• No results found

The Influence of Different Institutional Environments on Reactivity

N/A
N/A
Protected

Academic year: 2021

Share "The Influence of Different Institutional Environments on Reactivity"

Copied!
60
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

MSc International Business and Management 2019-2020 University of Groningen, Faculty of Business and Economics

Master thesis

The Influence of Different Institutional

Environments on Reactivity

A study within a CSR context

Daniel Král Lau Lam

S4137205

d.k.l.l.kral.lau.lam@student.rug.nl

Supervisor: Dr. Rieneke Slager Co-assessor: Dr. J. Shin

Word count: 13948

(2)

1

Abstract

Whilst we know from previous literature that companies respond to CSR rankings by improving their performance, we are yet to understand how home institutional environment influences this response. Therefore, this study explores the influence of different institutional environments on reactivity to rankings. Through the use of a configurational approach, together with a QCA, we reach five explanatory configurations for the outcome of study, reactivity. The combination of these configurations together with the theoretical framework allows us to confirm the possible configurational nature of reactivity, find that institutional environments do in fact influence reactivity within CSR rankings (LME vs CME), and, more importantly, find that an explicit CSR approach (LME) can lead to a better outcome within CSR rankings, such as the CHRB.

Keywords: Corporate Social Responsibility (CSR), Human Rights, Ranking, Reactivity,

(3)

2

Acknowledgements

First, I would like to thank my supervisor, Dr. Rieneke Slager, for all her support and guidance during this process. For all the meetings discussing the thesis that have sharpened my thoughts and ideas.

To my family for their continuous support, interest, and care. I especially want to thank my parents, for all the hard work and effort, values and guidance, that they have provided me throughout my life.

With this new experience in Groningen came new friends and my experience wouldn’t have been half as fun and enjoyable without the people I met. For all the discussions, library breaks, lunches, dinners, joys, that we shared during those months.

I would like to thank Matilde for her tremendous support through this new challenge, for her patience and kindness, and her continuous motivation.

(4)

3

Table of Contents

ABSTRACT 1 ACKNOWLEDGEMENTS 2 LIST OF FIGURES 5 LIST OF TABLES 5 LIST OF ABBREVIATIONS 5 1. INTRODUCTION 6

2.

LITERATURE REVIEW 9

2.1 REACTIVITY AND CSR RANKINGS 9

Corporate Social Responsibility, Rankings and Reactivity 9

2.2 INSTITUTIONAL THEORY AND CSR 14

Institutional environments and Corporate social Responsibility 14

2.3 Summary 18

3. RESEARCH METHODOLOGY 19

3.1 THE CHRB – THE CONTEXT 19

3.2 DATA COLLECTION 20

3.3 CONFIGURATIONAL APPROACH 21

3.4 CONDITIONS AND CALIBRATION OF THE QCA 22

3.4.1 Outcome 23 3.4.2 Explanatory conditions 24 Economy 24 Old 24 Serious allegations 25 Engagement. 25

Human Rights due diligence 26

(5)

4

5. DISCUSSION 33

6. LIMITATIONS AND FUTURE RESEARCH 40

7. CONCLUSION 42

REFERENCES 43

APPENDICES 50

APPENDIX A 50

Table 3: Truth table- Presence of Outcome 50

Table 4: Truth table- Absence of Outcome 51

Table 5: Necessary condition test 52

APPENDIX B 53

Table 6: Country, Company and Economy 53

Table 7: Total number count 53

APPENDIX C 54

Table 8: Companies included in the Agricultural Products industry 54 Table 9: Companies included in the Apparel industry 55 Table 10: Companies included in the Extractives industry 56 Table 11: Companies included in the ICT Manufacturing industry 57

APPENDIX D 58

Figure 2: CHRB Measurement Themes 58

Figure 3: CHRB Weighting of Measurement Themes 58

(6)

5

List of Figures

Figure 1: Implicit and Explicit CSR 17

List of Tables

Table 1: Calibration Table 28

Table 2: Outcome of the QCA Analysis 31

List of abbreviations

APG APG Asset Management

BHRRC Business & Human Rights Resource Center CHRB Corporate Human Rights Benchmark CME Coordinated Market Economies CSR Corporate Social Responsibility

fsQCA Fuzzy Set Qualitative Comparative Analysis IB International Business

IHRB The Institute for Human Rights and Business LME Liberal Market Economies

MNEs Multinational Enterprises NBS National Business systems

NGOs Non-Governmental Organizations QCA Qualitative Comparative Analysis SDGs Sustainability Development Goals

UN United Nations

UNGP Guiding Principles on Business and Human Rights

US United States

(7)

6

1. INTRODUCTION

In recent years, society is being characterized by its ‘grand challenges’: its pressing social, political, and environmental issues that go far beyond borders (Wettstein et al., 2019). These transnational challenges affect society in general, its communities and its people, and therefore they must be tackled through collaborative efforts. In this sense, the debate around Corporate

Social Responsibility (CSR) and human rights is gaining more and more relevance in the last

years (McWilliams & Siegel, 2001), as states and state agents cannot be solely held accountable for maintaining and respecting them (Muchlinski, 2001). In this regard, it is now expected that MNE’s integrate CSR and human rights as an extension of their responsibilities. However, the concept of CSR is surrounded by ambiguity (Matten & Moon, 2008), which leads to different approaches and strategies towards CSR within MNE’s (van Marrewjik, 2003). This is where CSR rankings such as the Corporate Human Rights Benchmark (CHRB), as mechanism of surveillance and control ( T.M. Porter, 1996; Sauder & Espeland, 2009) and public methods of evaluation (Martins, 2005) come into play in order to homogenize and incite engagement from MNEs towards CSR.

In the past few decades’ rankings have proliferated considerably in importance within our society (Hazelkorn, 2014), which led to an increase of various studies within the theoretical field of rankings (Rindova et al., 2018). As disciplinary practices that exercise power over organizations (Sauder & Espeland, 2009), rankings inherently generate reactivity, a response resultant from different factors (Espeland & Sauder, 2007), which can lead companies to improve their CSR (Chatterji & Toffel, 2010; Lewis and Carlos, 2019). Social rankings such as the CHRB, through standardized metrics of evaluation, rely on this understanding that rankings generate a response and influence those ranked (Sharkey et al., 2014; Rowley et al., 2016; Lewis, 2018), to drive companies towards the “agenda” they intend, as “organisations react to metrics in order to obtain favourable outcomes in the process of being publicly measured and ranked” (Slager et al., 2020: 2).

(8)

7 existing ambiguity within the concept of CSR (Dahlsrud, 2008) together with the understanding that reactivity can be configurational in nature as it can be influenced by various factors (Espeland & Sauder, 2007, Slager, 2020), presents itself as a plausible explanation. Especially when considering how different organisational factors can affect said response (Espeland & Sauder, 2007).

An organisational factor that is important to account for when considering reactivity in CSR rankings, is the home institutional environment of MNEs. While studies have shown how institutional environments affect CSR (Matten & Moon: 2008; Jackson & Apostolakou, 2010; Ioannou & Serafeim, 2012), there is still a gap as to how such effects affect reactivity.

Rooted in the distinction between liberal market economies (LME) and coordinated market economies (CME) (Hall & Soskice, 2001), there has been proven to exist a distinction in approaches towards CSR; explicit and implicit (Matten & Moon, 2008). Whereas explicit CSR is associated with LME’s, implicit CSR is associated with CME’s (Matten & Moon, 2008). This distinction in approach resultant from the institutional environment from which companies come from, brings us to the existing theoretical gap, that hasn’t further explored the extent to which this distinction reflects upon MNEs response, reactivity, towards CSR rankings.

In this study, we explore how MNEs home institutional environments influence corporate reactivity to CSR rankings. In doing so, we explore how an organizational factor such as geographic location, affects companies’ responses to social rankings, and therefore address the existing gap and add to the literature by pursuing a perspective on reactivity that hasn’t been further explored. The combination and understanding we currently have on these two perspectives, reactivity and CSR rankings, and institutional environments and CSR, then brings us to our research question: How do different institutional environments affect MNE’s

reactivity within a CSR context?

(9)

8 difference in institutional environments, here defined through the use of VoC (Hall & Soskice, 2001), affects reactivity.

Thought the analysis of our resulting configurations, combined with the theoretical framework, we provide a further indication on the configurational aspect of reactivity. More importantly, we find that the differences in institutional environments do in fact, influence reactivity within a CSR context (LME vs CME). And further add to this understanding, by presenting how this effect occurs as a result of the explicit vs implicit approach to CSR. Furthermore, we contribute by arguing how the explicit approach leads to a better outcome to CSR rankings such as the CHRB, as a result of the inherent values and norms of its market (LME). This study then contributes to the fields of CSR and IB, by providing additional insight into how differences institutional environments affect the reactivity of MNE’s, to CSR rankings such as the CHRB.

(10)

9

2. LITERATURE REVIEW

This chapter provides a comprehensive analysis of the theoretical framework relevant to this study, which links together theories and concepts in order to better understand the reasoning, rationale and conclusions achieved in this research. Two main perspectives are being presented in this literature. In the first part of this chapter, it is important to start analysing the existing theory on Corporate Social Responsibility, Rankings, and Reactivity, in order to understand the relation between these three notions. In the second part of this chapter, we explore how institutional level differences across environments affect the approach towards CSR. By combining these two perspectives, we can then pursue the previously overlooked aspect of how home institutional environment influences reactivity to CSR rankings.

2.1 CSR rankings and Reactivity

Corporate Social Responsibility, Rankings and Reactivity

More than ever, Multinational Enterprises (MNE’s) are being held accountable for maintaining, respecting and integrating corporate social responsibility and human rights as an extension of their own responsibilities (Wettstein et al., 2019), as the past decades have been characterized by major transformations at a global level. As governments can no longer be the exclusive actor addressing human rights and social responsibility, the public role and influence of MNE’s in this regard have grown in importance (Scherer & Palazzo, 2008).

CSR can be defined as, “CSR (and its synonyms) empirically consists of clearly articulated and communicated policies and practices of corporations that reflect business responsibility for some wider societal good. Yet the precise manifestation and direction of the responsibility lie at the discretion of the corporation” (Matten & Moon, 2008: 405). However, there are various definitions for this concept as a result of its continuous evolution, which entails a certain degree of ambiguity around it (Carrol, 1999; Dahlsrud, 2008; Hamidu et al., 2015).

(11)

10 growing awareness from MNEs towards CSR. An example of this is the growing response from CEOs and top managers to the pressures over social issues imposed by stakeholders (Bansal & Roth, 2000; Crilly, Zollo, & Hansen, 2012). However, despite its growth (McWilliams & Siegel, 2001), there is still a lack of consensus on its general definition (Carroll, 1999; Lockett et al., 2006; Crane et al.,2008). This ambiguity surrounding CSR (Dahlsrud, 2008; Matten & Moon, 2008;), allows MNE’s to use different strategies within CSR depending on their interests (van Marrewjik, 2003), which then leads to different ways to engage in CSR across the board. This, consequently, leads to the necessity to create international codes of conduct, and standards to regulate and evaluate MNE’s commitment with CSR and human rights (Aguilera et al., 2018).

Simultaneously, this awareness around CSR within MNE’s developed hand in hand with the increasing international concern over the possible negative impact of MNE’s operations (Moran, 2009; Kolk & van Tulder, 2010). As the topic of human rights has been neglected on the literature around CSR (Wettestein, 2019), there is a growing literature and debate around the necessity to assess, measure and mitigate the human rights impacts, through mechanisms such as human rights due diligence (Fasterling & Demuijnck, 2013; Fasterling 2017). This way, and through mechanisms such as human rights due diligence, the state is contributing to the transformation of human rights from a voluntary basis to a normative standard (Martin-Ortega, 2013). Where through establishing human rights due diligence processes and addressing human rights with specific policies within CSR, firms are better able to identify and prevent harmful impacts of their operations (McCorquodale et al., 2017; Obara & Peattie 2017).

(12)

11 This need to asses and measure a concept such as CSR, brings us to organizational rankings, commonly associated as public methods of evaluation (Martins, 2005). As in the past few decades’ rankings have proliferated considerably in importance within our society (Hazelkorn, 2014). As defined by Martins, organizational rankings are “publicly available comparative orderings of organizations, based on evaluation criteria determined by a ranking organization” (2005: 701). While rankings can in fact influence and affect MNE’s behaviours (Sharkey et al., 2014), they can do so in both a positive and negative way (Rowley et al., 2016; Lewis, 2018), as rankings create expectations and generate reactions, they therefore change the way people make sense of situations (Espeland & Sauder, 2007).

Here is where social rankings such as the Corporate Human Rights Benchmark (CHRB) come into play. Rankings, through the use of standardized metrics, help mitigate the ambiguity within CSR and force MNE’s to engage with it in a more homogeneous manner, leading MNE’s to engage in a more positive way with CSR and human rights (Chatterji & Toffel, 2010). Rankings therefore fill the gap between organizations and audiences, as they provide unbiased information to stakeholder as to better inform them (Schneiberg & Bartley, 2001; Bartley & Schneiberg, 2002).

While becoming more prevalent in our society (Hazelkorn, 2014), rankings commonly take three perspectives when considered: as means of information intermediation; as comparative orderings; and as means of surveillance and control (Rindova & Fombrun, 1999; Sauder & Espeland, 2009; Rindova et al., 2018). “Taken together, these perspectives emphasize that rankings perform a technical function (provision of information), a social function (conferring reputation and/or status and patterning resource exchanges), and a political function (control and power redistribution).” (Rindova et al., 2018: 2191). As a result of this growth in relevance, rankings have gained importance amongst MNEs environment, as they can impact the relationship with stakeholders (Rindova & Fombrun, 1999; Rindova & Martins, 2012), and generate a level of response on those being ranked (Espeland & Sauder, 2007).

(13)

12 Given that rankings induce a response, commonly associated as reactivity, a response resultant from different factors (Espeland & Sauder, 2007). CSR rankings such as the CHRB rely on this idea that rankings can thus provoke a reaction, that could ultimately steer those ranked towards the goal and metrics of those who create the rankings (Chatterji & Toffel, 2010). Two examples that support this notion are the work of Chatterji and Toffel (2010) and Lewis and Carlos (2019), where they provide evidence that CSR rankings can in fact influence MNEs. Chatterji and Toffel (2010) studied the responsiveness of MNEs to a corporate environmental rating and found that companies initially poorly rated, did in fact subsequently improve their environmental performance as a result of being ranked. These findings proved important as they provided evidence that rankings at a CSR level could in fact induce a positive response from the companies rated. Lewis and Carlos (2019) on the other hand, indicated that CSR rankings did in fact wield power when considering CSR, their study implied that CSR rankings did bring some accountability to those being ranked and did in fact influence performance.

While these studies provide evidence that CSR rankings can, in fact, influence a company’s social performance and engagement, prior research shows us that companies can respond differently to the same rating system (Delmas & Toffel, 2008; Philippe & Durand, 2011; Crilly et al., 2012). Crilly et al. (2012), for instance, through their in-depth study of multinational corporations, found that “firms facing identical pressures decouple policy from practice in different ways and for different reasons” (Crilly et al., 2012: 1429). Which raises the question as to why these responses vary. A factor that might help understand why responses may vary to the same raking, within a CSR context, is the ambiguity and lack of agreement on the concept of CSR. However, a factor that may take a more prominent role in this variance in responses, is the concept of reactivity itself, as it can be influenced by various factors. It is therefore important to understand the extent to which these responses to rankings vary (Crilly et al., 2012) given the parallel growing relevance of both rankings (Hazelkorn, 2014) and CSR (Bansal et al., 2000).

(14)

13 2007). As a result of these adjustments, whether intentional or not, understanding the consequences and responsiveness of said reflections and reactions is thus important, as they can affect the outcome at hand (Espeland & Sauder, 2007).

As we now know, rankings are conceptualized as disciplinary practices that exercise power over organizations “through incentives that are simultaneously seductive and coercive” (Sauder & Espeland, 2009: 64). While considering this coercive power that rankings have, we must then consider an important aspect within reactivity (Espeland & Sauder, 2007). The idea that this effect that rankings generate, reactivity, is influenced by organizational factors (Espeland & Sauder, 2007). As presented within the study of Espeland and Sauder (2007), rankings do in fact generate a response, reactivity, however, this level of response isn’t linear across the board. Factors such as geography, management, legitimacy, or company goals and ideologies, can influence the response of those being ranked (Espeland & Sauder, 2007). This notion is therefore of high relevance as it indicates that reactivity can have a configurational nature as it is a result of various organizational factors (Espeland & Sauder, 2007; Slager et al., 2020).

By having the understanding that reactivity can be configurational (Slager et al., 2020), one must therefore account that the response to rankings such as the CHRB can be influenced by multiple factors. Hence the configurational aspect, as it indicates how various organizational factors can influence reactivity (Espeland & Sauder, 2007), rather than a simplified view in which solely one factor influences reactivity alone.

These organizational factors are therefore important to consider upon studying reactivity, as they influence the response of MNEs to rankings such as the CHRB. However, studies that take into consideration these organizational factors are still scares. As such, there is a need to study how these different factors affect reactivity.

(15)

14 in response and this specific organizational factor, is that of Matten and Moon (2008), as it relied on an institutional perspective, in part provided by Varieties of Capitalism (VoC) (Hall & Soskice, 2001), to present evidence of a distinction at a CSR level.

2.2

Institutional theory and CSR

Institutional Environments and Corporate Social Responsibility

As commonly denoted, institutions are the ‘‘rules of the game’’ (North, 1990) as they tend to be embedded in societies and govern their behaviours (Meyer & Rowan, 1977). MNEs are no exception to them, as they operate within the social scope of institutions. MNEs activities are faced with various pressures from different institutional environments (Campbell et al., 2012) which therefore influence their decision making (Campbell, 2007).

Given the prevailing ambiguity in CSR’s definition, MNEs CSR activities are most commonly framed to their social contexts, and thus become a reflection of the prevailing institutions within them (Jackson & Apostolakou, 2010). Which brings us to studies such as that of Matten and Moon (2008), that relies on an institutional perspective, based on the institutional framework of Varieties of Capitalism (VoC) (Hall & Soskice, 2001) and National business systems (NBS) (Whitley, 1997: 1999), to explore a distinction at a CSR level resultant from differences at an institutional level environment. Through this lens of Matten and Moon (2008) we can explore how the geographical location (institutional environment) of a MNE affects its response, reactivity, to rankings, given the configurational nature of reactivity that allows organisational factors to influence it.

(16)

15 responses (Hall & Soskice, 2001; Porter, 2003). The main difference lies in how each economy organises itself. In LME, “firms coordinate their activities primarily via hierarchies and competitive market arrangements”, while in CME, “firms depend more heavily on non‐market relationships to coordinate their endeavours with other actors and to construct their core competencies.” (Hall & Soskice, 2001: 7). Within their work, the authors provide examples of such economies where countries such as the United States (US) or the United Kingdom (UK) are considered LMEs, and countries such as Germany and Japan are considered CMEs (Hall & Soskice, 2001).

The understanding of this institutional framework allows us to explore how an institutional environment affect reactivity, and then brings us to the work of Matten and Moon (2008), where through this distinction (LME vs CME), explored how and why CSR differs within institutional environments. By comparing two different regions (US and Europe), Matten and Moon (2008) made a distinction between explicit and implicit CSR. Whereas the US incentives and leads companies to take a more explicit approach, in Europe, CSR is driven towards a more implicit approach.

Explicit CSR is referred “as corporate policies that assume and articulate responsibility for some societal interests” (Matten & Moon, 2008: 409), in which CSR is a voluntarily part of a company or corporation’s strategy, rather than being a result of existing pressures or regulations, on the other hand, implicit CSR, presented as “corporations' role within the wider formal and informal institutions for society's interests and concerns.” (Matten & Moon, 2008: 409), is observed or expected when mandatory values, norms and rules steer companies CSR. Their study argued therefore that as the adoption of CSR propagates globally (Chapple & Moon, 2005; Puppim de Oliveira & Vargas, 2005), the approach in which companies would enact CSR would then be in part a result of the institutional context in which they were set, and thus, given the institutional differences across countries, a differentiation between explicit and implicit CSR would exist amongst regions.

(17)

16 This notion that differences in institutional environments influence CSR, is further supported through the understanding that distinct systems (political, financial, education and labour, and cultural) play a role in MNEs CSR, and that the political system, clearly linked with the distinction between LME and CME, plays a more prominent role in this regard (Ioannou & Serafeim, 2012). Furthermore, while further providing evidence of such an association (LME/explicit vs CME /implicit), studies such as that of Jackson and Apostolakou (2010), add to this understanding by finding evidence that companies within LMEs can achieve a higher performance at a CSR level, when compared to companies within CMEs. Studies such as that of Ioannou and Serafeim (2012) or Jackson and Apostolakou (2010), provide empirical proof on the role national-level institutions play in explaining the variation of companies’ performance at a CSR level and further add to this understanding of the institutional environment role on companies CSR.

Figure 1 – Implicit and Explicit CSR (Matten & Moon, 2008)

(18)

17 these effects affect company’s reactivity within CSR rankings. As the explicit vs implicit approach to CSR, rooted in distinct approaches within markets (LME vs CME), should therefore influence the response and engagement of those ranked.

However, it important to note that while in their initial paper, Matten and Moon (2008) alluded to the idea that a shift towards an explicit CSR approach could become more common depending on the strength of traditional institutions. In their most recent work (2020), a reflection on their 2008 paper, they built upon their initial conceptualization of explicit and implicit CSR and changed their initial perspective as a result of the changes that occurred in CSR through the last decade (Matten & Moon, 2020). Two new concepts where presented: explicitization and implicitization of CSR (Matten & Moon, 2020). These two approaches, in essence, were a result of a hybridization of both explicit and implicit CSR. Explicitization refers to “the process by which norms and rules associated with implicit CSR are adopted in explicit CSR policies, practices, and strategies” (Matten & Moon, 2020: 7). While, implicitization works in an inverse way, in which countries deemed more implicit, find themselves incorporating “new/reinvigorated values, norms, and rules for corporations that are informed by policies, strategies, and practices of explicit CSR” (Matten & Moon, 2020, p. 20).

This new paper then brings additional support in understanding that CSR in itself isn’t linear across the board, and thus creates and enacts different responses across subjects. As accurately said CSR “means something but not always the same thing to everybody” (Votaw, 1972: 25), and as such, it becomes easier to understand how and why responses towards CSR can vary across different institutional environments.

(19)

18

2.3 Summary

In recent years, CSR is growing in importance (McWilliams & Siegel, 2001). MNE’s are now being held accountable for maintaining and respecting social responsibility and human rights (Wettstein et al., 2019). However, the concept of CSR is surrounded by ambiguity (Matten & Moon, 2008), which leads to different approaches and strategies towards CSR within MNE’s (van Marrewjik, 2003). This is where CSR rankings, as mechanism of surveillance (Sauder & Espeland, 2009) come into play. Rankings inherently generate reactivity, a response resultant from different factors (Espeland & Sauder, 2007), which can lead companies to improve their CSR (Chatterji & Toffel, 2010). However, responses can vary within the same rating system (Crilly et al., 2012), which raises the question as to why they vary. An answer that seems plausible given the ambiguity surrounding CSR (Matten & Moon, 2008) and the configurational nature of reactivity (Espeland & Sauder, 2007; Slager 2020), is the difference within institutional environments (Hall & Soskice, 2001) as it leads to different CSR approaches (Matten & Moon, 2008). LMEs, associated with explicit CSR seem to have higher CSR performances when compared to CMEs, associated with implicit CSR (Jackson & Apostolakou, 2010). This distinction between approaches (explicit/implicit), based on differences within economies (Hall & Soskice, 2001, Matten & Moon, 2008), is what we then argue to lead to distinct responses to CSR rankings such as the CHRB. This research then becomes of relevance, as while we see how these institutional environments play a role in CSR, we are still to understand how they reflect upon reactivity within CSR rankings, given the lack of research in this regard.

The understanding of these two perspectives, reactivity (Espeland & Sauder, 2007) and differences in institutional environments (Hall & Soskice, 2001), together with the understanding that reactivity in itself can be seen as configurational (Slager et al., 2020), bring us to our theoretical gap in which we seek to understand how such institutional effects reflect upon the response of MNEs to social rankings, which then lead us to our proposition:

Proposition: The combinations of factors that lead to the presence or absence of reactivity will

(20)

19

3. RESEARCH METHODOLOGY

To provide a clear view and understanding of the methodology used, in this section we present the case context selected (the CHRB), followed by an introduction on configurational approach and Qualitative Comparative Analysis (QCA). We then present the selected conditions, their calibration and the four sequential steps of a QCA.

We rely on the understanding that reactivity can be configurational (Espeland & Sauder, 2007; Slager et al., 2020) and therefore resultant of various factors, to explore how a difference in institutional environments (LME vs CME) (Hall & Soskice, 2001) can lead to different responses (Explicit vs Implicit CSR) (Matten & Moon, 2008) from MNE’s within CSR rankings.

The focal case within this study, the CHRB, provides a detailed benchmark based on human rights standards that ranks MNE’s accordingly. This benchmark is set by various indicators and metrics, amongst which the United Nations Guiding Principles on Business and Human Rights (UNGP) and the Sustainable Development Goals (SDGs).

By using this benchmark’s 2019 data together with a configurational analysis, through a QCA, this research intends to explore whether differences in institutional environments do impact the reactivity of the MNE’s within the CHRB.

3.1 The CHRB – Case Context

The Corporate Human Rights Benchmark, initially launched in 2013, is a multi-stakeholder initiative that draws on investors, business and human rights, and benchmarking expertise from six organisations, APG Asset Management (APG), Aviva Investors, Business & Human Rights Resource Centre, The EIRIS Foundation, The Institute for Human Rights and Business(IHRB), Nordea Wealth Management. The CHRB was created as an open and public benchmark of corporate human rights performance, set by various indicators and metrics (UNGP, SDG, etc.), with the intent to compel change from those within its scope through the competitive nature of the market.

(21)

20 industries (Agricultural Products, Apparel and Extractives), while the current 2019 ranking counts with 195 companies; 95 of which were added between 2018 and 2019 allowing for the addition of a new industry (ICT Manufacturing).

Despite the available data and its numerical nature, the CHRB as a benchmark lacks sufficient information to pursue a quantitative research given the small number of companies under scrutiny. On the other hand, a purely qualitative approach may be seen as insufficient, given the available information provided by the CHRB. Given the structure and nature of this research, we use a configurational approach combined with a qualitative to analyse the CHRB data, given the size of the sample and the configurational nature of reactivity (Slager et al., 2020).

3.2 Data Collection

The data intended to support this study is mainly obtained from the CHRB, in which the CHRB publishes the results of its benchmark on companies yearly. This research then uses the data made available by the CHRB as the main support to further explore the reactivity of companies towards a CSR ranking.

The CHRB counts with published results for the years of 2017, 2018 and 2019, in which there are respectively 98, 100 and 195 results regarding the scores on a set of human rights indicators (see Appendix D: 59) of the largest publicly traded companies in the world (CHRB website). However, for the purpose of this analysis, the research only considers the results for the year of 2019. While a multiple-year analysis would be better suited when considering reactivity (Espeland & Sauder, 2007), as it would provide further insight on the evolution of MNE’s scores, given the changes in the methodology of the CHRB benchmark in 2016 and 2017, it makes sense to solely consider 2019 score. In order to analyse all three years, it would be required to align the metrics used in each specific year in order to correctly establish a score that would be equally measured across the three years, however, given the nature and context of this study, the research proceeds with the data from 2019 as it still counts with considerable information that is suitable for the research at hand.

(22)

21 In addition to the 2019 CHRB data, annual reports and corporate websites were used to extract additional information on certain companies in order to complement the information provided by the CHRB.

3.3 Configurational approach

We rely on the understanding that reactivity can be configurational (Espeland & Sauder, 2007; Slager et al., 2020) and adopt a configurational approach. This approach is adopted to analyse the data made available by the CHRB and explore the possible configurations that lead to our outcome of study, reactivity (Espeland & Sauder, 2007).

As presented by Misangyi et al. (2017), the usage of such an approach “enables a fine-grained conceptualization and empirical investigation of causal complexity through the logic of set theory.” (Misangyi et al., 2017: 255).

The use of such an approach allows us to explore causal complexity and conjunctural causation, in which various causal attributes (conditions) combine and form different and unique configurations that lead to an outcome of interest (Misangyi et al., 2017). A configuration, therefore, is a set of characteristics of the case/cases under study.

A configurational approach is built upon two bases, set theory and QCA. Set theory “uses Boolean algebra to determine which combinations of organizational characteristics combine to result in the outcome in question.” (Fiss, 2007: 1183). In our case, this is reflected in the combination of conditions that are expected to affect the outcome, reactivity (Espeland & Sauder, 2007). Moreover, given the focus of this study, special attention is given to configurations that count with the presence/absence of economy towards the outcome, as it directly resonates with our research question. Similarly, to set theory, QCA also relies on Boolean Algebra, but more importantly, QCA is designed to assess conjunctural causation empirically (Fiss, 2007; Misangyi et al., 2017). QCA is thus able to distinguish between the different combinations of attributes, the set conditions, and allows for a distinction of those that lead to the outcome of interest.

(23)

22 study is inherently configurational: various explanatory factors and their interplay simultaneously determine the outcome(s) studied” (Fainshmidt, 2020: 455), and it is for this same reason that studies within a quantitative scope can be perceived as overly simplistic as a result of the reliance on hypothesis that assume linear effects (Fainshmidt, 2020).

Whilst the analysis of reactivity could be done in a quantitative matter, as pointed by Espeland and Sauder (2007) there are various organisational factors that contribute to it, the geographical factor being one of them. As such, a configurational approach that considers various conditions and their combinations towards an outcome becomes more relevant when studying a concept such as reactivity (Slager et al. 2020).

The usage of a configurational approach is beneficial as it falls between both quantitative and qualitative research, allowing for the analysis of more complex IB phenomena (Fainshmidt, 2020). Moreover, this approach enables the analysis of medium-sized samples, that are considered too large for qualitative analysis, and yet too small for a quantitative approach (Fainshmidt, 2020).

Despite the understanding that this type of approach is still relatively novel, many studies with empirical and theoretical validity have relied on such an approach to explore concepts and ideas (Fainshmidt, 2020). For instance, within the context of national institutions, Witt and Jackson (2016) provided evidence on how a combination of both LME and CME can lead to a comparative advantage in high-tech industries. This is just one example of many that can be found within the work of Fainshmidt (2020).

The usage of a configurational approach is therefore relevant for this study as it allows us to explore the configurational nature of the concept of reactivity (Slager et al., 2020), and through the analysis of the possible configurations, we can then explore the extent to which differences within an institutional environment, here depicted as an LME and CME, affect reactivity (Espeland & Sauder, 2007).

3.4 Conditions and Calibration of the QCA

(24)

23 number of conditions leads to an exponential growth of possible combinations, which in hand complicates the interpretation of the results (Hill et al., 2019). As per QCA conventions (Greckhamer et al., 2018), the conditions described below, where selected taking into consideration a combination of both theory and case knowledge. In addition, this research expects the presence of all conditions to contribute to reactivity, as supported by both literature and case knowledge.

While the QCA analysis for the research is set as a fuzzy set Qualitative Comparative Analysis (fsQCA) given the non-crisp (non-binary) nature of the selected outcome, all selected conditions are defined on a crisp (binary) (see table 1: 28) basis. As such, the selected conditions reflect an unequivocally “presence” or “absence” in causal pathways (Hill et al., 2019). However, the usage of this approach can be perceived as a double-edged sword, while on the one hand, the usage of crisp conditions allows for an easier interpretation of the results, on the other, the oversimplification of the information present in the selected conditions can reflect on a loss of information. Nonetheless, using crisp conditions allows for a more simplified analysis of the conditions, that can thus more easily reflect the desired study of the influence of different institutional environments on the proposed outcome.

This research thus proposes one outcome and five possible explanatory conditions for the QCA analysis.

3.4.1 Outcome

(25)

24 and that the majority of those scores are low (see appendix B and C: 54-58), which led us to set the starting “in set” scores at 40% (see table 1: 28), that while not ideal, allow for an analysis of the level of response.

3.4.2 Explanatory Conditions

This analysis counts with five conditions. The economy condition captures the difference in institutional environments. The old condition captures pre-existence (preciously ranked) with the benchmark. Serious allegations captures exposure. Engagement captures a level of commitment with the CHRB, and Human rights due diligence captures a level of engagement with human rights.

Economy measures the market type of the MNE’s home country. This condition is by far the

most important to this study, as it is the one that allows for an institutional level effect analysis. This condition reflects whether a company’s home country is deemed as LME or CME as per the literature on VoC (Hall & Soskice, 2001). LME was set as the “in set” result, as it is our understanding that an LME is more prominent to competition within its market (Hall & Soskice, 2001), which in hand should lead MNE’s to more actively compete amongst themselves. In the sense that an LME company, leading in human rights (high CHRB score), will likely compel other companies from the same country, to “catch up”. The competitive environment expected of an LME is then a driver for a better CSR score. Jackson and Apostolakou’s (2010) work provide support to this argument, as they found that LME’s tend to score higher across various CSR dimensions when compared to CME’s.

The process by which the distinction of CME and LME (Hall & Soskice, 2001) was based, was a result of, existing literature indicating a country’s proneness to either LME or CME, together with literature that provided evidence of a country’s proneness to either explicit or implicit CSR, given the existing association between LME/explicit and CME implicit (Matten & Moon, 2008). For a more detailed view on the companies and countries set as LME and CME, revert to appendix B and C (p. 54-58).

Old. This condition is set on a combination of both case knowledge and literature. The

(26)

25 clear that the majority of companies previously ranked scored better when compared to new companies. Moreover, when looking to the data from previous years, it became evident that the majority of companies ranked by the CHRB through the years add a positive evolution of their scores, potentially caused by a learning effect.

When considering this information together with the literature at hand, and especially the literature on reactivity (Espeland &Sauder, 2007), it is then evident that the continuous presence on a ranking should lead to a greater level of response (high score), which then reflect upon those who were not previously ranked. As expressed in Espeland and Sauder’s (2007) work, law schools adapted to the ranking as it became more and more embedded in universities day to day lives. In that same regard, the same can be proposed for companies within the CHRB. We then set the presence of old (previously ranked) as having a positive effect towards the outcome.

Serious Allegations. Companies that are under public scrutiny are more likely to act according

to stakeholders demands in order to gain or maintain legitimacy (Rathert, 2016). A company’s CSR is but one additional form by which companies can try to achieve said legitimacy (Rathert, 2016). As such, companies under the public eye are more likely to actively engage in CSR practices. By following this rationale, we defined serious allegations as a condition. The presence of one or more serious allegations towards a MNE, should reflect on a greater level of awareness of the MNE at a CSR level, given the overall exposure to stakeholder. As such we expect the presence of serious allegations to contribute positively towards the outcome.

While the metrics that define whether an allegation is included or not in the CHRB data are very clear, no clear distinction is made on regards to the degree of the serious allegations considered, which is in part why this condition is presented as a crisp condition. The allegations reflected within the CHRB data are based on allegations from external sources (NGOs, news sites, governmental agencies, etc.), however, only those covered by Vigeo Eiris, BHRRC and RepRisk are considered (CHRB Methodology, 2020). The serious allegations within the CHRB are categorized into nine categories (see table 12: 60, for both categories and examples of allegations considered by the CHRB).

Engagement is a condition that reflects the distinction between companies that engage with the

(27)

26 the engagement assessment comprises of responses that fall in or outside of the formal assessment process of response to a letter sent by the CHRB. This condition reflects a level of commitment towards the CHRB, and allows for a distinction between companies that engaged with the CHRB and companies that fail, or choose not to respond. Given the available information provided by the CHRB, we are only able to make a simple distinction between companies that do and do not engage with the benchmark. A further analysis on the level or degree of engagement isn’t possible. We set the presence of this condition as positive towards the outcome, as a result of both case knowledge and theoretical support on a positive effect towards an existing level of engagement (Rowley et al., 2017; Pollock et al., 2018).

Human rights due diligence. Companies actively involved in human rights have certain

systems in place to ensure that human rights are taken into consideration. This condition measures MNEs presence of systems to ensure that human rights due diligence processes are in place. It distinguishes companies that have no level of human rights due diligence in place, from those that do (see table 1: 28). The selection of this condition is based on case knowledge and theoretical support (Martin-Ortega, 2013). While looking at the data, we see an indication of poor scores for companies that fail to have human rights due diligences in place. We then set the presence of human rights due diligence as the “in set” result that positively contributes to the outcome.

(28)

28 Table 1– Calibration Table

Name Description Data Source Calibration

Outcome Reactivity Benchmark score from the CHRB 2019 evaluation.

CHRB 2019 Data The final 2019 scores of the CHRB were converted into scores from 0 (no evidence) to 4 (advanced). Set membership was calibrated based on quartiles. The top quartile represents full membership (1) in the set of firms that are reactive, and the bottom quartile represents full non-membership (0) in the set of firms that are reactive. The second and third quartiles represent partial membership (0.33, 0.67). The division of the CHRB final scores was set from 0% to 9.9% (0), from 10% to 19.9% (0.33), from 20% to 39.9% (0.67), and finally from 40% onwards (1).

Engagement The company is formally or informally engaged with the CHRB.

CHRB 2019 Data Out of the set (0): Comapny has no form of engagement with the CHRB. Fully in the set (1): Company engaged formally or informally with the CHRB. Economy Distinction between LME and CME

based on country of origin of the MNE.

CHRB 2019 Data, VoC (Hall & Soskice, 2001) and additional literature

Out of the set (0): Company's country of origin set as a Coordinated market economy (CME). Fully in the set (1): Company's country of origin set as a Liberal market economy (LME).

Old Company that has been part of the CHRB in previous years, and has subsequently been ranked before.

CHRB 2019 Data Out of the set (0): Company that is not part of the previous CHRB benchmarks, 2018 and/or 2017. Fully in the set (1): Company that is part of previous CHRB benchmarks, 2018 and/or 2017.

Serious Allegations

Existence of one or more serious allegations associated to an MNE.

CHRB 2019 Data Out of the set (0): CHRB did not report or evaluate any type of allegations regarding the company. Fully in the set (1): CHRB reported or evaluate at least one allegations regarding the company.

Human Rights Due Diligence

MNE's presence of system to ensure that human rights due diligence processes are in place.

CHRB 2019 Data MNE with no system or mechanism (due diligence) regarding human rights, scores 0 on indicator B.2; wich brings us to calibrate the condition as follows:

Out of the set (0): Company scores 0 on Human Rights Due Diligence (B.2) Fully in the set (1): Company scores above 0 on Human Rights Due Diligence (B.2)

(29)

29

3.5 QCA Process

In line with QCA conventions, conditions can be derived from theory as well as case contextual knowledge (Greckhamer et al., 2018), which led us to the first stage of our data analysis. This first stage allowed for an initial analysis of the existing data that led us to the conceptualization of our five conditions, human rights due diligence, serious allegations, old, engagement and economy.

The second stage of the analysis involved the usage of the QCA method. Given the initial interest of using a fsQCA to explore causal complexity, an assessment of the possible conditions had already been performed in the first stage, derived from a combination of in-depth knowledge of the case (the CHRB) and a reliance on the theoretical background. As such, in the second stage, after calibrating the selected conditions, we proceeded to use the QCA software 3.0 in order to perform a QCA analysis. This process involved the application of the four sequential tasks fundamental for QCA (Scarpi et al., 2018), here presented in order to provide a better insight into the QCA process.

• The first step is to confirm whether or not necessary conditions are present, necessary conditions being conditions that are present whenever the outcome is present. There are no necessary conditions present as all conditions fall below the 0.9 benchmark (Greckhamer et al., 2018) (see table 5: 53).

• The second step is to construct a truth table; for this research, we have used the fsQCA algorithm in the QCA software to do so. The truth tables (see tables 3 and 4: 51, 52) include each possible combination of conditions.

(30)

30 indicators range from 0 to 1, we have set 1 as the frequency threshold, and 0.8 as the consistency threshold, in line with recommended levels (Ragin, 2008).

• The fourth and final step is to reduce once more the truth table into simplified combinations (Fiss, 2011). The QCA algorithm derives solutions that differ in complexity. The parsimonious solution derives configurations which are strongly simplified using Boolean algebra, whilst the intermediate solution takes into account only the most plausible simplifying assumptions (Ragin, 2008). This research, in line with QCA standards, uses the parsimonious and intermediate solutions to present core conditions (conditions solely present in the parsimonious solution), and peripheral conditions (conditions present in both parsimonious and intermediate solutions). The results from step 4 of the QCA analysis, allowed us to distinguish the relevant configurations from those that were not represented by the sample, commonly known as logical remainders (Hill et al., 2019). Following step 4, we were then able to obtain all relevant configurations that are represented by the sample and reached the final results table (see table 2: 31).

The third and final stage reflects an in-depth analysis of the resulting configurations from the QCA (see table 2: 31). Both solutions (presence and absence of outcome) are deemed strong given the level of consistency being above 0.8 (Rihoux & Ragin, 2009), which in turn indicates that a substantial proportion of the outcome is covered by the configurations. With regards to coverage, the solution with a presence (absence) of the outcome reflects 39% (42%) of all present (absence) outcomes.

At this stage, we interpreted the resulting configurations and used the existing theoretical framework to analyse both the configurations and the cases assigned to each of them. More importantly, we explored whether these configurations contribute in explaining the initially proposed effect of differences of institutional environments on reactivity.

4.

FINDINGS

(31)

31 equifinality, the notion that an outcome can be reached from different initial conditions and trough different paths (Fiss, 2011). While black circles indicate the presence of a condition, circle with an / inside reflect the absence of a condition. The blank spaces within each configuration represents a “don’t care” scenario for that condition, in which the condition can be either present or absent (Fiss, 2011).

Table 2 – Outcome of the QCA Analysis

Configuration 1 describes the presence of engagement, economy, serious allegations and

human rights due diligences, as core conditions for reactivity to be present. This reflects through the CHRB as an MNE that, is actively engaged with the CHRB, has an LME as a home country, has the presence of one or more serious allegations under scrutiny, and has at least a minimum level of human rights due diligences in place. The combination of these factors then leads to a presence of outcome, here measured as a positive score (see table 1: 28).

1 2 3 4 5

Engagement

• •

Economy

Old

Serious allegations

• •

Human rights due

(32)

32

Configuration 2 follows a very similar pathway to configuration 1, it counts with the presence

of engagement, old, serious allegations and human rights due diligences, as core conditions for reactivity to be present. However, configuration 2 has the distinction that instead of counting with the presence of economy (LME), it disregards this condition and counts with the presence of old, here reflected as a MNE’s that has been previously ranked by the CHRB.

Configuration 1 and Configuration 2, which reflect a presence of the outcome, are quite similar

when considering the core conditions that form them. However, these two configurations are distinct in a very interesting way. While both share three core conditions, they differ in regard to the conditions economy and old. Configuration 1 disregards old, as a “don’t care” condition, and configuration 2 indicates the exact opposite.

Given the similarity of these configurations, what becomes of interest to understand, is that these configurations point to the importance of the conditions old and economy when considering the outcome. As these are the conditions that in a sense, create equifinality when considering the presence of outcome. More importantly, this distinction between economy and old is extremely relevant as it points to the notion that the condition economy is what mitigates the absence of being previously ranked, for the outcome to be present, and the same applies for the condition old, as its presence mitigates the absence of a LME. All of this, while considering the presence of all other three conditions for both configurations.

Furthermore, both configurations present a high level of consistency, implying a level of relevance of these sets of conditions when considering the outcome.

Configuration 3 represents cases in which there is, an absence of engagement with the CHRB,

no presence of serious allegations under knowledge of the CHRB, and finally a complete lack of any form of due diligences in regard to human rights according to the CHRB. These cases represent an absence of outcome and are thus a reflection of poor/low scores. This configuration also reflects the importance of these conditions for the CHRB as it has a high level of consistency (92%) and has the largest number of cases.

Configuration 4 indicates the absence of economy, old and serious allegations as core

(33)

33

Configuration 5 indicates a third and final path to low benchmark scores, however, contrary to

the previous two configurations that solely rely on the absence of conditions, this third configuration counts with the presence of old (previously ranked by CHRB) as a core condition. Configuration 5 is then represented in the CHRB, as an absence of engagement (formal or informal) with the CHRB, a CME country of origin, previously ranked, and with a complete absence of any level of human rights due diligence. In addition to the distinction of a present core condition, this configuration also reflects the lower number of cases, contributing the least for the overall coverage of the overall solution.

These causal configurations of absence of outcome thus reflect a low overall score at the CHRB. But more importantly, similarly to the configurations on the presence of outcome, these causal configurations represent additional relevance when considering the “bigger picture”. For instance configurations 4 and 5 reflect absence of economy as a core condition, which in hand refers to a CME, and thus brings the notion that even when considered with the other core conditions, the fact is that a CME plays a role in leading to a lack of outcome, a poor score in this case.

Furthermore, it is important to note that the selected conditions seem to point to causal symmetry, where their presence points to the outcome and their absence to the lack thereof. Whilst we do not find a configuration that presents causal symmetry, the selected conditions provide indication of their overall importance as explanatory factors for our outcome of choice. As when observing the configurations, both configurations 3 and 5 share three conditions with both configurations 1 and 2.

(34)

34

5.

DISCUSSION

In this study, the broader question we pursue is to understand how differences across institutional environments affect reactivity, more specifically within a CSR context given its growth and ambiguity (McWilliams & Siegel, 2001; Dahlsrud, 2008). By relying on the understanding that reactivity can be configurational (Espeland & Sauder, 2007; Slager et al., 2020), we explore this question by analysing the CHRB through the use of a QCA. This configurational approach provides us with combinations of conditions, that produce distinct levels of the outcome, reactivity (Espeland & Suader, 2007)), and provide nuance to it. While our use of a QCA analysis complements and adds to the existing framework of causal configurations through the use of a fsQCA, our resulting configurations allow us to go beyond a rather simplified view of a reactive response (Pollock et al., 2018), and enable a more complex analysis of the different conditions that come into play when considering the outcome of study. As such, it is important to explore and explain the implications these configurations point to, and what that represents to both our study and theory.

An overall notion we can take from the research, is the understanding that reactivity can be configurational (Espeland & Sauder, 2007; Slager et al., 2020). The five resulting configurations, and the combination of core conditions within them, point to the presence or absence of outcome. Which in turn contributes to idea presented by Slager et al. (2020) on the potential configurational nature of reactivity and how it can be explained by combination of multiple factors.

Whereas this reasoning is important, as it contributes to the configurational understanding of reactivity. There are three illations that we need to reflect upon, resultant from our theoretical understanding, combined with the five configurations and combination of conditions within them.

(35)

35 serious allegations in our view, adds to the argument presented by Rathert (2016), in which MNE’s seem to more commonly use CSR as a mechanism to achieve legitimacy. As such, the presence of serious allegations implies a level of accountability from stakeholders, which then leads MNEs to address such issues in order to retain legitimacy, given the link between CSR and legitimacy (Rathert, 2016). Our rationale being that once a MNE is faced with a serious allegation, it then inevitably becomes aware of the problem. To retain legitimacy the company has to address it, which implies gaining knowledge over the issue. While an absence of allegations simply won’t create such a chain of events. Secondly, the need for a presence of human rights due diligence, lines with the notion that establishing human rights due diligence processes and addressing human rights with specific policies within CSR, can be beneficial for MNE’s (McCorquodale et al., 2017; Obara & Peattie 2017). Finally, the importance of engagement towards the outcome, supports the work of Pollock et al. (2018) and Rowley et al. (2017), where they explain how such interactions (engagement) help negotiate reactivity towards the specific metrics.

However, it is important to note that, when considering these three core conditions presence towards the presence of outcome, we must also account for the need of a presence of the conditions economy and old. As the presence of these conditions in configurations 1 and 2 respectively, builds upon existing research that would expect more pressure (serious allegations), capacity (human rights due diligence), and dialogue (engagement) with the metric provider, to impact reactivity (Slager et al., 2020). The presence of these two conditions, builds upon this notion, by adding that the outcome can also be influenced by economy and old, here reflected as the presence of an LME, and a MNE previously ranked by the CHRB.

(36)

36 Apostolakou, 2010), therefore indicating how an institutional effect (Matten & Moon, 2008; Ioannou & Serafeim, 2012) can influence reactivity.

Furthermore, we also need to consider the implications of the potential interchangeable effect of these two conditions towards the outcome, as both configurations only differ in this regard. The “don’t care” aspect of economy and old within each configuration (1 and 2), could indicate that companies that have been previously ranked, benefit through what we posit as a learning effect, which then enables them to mitigate an institutional level effect (LME vs CME). In that same regard, the positive institutional effect of a LME, mitigates the need for being previously ranked to achieve a positive outcome.

The second important illation we take from the resulting configurations, is the overall presence or absence of the condition economy throughout three configurations, but more importantly, the absence of economy in configurations 4 and 5.

The presence/absence of economy is important, as it helps us understand and answer our research question. Economy, by being present in the configurations, is part of the combination of factors that lead to the presence/absence of the outcome, which then enables us to argue its effect upon reactivity. This presence/absence contributes to the existing theoretical framework by providing evidence of an institutional effect (Hall & Soskice, 2001), in reactivity (Espeland & Sauder, 2007). While we understood that an institutional level effect played a role in CSR (Matten & Moon, 2008), we can now provide evidence of said effect in reactivity, and therefore contribute to the literature of reactivity (Espeland & Sauder, 2007) and institutional environment differences (Hall & Soskice, 2001; Matten & Moon, 2008; Ioannou & Serafeim, 2012).

(37)

37 The presence and absence of economy through configurations 1, 4 and 5, indicates a degree of influence from both LME and CME. Whereas the first reflects a positive outcome, the second associates itself with a poor outcome. This indication therefore supports the theory of Varieties of Capitalism (Hall & Soskice, 2001), as it further provides evidence on the distinction between economies. However, the presence and absence of this condition has deeper implications, which brings us to the third illation.

While the second illation contributes to the understanding that differences in institutional environments (Hall & Soskice, 2001) affect reactivity, within a CSR context, we are still to explore how these differences affect reactivity. Which brings us to the third illation. Here, is where the understanding between the association of LME with explicit CSR and CME with implicit CSR (Matten & Moon, 2008), comes into play. However, we must first acknowledge the results indication, as they infer that a LME is better for the outcome (higher score in the CHRB), whereas a CME is actually detrimental. This then implies that a MNE from a LME, would actually score better in a CSR ranking as a result of their institutional environment, which allows us to build an argument as to why an explicit approach (LME) is better than an implicit approach (CME), in public CSR rankings.

What we argue as to why one is better than the other, in this context, is the distinctions within their inherent institutional environment. While in LME’s, companies tend to be more reliant on the market, in CME’s, companies coordination tends to be more strategic (Hall & Soskice, 2001). As a result of this distinction, companies within LME’s are more likely to be competitive by nature, as they are accustomed to a market regulation in which companies act more independently (Hall & Soskice, 2001). Where LME’s tend to have more explicit CSR, in which there is a greater degree of openness towards CSR and CSR practices, companies are more open and vocal about their CSR, on the other hand, CME’s are more linked to an implicit CSR approach, in which companies abide by CSR standards, but are however more discrete about their practices (Matten & Moon, 2008).

(38)

38 standards of CSR as a result of the surrounding competition in its environment. On the other hand, a CME company is less reliant on competition to dictate their CSR engagement, and given their implicit nature is less likely prepared to attend to public standards of entities such as the CHRB. We therefore build upon Matten and Moon’s (2008) work, by adding the idea that an explicit approach towards CSR, which stems from a LME’s principals, can be beneficial towards public CSR rankings such as the CHRB. This argument, together with our findings, then provides a reasoning for both how and why, a difference in institutional environments can affect reactivity.

Whereas these illations rely on both theory and resulting configurations, there are there are still two illations to be added that are based on the theoretical framework and case knowledge. The first regards the CHRB. Given the argument presented above, that companies from a LME which have an explicit approach to CSR, are better when considering public CSR rankings, as a result of their “explicitness” or openness to CSR (Matten & Moon, 2008). What then becomes interesting to explore, is the extent to which rankings such as the CHRB influence an MNE’s approach to CSR. This indication that MNE’s from LME’s can perform better as a result of their explicit approach to CSR, then brings room to argue that the CHRB can be assessing and pushing towards an explicit CSR “agenda”, given the positive association with LME’s and CSR (Jackson & Apostolakou, 2010).

What we then look to argue upon, is that CSR rankings such as the CHRB, can be leading MNE’s from CME’s to engage in the process of implicitization (Matten & Moon, 2020). By pushing what we name as an explicit “agenda”, through metrics and standards that favour a more open engagement. MNE’s, by abiding to their metrics, having human rights due diligence, actively engaging with them, or addressing serious allegations, as seen in

configurations 1 and 2, they perform better in the ranking. Whilst we don’t further explore this

idea, it seems interesting to consider, as it lines with the use of rankings as a mechanism of surveillance and control (Espeland & Stevens, 2008). By pushing this explicit “agenda”, CSR rankings such as the CHRB, can be leading MNE’s with implicit approaches to CSR towards adding explicit CSR aspects, therefore leading them to implicitization (Matten & Moon, 2020).

(39)

39 we see that they rank very low comparatively to the companies previously ranked, and that their scores fall in line with those of the companies initially ranked in 2017 (CHRB key findings, 2019). This then points to the understanding that the CHRB is in fact generating reactivity (Espeland & Sauder, 2007). But, more importantly, these low scores of the majority of MNE’s only corroborate the argument presented by Wettstein et al. (2019), as this is a clear indication of the lack of engagement from MNEs towards human rights. Finally, these low scores, especially from the newly added companies, point to the necessity of rankings such as the CHRB, that provide an added mechanism that can help improve human rights engagement within MNE’s.

The pursue of a possible answer for our research question, through the use of a QCA analysis, allowed us to complement the existing theory on reactivity (Espeland & Sauder, 2007), institutional environment differences (Hall & Soskice, 2001; Ioannou & Serafeim, 2012), and more importantly institutional environment differences within CSR (Matten & Moon, 2008; 2020). The use of a configurational approach allowed for a more nuanced analysis of the outcome of study, reactivity (Espeland & Sauder, 2007), and provided us with five distinct combinations of conditions. Our findings have thus enabled us to contribute to the existing theory on reactivity and institutional environment differences, within a CSR context, by providing evidence of how the presence and/or absence of the selected conditions contribute to the outcome of study. But more importantly, it provides a rationale as to how this distinction (LME vs CME) in institutional environments, affects reactivity. Furthermore, the study contributes to the work of Matten and Moon (2008), by providing an argument on the LME/explicit CSR association as a possible explanation for a better CSR performance when in the context of CSR rankings. This paper then corroborates the argument presented by Jackson and Apostolakou (2010) and adds to it by indicating how LME/explicit CSR can lead to better performance within CSR rankings such as the CHRB.

(40)

Referenties

GERELATEERDE DOCUMENTEN

- Door zijn uiterlijk moet de brommobiel feitelijk als een 'soort personenauto' worden beschouwd en niet als een 'soort bromfiets'; De brommobiel mag niet op het

To make our exposition more concise we sometimes read AG:x ϕ as ‘the agents in G accept that ϕ while functioning together as members of institution x’.2 For example, AG 1

The institutional environment of Spain, considered as a country with a high regulative, normative and cognitive distance in comparing with the Netherlands, is with

In this paper, I find that the share of institutional ownership has a positive influence on acquisition returns therefore supporting the theory that

The regressors include are total cash divided by total assets (cash), total debt divided by total assets (lev), research and development expenses divided by total assets

relationship between the (lagged) Size of firm, Financial return, and Tobin’s Q control variables and CSR decoupling indicate a highly significant relationship with regards to

Pressure resistant investors were expected to have a significant positive influence on CSR activities, because contrary to pressure sensitive investors, they do not have

In the following section the PCA is described and conducted on the political variables and credit variable in order to reflect the influence of the political state of a country on the