Cover Page
The handle http://hdl.handle.net/1887/57990 holds various files of this Leiden University dissertation
Author: Groot, B.M.
Title: Selling cultural heritage?
Issue Date: 2017-09-26
Chapter 3
Research Method
The study uses a qualitative approach with documentary analysis as the key data source, and using the grounded theory method to guide the data collection. This was chosen as an appropriate method as the aim of this research was to engage in an explorative assessment to discover new inductive theory given the lack of existent research in this topic. The objective of this study is to investigate the MNC interaction with cultural heritage responding to the following research questions:
1. Why do MNCs care about cultural heritage? What values matter for MNCs?
2. How are MNCs involved with cultural heritage across industries? What examples exist from wider industries (beyond the extractive industries and tourism sectors)? What is the strategic rationale for this involvement?
3. What are the implications for regulation across industries and for the creation of private- sector partnerships?
The purpose of this chapter is to describe the chosen research methodology, to explain the sample selection, to describe the procedure used to design the data collection, and to provide an explanation of the data coding and analysis procedures. A list of the selected corporations and a further description of the coding process is also provided in the research findings in Part Three.
3.1 Document analysis
Document analysis is a systematic procedure for evaluating documents, which can be both printed and electronic materials. These can include advertisements and marketing materials, agendas, background papers and prior literature, diaries and journals, event programs, maps and charts, newspaper articles, press releases, program forms and applications, organizational and institutional reports, survey data, and other public records (Bowen 2009). Like other methods of qualitative research, the documentary data requires interpretation and analysis in order to elicit and meaning and discover new theory or knowledge (Corbin and Strauss 2008, Bowen 2009). Document analysis is often used within a triangulation method involving interviews, participant observation, or questionnaires and surveys; or used as a stand-alone method (Bowen 2009).
The key advantages of this method are its noted availability, efficiency, and cost-effectiveness. In
addition, documents are readily available and un-obtrusive. Document analysis requires detailed sourcing to specific the exact documents used, but with this data it is a highly stable method where the researcher’s presence does not affect the data, and where there is high potential for future comparative studies. This focus on publicly available information protects for future validity (for example, to compare the same research at a later date and provide a comparative research over time or across industries). By using public reports and documents the researcher has full access to a large sample to investigate and there was no risk of companies choosing to not participate. It also ensures a current picture as the reports are usually published annually or bi-annually and available directly online via the company website, and this also allows for future research to provide comparative assessments. As the majority of companies investigated were public limited companies (98%) there were also no accessibility issues as there were publicly available organizational reports for almost all companies. The only exception is in the case of privately- listed companies where no reports could be found, in this research being Rolex, and the exception of
Amazon where there is no publicly available corporate social responsibility report or discussion of social responsibility activities, although a corporate annual report is available and the absence of institutional documents could be covered by other documents. Corporate reports overall are therefore available, accessibly, and authentic and useful (in so far as they reflect the publicly-communicated and formal strategy of the MNC as a corporate unit).
The key disadvantages of this approach is that documents may not always be publicly available and even when available they may have low or insufficient detail. Most importantly, documents reflect a bias based on their source and intended creation. This issue of biased selectivity is especially relevant for corporate reports or in an organizational context where the selection of documents is likely to be aligned with internal procedures or policies, and with the organizational agenda or based on the purpose of its creation and the target audience, and may also be biased based on the “unit that handles record-keeping”, typically human resources (Bowen 2009, 32). Findings are limited to the MNC stated perspective, and moreover it is the MNC globally stated perspective. While this is valid to confirm the “existence’ of codes it cannot unpack the layers of how companies are involved and the drivers from different stakeholder perspectives both internally (such as employees) and externally (such as local stakeholders). Linguistic limitations as well as challenges of scope do not allow me to reflect the multi-dimensional host and home country statements, but this of course limits the analysis to this “global” MNC communication and identity. Many MNC subsidiary may publish complimentary local reports in their host countries (such as BMW China which has been active in cultural heritage interactions but where these examples are not reflected explicitly in the BMW global reporting).
The triangulation of data is done entirely through textual document analysis, focusing on available corporate and public documents. Email questionnaires and online surveys were also considered but rejected due to their many limitations that have been commented on in previous research that has tried to do similar assessment of the private-sector interaction (Starr 2013). Although questionnaires are relatively easy to set-up and easy to distribute to a large number of participants, they may be biased towards a certain type of participate who is biased to online work or has the time to complete the survey. As from other researches there are several learnings that CSR communities within large firms are unlikely to respond to mail survey approaches because they are already overwhelmed with email questions and surveys (Knox, Malak and French 2005, Starr 2013). Starr 2013 emphasized the lack of success of this approach due especially to the low response. In her research she comments likewise on the challenges of getting the right people in corporations to respond to an email survey, varying from practical issues like finding the right email address, to the problems of getting people to respond (sometimes a survey just arrives at a busy time and doesn’t get answered), as well as related challenges ensuring the sufficient quality of replies. In her 2013 research even “model” companies for cultural heritage such as American Express did not answer (one might assume employees of this company would be more empathetic to this field of research and therefore biased to respond due to the strong corporate vision of historic
preservation). Interviews are an alternative and highly valid method for in-depth analysis and would be likely to provide a multi-stakeholder perspective, but this makes them more suited for a focus on one or two key cases and not in this research with a focus on the discovery of new examples and cases.
3.2 Overall data collection and sample
The research uses the grounded theory to guide the data collection. The aim of grounded theory is
inductive discovery from the data. In grounded theory the data collection is conducted and then coded and grouped into concepts and through this allows the discovery of theory. Grounded theory provides
different heuristic tools to approach the coding via an open approach to help identify salient categories through constant comparison and iterative coding (Corbin and Strauss 2008). Open coding was suitable for this research where the theory of what companies are doing and why on a cross-industry examination, did not yet exist. In terms of the approach to coding, Strauss and Corbin suggest an analytical approach
that move beyond description only: in the way the codes serve as concepts that are uncovered through an iterative process of data collection and analysis.
3.3 CSR reports & Sample Selection
CSR reports were chosen for the analysis as they are one of the few (relatively) consistent and easily accessible document that can be assessed and compared across industries. The term CSR report refers broadly to refer to any similarly named or intended report, code of conduct, or comparable document of the MNC societal responsibility mission and impact and with the presumption that these codes of conduct or responsibility reports, in their existence and content, can be seen as informative about the context in which the MNE is operating. Codes reflect the MNCs thinking or strategy related to vital concerns and their perception of potential vulnerabilities, and the presence of a code may indicate a stronger
stakeholder importance and therefore stronger MNE concern with this potential topic (Reich 2005).
Using CSR reports as focused document analysis method is common for several previous studies, as discussed below. However, as stressed in these prior researches these reports should be analyzed in light of their broader context in order to draw any conclusions about impact. Schultz, Castello and Morsing 2013 suggest that CSR reports are a key source for determining MNC responsibility and its perceptions along with other public data, including Facebook pages and media reports. Knox, Malak and French 2005, provide an analysis of leading FTSE companies based on their CSR reporting and argue that the biggest companies like the extractive industries and telecom are most adept at prioritizing and identifying stakeholders and linking their CSR programs or codes to business and social outcomes. Martinez, Perez, and del Bosque 2013, provide a research of CSR in Melia Hotel, combining an analysis of reports and public media data with fourteen interviews. Reich 2005 assesses the differences in key strategic priorities across CSR reports from companies based in USA, Germany and Japan, and finding that these remain divergent and nationally differentiated, and not globally converging. Gordon and Miyake 1999, provide an analysis of 246 voluntary corporate codes of conduct in an OECD working paper that assesses the key strategic topics being prioritized by MNCs.
The data collection was based on public documents available largely online through corporate websites, and typically in English, published at the level of the ‘home’ MNC (i.e. not the subsidiary reports). These CSR reporting document vary by corporation, resulting in a range of names and types of documents (Mission and Vision statements, Code of Ethics, CSR report, Suppliers Reports). The reports used in the research were always the latest version publicly available at the time of research (based on access from January 2014 to Sept 2014). All reports were downloaded directly from a company website and typically included the corporate annual report and the corporate social responsibility report. This was triangulated with background research looking into key media and broadcasting commentary of the company as well as a search internally of the corporate website for other key documents, case studies or references to key corporate projects and activities.
Sample selection using RepTrak™
The sample of companies for the CSR report analysis was chosen using an external ranking by the Reputation Institute, which published an annual RepTrak™ list of The World's Best Regarded Companies for CSR. The research focuses on the list from 2013, this being the most current at the time of the
research. The reports or codes of conduct of the top fifty (n=50) ranked companies in the list were selected for analysis and then a content analysis was conducted based on these reports (Appendix A).
Why use a list for the selection of MNCs? The main reason for this focused selection was to provide a specific analysis across industries and to avoid a biased selection of cases. The research question required a broad design, and this ranking allowed for an objective list and rationale for selection. Failing to use any
list would result in a subjective selection of corporations based on my own geographic region, linguistic abilities, and literature and broadcasting exposure. Using a ranking or list system therefore helped to mitigate this potential familiarity bias.
There are several available “global” lists related to international corporate social responsibility and reputation (more than 183 public lists worldwide from 38 countries, according to the Reputation Institute per their website in 2014). Civic society lists could also be used to provide a selection of companies: for example, Greenpeace’s Guide of Green Companies (Greenpeace 2012) for green electronics, or Oxfam’s “Behind the Brands” list which provides a company score-card for the “Big Ten” food related corporations and their social responsibility policies regarding agricultural sourcing from developing countries (Oxfam 2015). However, no such listing for cultural heritage exists today. The research therefore turned to other fields to find a suitable ranking. The most commonly used lists are the
RepTrak™ list and the Forbes list of America’s 100 Most Trustworthy Companies and Western Europe’s 50 Most Trustworthy Companies (existing only as of 2014)and the Forbes List of the World’s Most Ethical Companies (existing since 2011). These Forbes listings of Trustworthy Companies was deselected for this research because of the specific financial focus and their lack of worldwide stakeholder inclusion.
The 100 Most Trustworthy companies list by Forbes is highly USA-centric, and moreover it focuses on the measure of trust as understood from a total shareholder revenue perspective based on average scores for trustworthy corporate behavior and accounting and governance risk. The list of The World’s Most Ethical Companies was deselected it due to focus which is too internal for this purpose, looking internally on how MNCs are managing their ethics and the overall ethics of trust within the corporation.
In contrast, the RepTrak™ ranking invites 55,000 (as of 2013) consumers across fifteen countries to rank their perception of global corporations’ social responsibility, giving best regional indices and a final worldwide index. The Reputation Institute system measures company’s ability to deliver on stakeholder expectations in seven key areas: their products and services, innovation, workplace, governance, citizenship, leadership, and financial performance. To qualify for the ranking each company has to already be present with an above-average score based on the scores from the five precedent years (in this case 2007-2012). The criteria for qualification also limits companies based on size, revenue, and
multinational presence, this to ensure familiarity among the stakeholders being interviewed (with a minimum revenue of US$1 billion globally) (Reputation Institute 2013). The benefits of the RepTrak™
Institute list compared to other listings is that, first, it provides a global ranking. To compile the final global list, the Reputation Institute combines key ranking companies from the regional lists provided (for 30 countries). Second, it focuses on CSR from the stakeholder perception with a vast scope of data behind its selection with thousands of stakeholder responses through interviews. Third, it is an established annual ranking, provided since 2007 and with a growing update of respondents and countries on an almost annual basis, allowing also for enhanced reliability and validity through extension to future comparative or longitudinal researches. RepTrak rankings have also been used in recent academic researches, such as by Ahn and Wu 2015, who used reputation data from RepTrak on the level of country analysis. Their research used the RepTrak as a ranking of country reputation score in research on nation-branding (analyzing how internal cultural policy affects country-level reputation, such as whether a cultural ministry exists).
However, there are still several limitations of this list as ranking method. First, it does not capture or explain who the stakeholders are who have been asked to participate, which leaves the question of the
‘most reputable’ companies according to whom? Second, it is focused on perceived reputation according to ‘public’ view. This is an assessment therefore of perceived reputation and not a normative assessment of company practices. Third, the sample is limited by the scale and size of revenue, due to the
requirement related to income (US$ 1 billion). This selection methodology biases the samples to large corporations: due to the size criteria of the ranking it is also limited to the largest multinational enterprises (measured by annual revenue) rather than a more balanced mix of different small and medium sized
private-sector corporations. It is likewise limited by requirements of awareness (the interviewees need to be familiar with the corporations, therefore it will show a preference for corporations that have higher public awareness). The method asks to respondents regarding their familiarity with the company, meaning that companies which do not have high awareness will also not make the top 50 or 100 list. This therefore limits the inclusion of companies based not only on their activities but also their “globalism” and likely on their media awareness. Finally, the list focuses explicit on corporate reputation, but it does not allow for a more granular understanding at the brand-level.
The sample size was original approached with an estimated sample of twenty companies, that was later extend to fifty (n=50) companies due to the lack of industry variety within the top twenty corporations in the list. The expanded sample size of fifty corporations allowed for broader discovery across sectors. Two firms were eventually excluded due to no (public) code or data related to their CSR reporting. The data collection is based on a snapshot based on the situation in 2012-2013 during the data collection for this sample.
3.4 Broader survey of MNC examples
Chapter Five provides a wide survey of cultural heritage interactions looking at examples from marketing and advertising. This is an area that is largely absent from the CSR reporting, but it is an important area of documented data where corporations are involved with cultural heritage. The survey provides rich
descriptions of the specific phenomenon of cultural heritage interactions by corporations, allowing for broader theory building and generalization across industries. The cases are analyzed through a framework of push (reactive) and pull (proactive) strategies which allows for an empirical discussion of both how MNCs are involved with cultural heritage, and why MNCs are involved with cultural heritage.
The data sources used include a wider range of documents, including existent case studies and surveys within secondary literature, broadcasting documents, commentary and materials from social media and blog sites, marketing and advertising materials, newspaper articles, organizational and institutional reports, press releases, and reports on corporate activity references by media, communication and advertising agencies. Additional data came from informal interviews with different managers and corporate executive working in the private-sector in the fields of marketing, communications, media broadcasting, and advertising agencies (n=26) conducted from Jan-July 2015; and from personal communication with researchers and experts in the field of heritage management. This broader documentary analysis is necessary to include wider industries that were not included in the RepTrak listing, and to expand the research beyond the scope of corporate social responsibility and responsibility reporting. Similar textual analysis has also been used through in anthropological research to provide critiques of advertising and marketing campaign (see discussion in Benson and Kirsch 2010, 464) although where the focus in previous anthropological research has been on an ethical basis, this thesis intends a similar discussion concerned with the MNC drivers, motives, and strategy.
In terms of the research design, the survey has two key limitations: first it is not an exclusive list of examples. Moreover, it is biased towards English language resources, with a regional bias towards examples from Europe where I was located during the data collection. It should therefore be seen as a useful survey but it is of course biased in the selection of corporations, and it is not intended as an exclusive list of all corporate activities or of each corporation’s global involvement.
3.5 Selection of case studies
Chapter Six provides a more in-depth analysis of four key corporations: Rio Tinto, American Express, Google, and Unilever. The theoretical sampling of cases is chosen because they are especially revelatory examples from four diverse sectors. Analyzed together, they allow for contrasting perspective of
corporate involvement and for inductive theory building with each case giving insight into different industry environments. The case of Rio Tinto is important for the perspective of regulatory pressures, focused on the government involvement and the corporations appeal to social-license to operate. The case of American Express is important for the combination of philanthropic actions and cause-marketing. It also highlights important theory around the distinction of philanthropic and branding functions, looking at American Express Foundation as a separate entity from the American Express marketing unit. The Google example is chosen as an illustration where cultural heritage overlaps with core business objectives. Finally, the Unilever examples highlights the importance of theory building defined at the brand-level of analysis, instead of at the corporate level. These cases highlight the specific theme of the value of cultural heritage for MNCs and why cultural heritage matters to corporations. As will be seen in the analysis, the multiple case approach allows for the extension of theory to cover broader contexts.
3.6 Selection of key UN instruments
Chapter Seven takes a different approach looking at key UN instruments affecting the corporate strategy and interactions. Originally this area of research was not included in the research design, but it was added during the research due to the importance of key codes revealed from the CSR reports and wider survey.
The instruments are selected based on the coding and data collection in the previous chapters; in other words, selecting those UN covenants and instruments that are featured most prominently within the MNC reporting and which are therefore assumed as the most relevant for the research. The analysis of these UN instruments is based both on secondary literature and through the respective UN databases.
3.7 Limitations
Codes used by MNCs for their CSR reporting and discussion vary in their sophistication and scope ranging from one-page mission statements to full reports and impact assessments. There is a significant variance in the design and content of the documents analyzed, the number of documents available, and the type of content presented. Moreover, there is no standard lexicon for codes overall within the field of CSR nor specifically to cultural heritage. This leaves much open to subjective interpretation on the part of the researcher (Reich 2005). MNCs also lack a standard nomenclature and terminology for their coding process, and there is no uniform approach across MNEs concerning what to include and how to describe them. As seen across CSR reports, codes often lack provisions for monitoring and enforcement and enforcements that do exist are self-regulatory (Reich 2005). These codes of conduct may be created for very different audiences such as guidelines for employees, guidelines for suppliers or partners, or a statement of the company’s commitment to the public (Gordon and Miyake 1999).
The sample of corporations will also be biased towards the largest multinationals, and this leads to some ethical considerations as often it is those companies with the largest revenues and the broadest web of operations and communication which inflict the largest (negative) externalities and impact (Klein 2010).
Whilst this is an important ethical question regarding the theme of engagement with the private-sector overall, it has no impact on the findings of this research regarding the MNC role.
The document analysis uses a variety of data sources beyond these corporate reports, but the data is still biased towards institutional documents. Specifically, this research does not focus on the local stakeholder interests in the projects and examples studied. Local interests are sometime reports in public broadcasting, news articles, and similar blog and opinion pages, but this is often decontextualized from local
stakeholders even with access to blogs and social media there is only a very limited perspective of the
“local community” given in this research. However, it is not seen as a weakness because the focus is on the development of new theory regarding how and why multinational corporations interact with cultural heritage and not on any normative assessment of these interactions and their (local) impact.
