability.

3.4 Signalling Theory

(24)

19

seller and the buyer, “starts from the situation of informational asymmetry, where the seller knows more about his or her quality than the potential buyer” (Utgård 2018, pp. 385-386). When applied to CSR, it follows that:

... a firm will signal CSR performance when the reward from sending such signal is higher than the cost, and when it is less costly to send such signals for good

performers than for bad performers. [In such a situation] it may be profitable [for firms] to signal CSR quality (Utgård 2018, p. 386).

CSR signals (or lack thereof) are often used as cognitive shortcuts and are employed by external parties to distinguish ethical businesses from unethical ones (Utgård 2018; Zerbini 2017). Hishan and Ramakrishnan (2015) consider corporate transparency important in signalling trustworthiness to stakeholders; while Alam and Akter (2019) add that disclosure is necessary to satisfy investors and to give shareholders confidence that the business is “less likely to be encumbered by regulatory oversight” (Hishan and Ramakrishnan 2015, p. 53).

The underlying rationale is that CSR disclosure has implicit value beyond reducing information asymmetry (Zerbini 2017) and can result in payoff for companies in the form of reputational improvement, increased legitimacy and financial performance. However, one caveat is that for many signals to work, they need to be costly to be perceived as credible (Utgård 2018). While Zerbini (2017) recognises a wide variety of CSR signals including codes of practice, dedicated CSR staff, training and incentive programs, corporate disclosures, trustmarks (e.g. certifications and memberships) and performance ratings; Utgård (2018) cautions that CSR-washing is of concern to stakeholders and as such, rational observers will not interpret general statements and other half-hearted attempts at CSR communication as credible signals.

3.5 CSR Communication Framework

The below Framework for CSR Communication was established by Du, Bhattacharya and Sen to improve the implementation and understanding of corporate communications. The left side of the framework is concerned with what to communicate (message content), and where to communicate (message channel); while the centre highlights the internal and external contingency factors that can amplify or dampen the effects of a CSR message; finally, the right-hand side lists the diverse outcomes that can result from successful communication of CSR.

(25)

20

Figure 1: Framework for CSR Communication Source: (Du, Bhattacharya and Sen 2010, p. 11)

3.6 Synthesis of Theories

Prior studies on CSR have applied a variety of theories to the study of CSR communication. These include stakeholder theory (Nurhayati et al. 2016; Perry and Towers 2009; White, Nielsen and Valentini 2017), institutional theory (Kozlowski, Searcy and Bardecki 2014; Woo and Jin 2015) and signalling theory (Alam and Akter 2019; Hishan and Ramakrishnan 2015), attribution theory (Childs, Woo and Kim 2019), agency theory (Nurhayati et al. 2016), and media agenda setting theory (Islam and Deegan 2010). However, legitimacy theory is by far the most common theory applied to CSR in the reviewed academic literature (Alam and Akter 2019; Birkey et al. 2018; Islam and Deegan 2010; Kamal and Deegan 2013; Nurhayati et al. 2016).

(26)

21

Figure 2: Relationship Between Theories that Concern Legitimacy Source: (Chen and Roberts 2010, p. 653)

Recognising the multitude of motivations for CSR disclosure, and the inadequacy of a single theory to explain the phenomena, Chen and Roberts (2010, p. 651) state that:

It must be acknowledged that some business entities initiate social activities based on direct interactions with stakeholders, whereas others may also undertake similar activities to manage their societal level of legitimacy.

As such, it is appropriate to apply multiple perspectives when analysing CSR communications. As suggested by Chen and Roberts (2010), legitimacy theory is most appropriate when a researcher is primarily focused on how firms manage their public image and is a suitable framework for evaluating social and environmental disclosures. Furthermore, the authors go on to promote stakeholder theory as the most appropriate framework for investigating voluntary participation in “activities benefiting society or the natural environment without explicit self promotion or publicity” (Chen and Roberts 2010, p. 662).

(27)

22

legitimate and illegitimate simultaneously. Signalling theory adds to the arguments presented in this thesis by recognising CSR communication efforts as signals employed by firms in the pursuit of legitimacy.

4 Methodological Framework

4.1 Approach

The research component of this thesis will consist of a content analysis applied to three CSR communication channels employed by apparel brands; namely, formalised CSR reports, corporate websites and social media communications. A mixed-methods approach will be applied for the research component of this thesis resulting in both quantitative and qualitative evidence being presented in the results section. While some academics view the two approaches as distinct, Krippendorff (2013, p. 22) questions the validity and usefulness of distinguishing between qualitative and quantitative content analysis, stating that “all reading of texts is qualitative, even when certain characteristics of a text are later converted into numbers”. This thesis takes a similar view, recognising the usefulness of both quantification and qualification in exploring the concepts and themes relevant to CSR communications.

4.1.1 Ethnographic Content Analysis

Krippendorff (2013, p. 24) defines content analysis as “a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use". The methodology has significant benefits, including that when well-executed, it is transparent, flexible, unobtrusive, and can deal with unstructured data (Bryman and Bell 2015). With the advent and subsequent popularisation of computer-aided qualitative data analysis software (CAQDAS), there is also the ability to deal with large quantities of textual data efficiently. Content analysis has been widely applied in the field of CSR communications research (Mann et al. 2014) and has been effectively applied to analyse CSR communications across various channels. Krippendorff (2013, p. 36) emphasises that most content analyses utilise data that was not created for use in research; instead, they are intended to "be read, interpreted, and understood by people other than the analysts", thus reducing the potential for response bias.

(28)

23

ECA follows a recursive and reflexive movement between concept development– sampling–data collection–data coding–data analysis– interpretation. The aim is to be systematic and analytic but not rigid.

ECA has been selected as a research approach, over more conventional content analysis methods as it encourages discovery, allows for the integration of narrative data and allows concepts to emerge during the research (Altheide and Schneider 2017). To fulfil the aim and purpose of this thesis, a more flexible, iterative approach was desirable, as was the ability to illustrate themes emerging in the discourse through the use of quotes (Bryman and Bell 2015) and appropriate, qualitative, visual data.

4.2 Sampling

As a systematic account of apparel companies does not exist (Colucci, Tuan and Visentin 2020), it is instead necessary to use cluster sampling "which is the process of selecting content units from clusters, or groups, of content" (Riffe, Lacy and Fico 1998). As the research objective is to investigate the communication patterns of apparel brands who are engaging in CSR communication, the sampling frame was created by consolidating lists of apparel retailers who participate in various, reputable CSR-related activities. To create the sampling frame, the names of companies who participated in the Fashion Transparency Index Survey 2020 (n= 237), are members of the Textile Exchange (n= 442), the Fairwear Association (n= 110), listed in The Fashion United Top 100 Index (n= 100), or are signatories of the Uzbek Cotton Pledge (n=

144), were consolidated; after which, a random number generator was used to select brands for

participation in the study (n=30).

Each selected brand was assessed to ensure they met the inclusion criteria by communicating in English and being a B2C brand. If a group of companies was selected, one brand from the group portfolio was randomly selected for inclusion in the study. According to these criteria, any unsuitable brands were discarded, and a new brand randomly selected in replacement. This process was continued iteratively until a full sample of 30 brands was achieved. By sampling in this manner, the intention was to achieve a sample of companies who are representative of different CSR priorities and potentially, different CSR-related communication strategies. This process aims to reduce selection bias (Colucci, Tuan and Visentin 2020) and achieve diversity in business types.

(29)

24

exchange and sampled the largest five of the 30 apparel companies listed for their research. Unable to find an appropriate list to sample from, Fulton and Lee (2013) used a Google keyword search to generate a list of appropriate companies to study. Colucci, Tuan and Visentin (2020) take by far the most involved and ambitious approach, consolidating The BoF500 list, SAC members (now 72 brands), the clean clothes campaign, Fur Free Retailer, Digital IQ Index Fashion, Greenpeace Toxic Threads report, then assessing the presence or absence of various forms of CSR communication.

4.2.1 Sample Size and Description

Determining an appropriate sample size presents a challenge as there is no universally accepted criteria for sample size (Neuendorf 2017). Bengtsson (2016) posits that a sample size between one and 30 subjects is adequate for most qualitative studies, while Krippendorff (2013) specialising in content analysis, emphasises that sample size should instead be determined by informational requirements, ensuring that the research question can be answered with sufficient confidence.

Table 1provides descriptive statistics of the brands included in the study. The sample is diverse, with 11 countries represented, as well as businesses of differing sizes, and a mix of publicly traded and privately held organisations.

Table 1: Descriptive Statistics of Sampled Companies

Brand Headquarters Public/Private Size*

Acne Studios Sweden Private Large

American Eagle Outfitters USA Public MNC

ASOS UK Public Large

Boohoo UK Public Large

Burberry UK Public Large

C&A Belgium Private MNC

Canada Goose Canada Public Large

Carhartt USA Private Large

Columbia Sportswear USA Public MNC

Cotton On Australia Private MNC

Esprit Holdings Hong Kong Public MNC

Faherty Brand USA Private Medium

Filippa K Sweden Private Large

From the Road USA Private Micro

Glassons New Zealand Public Large

Kathmandu New Zealand Public Large

Levi Strauss USA Public MNC

(30)

25

Mini Rodini Sweden Private Medium

Moncler Italy Public Large

Next UK Public MNC

Nike USA Public MNC

Odd Molly Sweden Public Medium

Pact Apparel USA Private Medium

Rowlinson Knitwear UK Private Small

Suistudio The Netherlands Private Small

Tact & Stone USA Private Micro

United Colors of Benetton Italy Public MNC

Uterqüe Spain Public Large

Vivienne Westwood UK Private Large

MNC = Multi-national Corporation *OECD of definition business size applied.

4.2.2 Sampling Limitations

The choice of cluster sampling introduces additional sampling error compared to probability methods (Riffe, Lacy and Fico 1998); however, is an appropriate method when “analysts cannot enumerate all units of analysis but find lists of larger groups of such units, or clusters” (Krippendorff 2013, p. 117). By utilising clustering, the variance may be exaggerated (Krippendorff 2013) or conversely, units may display reduced variance as the sampling process inadvertently groups units that inherently similar in nature (Riffe, Lacy and Fico 1998). Although cluster sampling is an appropriate method for the type of research being conducted in this thesis, it is important to understand the limitations of the methodology as although participating brands are selected randomly, the sample may not be representative (Riffe, Lacy and Fico 1998). Krippendorff (2013, p. 118) adds that generalisation is not the most important issue in content analysis but instead that texts should be chosen based on their “likely ability to contribute to decisions on rather specific research questions”.

4.2.3 Unit of Analysis

Several units of analyses will be collected and analysed to compare the CSR communication channels of apparel brands. Firstly, the most recently published annual CSR report will be examined for each company; in the absence of a dedicated CSR report, the annual report will be used in its place. The second unit of analysis will be the CSR section of the company website, and finally, Facebook posts for the 12 months prior to the date of data collection will be collected and analysed.

(31)

26

Birkey et al. 2018; Colucci, Tuan and Visentin 2020; Da Giau et al. 2016; Fulton and Lee 2013; Kozlowski, Searcy and Bardecki 2014; Mann et al. 2014; Ramakrishnan, Hishan and Kanjanapathy 2016; Wong and Dhanesh 2017; Woo and Jin 2015) as their unit of analysis, with less focus on CSR discourse on social media or other, less formal, channels.

4.3 Data Collection

All data was collected over a three-day period, between 23rd – 26th April 2020. All content was

digital and as such, was downloaded, labelled and stored as PDFs to control for changes in online content that can occur over time (Wong and Dhanesh 2017).

4.3.1 CSR Reports

CSR reports were searched for on company websites, group websites and finally the GRI database. In the event that a CSR report was not discovered, an annual report was downloaded in its place; if no annual report was available, the report was noted as unavailable and thus excluded from the analysis. As different companies publish their reports at different times of the year, the most recently published report was downloaded for each brand in the sample.

4.3.2 CSR Section of the Company Website

Vast differences in the layout of websites were observed, and as such, a strict methodology for data gathering was established and applied to ensure consistency. Data was gathered only from the CSR section of each website, with all webpages in this section converted to PDF and downloaded for analysis. In the event that the CSR section linked to content on different parts of the website, these linked pages were also included as long as they were published on the company domain. Any external links were disregarded. Policy documents were noted when available, but not included in the content analysis.

4.3.3 Facebook Data

A list of Facebook usernames was consolidated for each of the brands and the software Facepager was used to download posts dated 23rd April 2019 – 23rd April 2020. This

information was downloaded as a spreadsheet and converted to a PDF to ensure compatibility with the ATLAS.ti qualitative analysis software.

4.4 Software

(32)

27 4.4.1 Facepager

Facepager is an open-source program designed to facilitate the fetching of publicly available data on the basis of APIs4 and webscraping5 (Jünger and Keyling 2020). Due to the increasing

commercial value of data, companies such as Facebook are making it increasingly difficult to obtain comprehensive access to their raw data, resulting in financial or technical barriers for academics looking to utilise this rich source of data (Batrinca and Treleaven 2014). Facepager works by providing a more user-friendly interface to the Facebook API, providing an easier method of accessing data that would otherwise be out of reach for researchers who lack appropriate financial or programming resources.

4.4.2 ATLAS.ti

Considered one of the major software packages for qualitative data analysis, ATLAS.ti allows for the documentation of “analytic decisions in a transparent, reflexive, rigorous and systematic way” (Paulus and Lester 2016, p. 405). By electing to use CAQDAS, researchers can improve the reliability and time efficiency of their content analysis (Krippendorff 2013). Furthermore, ATLAS.ti can improve academic rigour by facilitating “reliability, reflexivity process, and perhaps most importantly [by] improving credibility, conformability and dependability” (Rambaree 2013, p. 12).

While there are significant benefits to using CAQDAS, it is important to remember during implementation, that while helpful, computer programs are soulless and require proper research design and human creativity (Bengtsson 2016, p. 12) to achieve optimum results. Krippendorff (2013) and Rambaree (2013) agree, adding that computers lack the ability to think. Rather than treating this as a disadvantage, Krippendorff (2013, p. 365) states that:

The advantage of such analytical aids lies in the assurances they can provide that text explorations are systematic, effectively countering the natural tendency of humans to read and recall selectively.

Responding to the reluctance of some researchers to adopt CAQDAS, Paulus and Lester (2016, p. 424) argue that “rather than taking control away from the researcher, ATLAS.ti enables the analyst to engage, through queries and networks, in deeper levels of analysis than is possible by hand”; furthermore, they emphasise that “rather than creating distance, [the inbuilt tools in ATLAS.ti] support a very close reading of the data” (p.424).

4 Application Programming Interfaces (APIs) are the interfaces provided by online services that allow for the

scraping and storage of social media data that could not otherwise be gathered (Batrinca and Treleaven 2014).

5 “Scraping—collecting online data from social media and other websites in the form of unstructured text and