Qualitative Interviews

The study’s primary data source is comprised of qualitative, semi-structured interviews conducted with actors throughout the Bangladesh RMG industry whom were either directly or indirectly involved with the Accord and Alliance. Standard themes were identified which guided the

39

creation of the interview guide. Accord and Alliance brand members served as the principal research subjects, so the themes for this version of the interview guide consisted of:

• Creation and/or signing of the Accord/Alliance agreement o What caused them to join

o Why they joined, and why that particular PG instead of the other

• Perceived roles and responsibilities of different actors groups as related to ensuring the sustainability of international supply chains in both an ideal world as well as the real world: brands, governments, international governing bodies (e.g. ILO), civil society organizations, consumers

• Dynamics with the other stakeholders: Accord/Alliance organization, other brand members, involved civil society organizations, customers

• Perception of the Accord’s legal enforcement clause

• Operationalization of the Accord/Alliance within the organization: resources required/dedicated, change in sourcing practices, position of CSR within the organization and sourcing process

Interview guides for other actor groups like trade union members focused on these same general themes but were adapted fittingly to the research subject. Whilst always including these themes, interviews were also tailored to specific research subjects. For example, H&M negotiated but refused to sign the agreement which preceded the Accord, but it was boastful afterwards for being the “first” brand to sign the Accord. So, H&M’s interview also included many questions related to the processes and decisions surrounding those events. Similarly, interview guides were also adapted to reflect terminology consistent with that of the research subjects. For instance, whilst most companies refer to their employees as such, Target calls theirs “team members”, so this terminology was adopted for the sake of ensuring clarity of concept for the interviewees. Example interview guides can be found in Appendix 3: Interview Guides. Three have been included which illustrate the tailored nature of the research: Gap as an Alliance interview, Bestseller as an Accord interview, and Danish Fashion & Textile as an “expert”, non-brand source with extensive knowledge of the industry and actors. While the wording of the questions may change, the themes remain consistent.

40

The study included interviews with representatives of brands, labor unions, NGOs and the Accord and Alliance organizations themselves. The study first targeted those closest the phenomenon at hand: the Accord and Alliance organizations. The Accord was very generous with its time and insights, something it had done for many researchers. The Alliance, however, was much more closed and greatly limited access by researchers; fortunately, after more than a year of persistence and networking the researcher finally gained good access to the Alliance’s leaders and board of directors. For brands, the study targeted those which played a role in the establishment of the initiatives, such as Gap, H&M, PVH, Bestseller and Target. Other key players in the Accord and Alliance were also targeted, which included IndustrALL Global Union, Clean Clothes Campaign and the ILO. From these interviews, snowball sampling was employed, with the author gaining referrals to additional research subjects. Saturation was reached when interviewees referred the researcher to those that had already been targeted for the research.

The author also sought out experts not directly involved in either of the PG efforts to gain an

“outside” perspective, but still within the industry; indeed, many of the organizations were those mentioned by research subjects during the interviews. The Sustainable Apparel Coalition and IDH Sustainable Trade Initiative are examples of organizations which weren’t directly involved in the Accord or Alliance, but were brought up by multiple interviewees during the course of the research. Connections to research participants were made either through the author’s own network or though “cold calling”, usually via targeted LinkedIn messages or emails. A large number of the research subjects came from Scandinavia, reflective both of the author’s location – and hence, network – in Denmark, as well as the actions of many of the Scandinavian brands as leaders within the Accord.

Interviews were conducted primarily in two “rounds”. The first eleven interviews were conducted by Master’s thesis students supervised by the author. The interview guide was developed in tandem, and the interviewees were selected together. These served as pilot interviews for the main data collection of the PhD study. In this spirit, the author obtained the full transcripts from the interviews but conducted her own coding of the data. The author herself conducted all of the remaining 30 interviews. Data from the pilot was only used in this PhD if it provided unique perspectives or insights that the author was not able to collect; for example, one interviewee from the pilot study declined to participate in a follow-up in interview, and in others, the passage of time sometimes also resulted in different reflections and insights from respondents. Interviews

41

were conducted in person when possible, including during fieldwork in Bangladesh in June 2016.

Research travels allowed the author to conduct in person interviews in Canada, Denmark, the Netherlands, the UK and the U.S. When meeting in person was not possible, interviews were conducted via Skype or phone. In all, 36 research subjects from 30 organizations participated in a total of 41 interviews; some interviews included two research subjects at a time, and other subjects were interviewed twice.

Table 2. Account of interviews by category.

Organizational Affiliation n=

Accord members 23

Represents approx.

10% of members

Alliance members 8

Represents 28%

of members

No official affiliation 10

HQ Location

North America 9

Scandinavia 18

Continental Europe + UK 9

Bangladesh 5

Sector

Private sector (brands) 22

Civil society organizations (labor

unions, NGOs) 12

Accord and Alliance organizations 6

Public sector (ILO) 1

Those interviewed twice include many of those whom were previously interviewed by the author’s Masters students, which offered the opportunity to ask follow-up or more specific

42

questions based on the subjects and themes from the first round of interviews. Bestseller was an example of this, and the interview guide provided in Appendix 3: Interview Guides reflects the areas of follow-up that a second round of interview afforded. Other research subjects interviewed multiple times were staff of the Accord and Alliance. An overview of the research subjects can be found in Table 2. In all, the interviews yielded a total of 42+ hours of audio recordings. The substantive content portions of the interviews were fully transcribed; sidebar or irrelevant conversations – such as about the weather, travels, children or the like – were not. All told, the study-related portions relevant for the research resulted in around 900 pages of single-spaced transcripts.