University of Groningen
Tensions in sustainable supply chain management: instrumental, institutional, and paradoxical perspectives
Xiao, Chengyong
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Xiao, C. (2019). Tensions in sustainable supply chain management: instrumental, institutional, and paradoxical perspectives. University of Groningen, SOM research school.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
TENSIONS IN
SUSTAINABLE SUPPLY
CHAIN MANAGEMENT
i
Publisher: University of Groningen Groningen, The Netherlands
Printed by: Printpartners Ipskamp B.V. ISBN: 978-94-034-1330-3 eISBN: 978-94-034-1329-7 Layout: Chengyong Xiao Copyright 2019© Chengyong Xiao
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system of any nature, or transmitted in any form or by any means, electronic, mechanical, now known or hereafter invented, including photocopying and recording, without prior written permission of the publisher.
ii
Tensions in sustainable supply
chain management:
instrumental, institutional, and
paradoxical perspectives
PhD thesis
to obtain the degree of PhD at the University of Groningen
on the authority of the Rector Magnificus prof. E. Sterken
and in accordance with
the decision by the College of Deans. This thesis will be defended in public on Monday 18 February 2019 at 16.15 hours
by
Chengyong Xiao
born on 25 June 1987iii
Supervisors
Prof. D.P. van Donk Prof. J.T. van der Vaart
Co-supervisor
Dr. M.M. WilhelmAssessment Committee
Prof. J. Surroca Prof. M. Pagell Prof. C. Busseiv
谨以此书献给我的母亲 陈寿珍女士 她一直把我视为她的骄傲
v
Acknowledgement
Completion of this thesis was a challenging, mostly exciting, inevitably frustrating, and eventually rewarding journey. It was a journey of six years, during which I experienced many ups and downs. More accurately, several overwhelming “downs” preceded several thrilling “ups”, which was succeeded by another overwhelming “down”. For me, this journey is a perfect illustration of one of my mottos used in the high-school period: no pains, no gains; although I am not completely convinced that my gains are sweet enough to justify all the bitter pains that agonized me in the past six years. By no means do I believe in hedonism, but I do believe in the value of having a balanced life, in which gains can somehow outweigh pains. Finally, having scrambled out of this suffocating tunnel, I would like to express my sincere gratitude to the people and organizations that helped me, directly and/or indirectly, for the completion of this thesis.
First and foremost, I would like to express my sincere gratitude to my supervision team: Prof. Dr. Dirk Pieter van Donk, Prof. Dr. Taco van der Vaart, and Dr. Miriam Wilhelm for their continuous support, motivation, and immense knowledge. I first met Taco back in the autumn of 2011, during the “job interview”. A Chinese proverb says that
千裡馬常有，而伯樂不常有 (Swift horses are usually found but not the same as ‘Pak Lok’, the person who has good judgment to spot swift horses). Taco has been my Pak Lok, who has shown consistent support from the very beginning, although I am far from being a “swift horse”. All through these years, his immense knowledge, network skills, and negotiation skills have greatly contributed to the completion of all three projects. Thank you, Taco, for introducing me to our case company and introducing me to a broader academic network. Taco and Dirk Pieter formed my supervision team in September of 2012, when I officially kicked off my PhD project in Groningen. Many thanks go to Dirk Pieter for his substantial support which came in many forms: detailed, critical, and constructive comments, which helped me to improve my academic writing skills substantially; expansive thinking, which provided gleams of sun shines when my knowledge and creativity seemed to have failed me; high academic standards, which prevented the papers from being sent out prematurely, etc. Miriam started to act as my daily supervisor shortly before the summer of 2013, which further increased the strength of the supervision team. Her immense knowledge, skillful guidance, and consistent support have contributed significantly to the completion of my thesis. Miriam and I have
vi
been working on four projects, but only one of them is included in this thesis. As such, this thesis is just a partial reflection of Miriam’s contribution to my academic improvement. Without her participation, my PhD journey would have been very different. I feel extremely lucky and gratitude to have such a wonderful supervision team.
Second, many thanks go to the SOM, which has provided the logistical support for completing this thesis. Ellen is always ready to help: course registrations, travel application and reimbursements, etc. Arthur has prepared more than a dozen of letters for me: Visa applications, ID extensions, conference registrations, etc. Rina has been so generous that I was always given adequate travelling budgets to attend one or sometimes even two international conferences each year. Special thanks go to Rina, Taco, and Dirk Pieter for extending my scholarship for a period of 6 months in 2016. Last but not least, many thanks go to the PhD coordinators, Linda, Justin, Jasper, and Kristian, for their kind support at different phases of this journey. The secretaries of Operations Department, Durkje, Marjo, Renny, Linda, Ellen, Heleen, and Irene, also provided valuable support during the past years.
One of the most exciting, enjoyable, and memorable parts of this journey was the field study for data collection. I was very lucky to have received significant support from our case company COSMOS. Many thanks go to Emile, Zhihai, Steven, Ellen, Jenny, Amy, Carl, Tommy, Ann, Tao Jian, and Yolanda. With their support, I could have access to the ten suppliers and did comprehensive and interesting interviews with more than 60 interviewees. Many thanks go to all my interviewees from COSMOS and the ten suppliers for their time, hospitality, and interesting and insightful stories.
Special thanks go to my assessment committee, Prof. Dr. Christian Busse (University of Oldenburg), Prof. Dr. Mark Pagell (University College Dublin), and Prof. Dr. Jordi Surroca (University of Groningen), for their time and efforts put in assessing my thesis. Thank you all for your positive assessment and constructive comments.
Besides these people and organizations, I would also like to express my sincere gratitude to several groups of people that have helped me in many ways in the past six years. First, my sincere thanks go to my UCAS Group: Yanping Zhao (赵艳平), Xin Fan (范昕), Yuwan Duan (段玉婉), Quanrun Chen (陈全润), Xuan Zhang (张璇), Ye Liu (刘
vii
晔), Suxiao Li (李苏骁), Binqian Yan (闫冰倩), Qiubin Huang (黄秋彬) & Jenny (翁健莉
), Yan Yan (闫岩) & Jingjing Zhang (张晶晶), Jiyuan Wang (王吉元) & Ya Gao (高雅), Kailan Tian (田开兰), Huan Liu (刘欢), and Chenming Peng (彭晨明). Inevitably, this is a very dynamic group, but it has shown extraordinary cohesion and solidarity all through these years. Special thanks go to Yanping, Quanrun, and Yuwan for helping me to settle down in Groningen back in 2012. Ye Liu has provided timely and professional methodological support, which helped me to finish the first two chapters of the thesis.
Many thanks go to my fellow PhD students in the OPERA Group: Oskar, Nick (Ziengs), Nick (van Beest), Monique, Minou, Sarah, Nonhlanhla, Robert-Jan, Hendryk, Bart, Aline, Sabine, Roel, Anne, Lisanne, Mitchel, and Lisa…… Thank you, Oskar, for your sweets and melatonin. Thank you, Monique, for helping me to identify and report the BSN fraud to the Belasting Dienst. Thank you, Hendryk, for the BBQs and for helping me to move out of the student house. Thank you, Aline, for providing me with first aid by the side of Neuchatel Lake. Thank you, Bart, for your skates. Thank you, Lisa, for taking over the premaster workshops when I was busy with taking care of my daughter and wife… Special thanks go to my office mate Robert-Jan. It has been a great pleasure to share Dui-651 with you in the past three years. We had a lot of interesting conversations, and you have provided me with many helps. Special thanks go to Bart and Robert-Jan for helping me to finish the last mile of the Marathon, namely preparing the Dutch summary (Samenvatting) included at the end of the thesis.
Many thanks go to my Badminton Group in Groningen: Zhe Sun (孙哲), Jieqiang Wei (韦杰强), Qibing Xie (谢其兵), Yanping Zhao (赵艳平), Shuhai Zhang (张书海), Suping Peng (彭素萍), Lu Zhang (张璐), Ning Ding (丁宁) & Lefon Ching (陈乐方), Xuewen Zhang (张学文) & Min Wu (武敏), Chenglong Deng (邓成龙), Huatang Cao (曹 华堂), Miaozhen Huang (黄妙珍), Shili Chen (陈仕莉), Yi Yu (余意), and Yueyang Sun (
孙月阳). Thank you, Dut Van Vo, for playing with me and training me in 2013. Special thanks go to our great trainer, Rik Wilbers, for his professional and patient instructions and trainings all through the years.
viii
I would also like to express my sincere gratitude to the following friends who have made my time at Groningen enjoyable and memorable: Tao Yuan (袁涛), Jiquan Wu (吴 继全) & Zhonghong Ma (马忠洪), Weiguo Xia (夏卫国), Jonanthan Gray, Qingkai Yang (杨庆凯) & Yuhan Wang (王煜涵), Binqi Tang (唐滨琪) & Shili Chen (陈仕莉), Yan Shao (邵闫) & Fan Wang (王帆), Shuai Feng (冯帅), Cong Duan (段丛), and Aobo Jiang (姜奥 博). Thank you, Tao, for taking good care of me for years. Your hospitality and delicious foods constantly made me indebted. Thank you, Jiquan and Zhonghong, for providing a shelter for me in 2014 and 2015. Special thanks go to Qingkai for taking care of my mother-in-law at Martini Ziekenhuis. Special thanks go to Mengmeng for taking good care of my wife and daughter while I was teaching in Sri Lanka. Special thanks go to Shili for taking care of Li and Sofie at UMCG, and Binqi for preparing the delicious and nutritional food for us when Sofie was born. Special thanks go to Fan and Yan for your delicious food and great cares… A friend in need is a friend indeed!
I would also like to express my sincere gratitude to several friends for their consistent support all through the years beyond my PhD journey: Zelin Hao (郝泽林), Jian Chen (陈剑), Long Liu (刘龙), Chuzhao Wang (王初照), Shaoyong Wang (王少永), and Jiajia Wu (吴嘉佳).
Special thanks go to Dr. Qian Wang (王谦) for her generous and consistent support in the past ten years. I am very lucky to have you as the supervisor for my master’s program at UCAS.
Last but not least, my sincere gratitude goes to my wife Li Jiang. Your consistent and timely encouragement, among others, has prevented me from being overly complacent and motivated me to aspire for a higher goal. Your talents in organizing and calm personality are a great complement to me. Special thanks go to Li for transcribing basically all my interviews in 2014. Moreover, special thanks go to Li for taking great care of our daughter Sofie.
ix
Table of Contents
Chapter 1... 1
Introduction ... 1
1.1 Sustainability in global supply chains ... 1
1.2 Tensions in sustainable supply chain management ... 2
1.3 Instrumental, institutional, and paradoxical perspectives toward SSCM ... 5
Chapter 2... 10
When does corporate sustainability performance pay off? The impact of country-level sustainability performance ... 10
2.1 Introduction ... 10
2.2 Theoretical foundation and hypotheses development ... 12
2.2.1 The CSP-CFP relationship ... 12
2.2.2 The impact of country-level sustainability performance on the CSP-CFP relationship ... 13
2.3 Methods ... 16
2.3 1 Sample and data ... 16
2.3 2 Measurements... 17
2.3.3 Common method bias ... 19
2.3.4 Data analysis ... 20
2.4 Results ... 21
2.4.1 Construct validity of corporate sustainability performance ... 21
2.4.2 Hypothesis testing with HLM ... 22
2.4.3 Simple slope analysis ... 24
2.5 Discussion ... 25
2.6 Conclusions ... 28
Chapter 3... 30
When are stakeholder pressures effective? An extension of slack resources theory ... 30
3.1 Introduction ... 30
3.2 Theoretical foundation and hypotheses ... 33
3.2.1 Stakeholder pressures and corporate sustainability management... 33
3.2.2 Slack resources theory ... 34
3.2.3 Country-level sustainability performance ... 35
3.3 Methods ... 37
3.3.1 Sample and data ... 37
3.3.2 Measurements... 38
x
3.3.4 Endogeneity ... 42
3.3.5 Data analysis ... 43
3.4 Results ... 45
3.4.1 Hypothesis testing with HLM ... 45
3.4 2 Simple slope analysis ... 46
3.5 Discussion ... 48
3.6 Conclusions ... 52
Chapter 4... 56
Inside the buying firm: exploring responses toward paradoxical tensions in sustainable supply chain management ... 56
4.1 Introduction ... 56
4.2 Theoretical background ... 58
4.2.1 Paradoxical tensions in SSCM ... 58
4.2.2 Managerial responses to paradoxical tensions in SSCM ... 60
4.3 Methodology ... 61
4.3.1 Case selection ... 61
4.3.2 Data collection... 63
4.3.3 Data analysis ... 67
4.4 Findings ... 67
4.4.1 The tension between sustainability standards and the Chinese socio-economic environment ... 67
4.4.2 Purchasing managers’ responses to sustainability tensions ... 69
4.4.3 Sustainability managers’ responses to sustainability tensions ... 71
4.5 Discussion ... 75
4.5 1 Contributions to theory ... 78
4.5 2 Managerial implications ... 80
4.5 3 Limitations and future research ... 81
4.6 Conclusions ... 82
Chapter 5... 85
Conclusions ... 85
5.1 Main findings ... 85
5.2 Managerial implications ... 87
5.3 Future research directions ... 89
1
Chapter 1
Introduction
“The greatest threat to our planet is the belief that someone else will save it.” - Robert Swan
“We are living on this planet as if we had another one to go to.” - Terry Swearingen
1.1 Sustainability in global supply chains
Global supply chains in industries such as consumer electronics, textile and fashion goods, and automobiles have provided the stimulus of economic development for emerging countries like China, India, and Brazil. Notwithstanding the contributions to overall economic growth, poverty alleviation, and improved living standards for emerging countries, multinational companies orchestrating global supply chains are widely criticized for undermining the environmental integrity and social equity of emerging markets (Locke, 2013; Neilson & Pritchard, 2011). The main reason is that emerging-market suppliers have been generally lax in protecting the environment and safeguarding the physical and financial wellbeing of workers to keep their cost competitiveness (Ngai, 2005; Ngai & Chan, 2012; Yu, 2008). Since the 1990s, as multiple stakeholder groups including consumers, non-governmental organizations, and environmental activists began to hold Western buying firms accountable for the labor standards and environmental practices of their suppliers in emerging countries, sustainability has emerged as an important element of supply chain management (Krause, Vachon, & Klassen, 2009; Wu & Pagell, 2011). That is to say, supply chain management has become sustainable supply chain management (Carter & Rogers, 2008).
Sustainable supply chain management (SSCM) is broadly defined as the strategic and transparent integration and achievement of social, environmental, and economic goals in the systemic coordination of key intra- and inter-organizational processes for improving the long-term viability of individual companies and their supply chains (Carter & Rogers, 2008; Golicic & Smith, 2013). Specifically, SSCM covers two main aspects: (1) the environmental aspect refers to the impact of supply chain activities on the natural
2
environment and covers issues of energy consumption, waste minimization, resource preservation, and pollution abatement (Linton, Klassen, & Jayaraman, 2007); (2) the social or human aspect focuses mainly on the physical, financial, and mental/emotional wellbeing of workers in global supply chains, including labor practices and occupational health and safety (Krause, Vachon, & Klassen, 2009; Wilhelm et al., 2016). In the past two decades, Western buying firms in diverse industries have been making efforts to improve sustainability performance of their global supply chains, but the effects, as well as the efforts, have been rather limited (Lund-Thomsen & Lindgreen, 2014).
Extensive empirical explorations in the past two decades (Abbasi & Nilsson, 2012; Busse et al., 2016; Ngai, 2005; Walker, Di Sisto, & McBain, 2008; Yu, 2008) show that SSCM is an inherently complicated process, mainly due to the multiple tensions that can exist between social/environmental and economic goals (Hahn et al., 2015), tensions between short-term and long-term goals (Slawinski & Bansal, 2015), interest conflicts between different groups of stakeholders (Chung, 2015; Neilson & Pritchard, 2011), and conflicts between sustainability standards and the socio-economic contexts of emerging countries (Lund-Thomsen & Lindgreen, 2014). Besides these tensions, developing sustainable supply chains in an emerging-market context is also impeded by the suppliers’ limited slack resources, inadequate management capabilities, and unconducive institutional environment (Abbasi & Nilsson, 2012; Busse et al., 2016; Ngai, 2005; Walker, Di Sisto, & McBain, 2008; Yu, 2008). In the ensuring sections, we first provide a succinct overview of the SSCM literature, with the aim of highlighting the major barriers to improving sustainability in global supply chains. Specifically, the multiple tensions embedded in SSCM processes are introduced in more detail. Second, we outline the three empirical chapters of this thesis, each of which applies a specific theoretical perspective to explore tensions in SSCM.
1.2 Tensions in sustainable supply chain management
First, significant tensions can exist between the economic and the social/environmental dimensions of sustainability. Providing decent salaries for all workers entails substantial increases in suppliers’ operational costs (Ngai, 2005; Ngai & Chan, 2012), and mitigating environmental impact often calls for significant technological and financial investments (Wu & Pagell, 2011). These tensions can unfold themselves more prominently in an emerging-market context, where cost is still the main order winning criteria for small-
3
and medium-sized manufacturers that have yet to develop core technological capabilities to stand out within fierce competition. Yu (2008) did an in-depth field study of a footwear manufacturer in Fuzhou (China), which was struggling to implement Reebok’s Supplier Code of Conduct. It was found that the cut-throat competition in the footwear manufacturing industry had driven down the profitability of manufacturers to a level that compliance with the labor standards of Reebok would almost eradicate the supplier’s marginal profitability. Likewise, Ruwanpura and Wrigley (2011) found that cost pressures were so overwhelming in the Sri Lankan apparel industry that most suppliers in the industry had to circumvent the labor and environmental standards of Western buying firms just to survive. Faced with significant tensions in SSCM, emerging-market suppliers would generally choose to suppress sustainability, as they typically have rather limited slack resources for improving sustainability and the institutional pressures for compliance are still relatively weak.
Second, SSCM also entails temporal tension between short-term and long-term goals (Hahn et al., 2015; Slawinski & Bansal, 2015). The Brundtland Commission on Environment and Development had put the temporal tension at the heart of sustainable development, which is defined as “the kind of development that meets the needs of the present without compromising the ability of future generations to meet their needs”. Intertemporal tensions at the society or system level unfold more vividly at the organizational level in the processes of managing corporate and supply chain sustainability (Hahn et al., 2015; Slawinski & Bansal, 2015). Firms have to balance their short-term (economic) goals with their long-term needs and those of a wide range of stakeholders, including workers, local community, and the society. It is not surprising to see that economic short-ism (Slawinski & Bansal, 2015) is the dominant approach for individuals and organizations to deal with the intertemporal tensions between short-term and long-term goals. Economic short-ism may well be more rampant in an emerging-market context, where the laws on social equity and environmental integrity are less strictly enforced (Chan, Pun, & Selden, 2013; Ngai & Chan, 2012). However, as several global sustainability issues such as environmental pollution, global warming, and climate change become more salient, it is foreseeable that major stakeholder groups will further pressurize emerging-market suppliers to take more responsibility for sustainable development. As such, economic short-ism may gradually become less rampant in an emerging-market context.
4
Third, tensions can also emerge between the sustainability standards advocated by Western buying firms and the socio-economic reality in emerging markets (Busse et al., 2016; Neilson & Pritchard, 2011). Sustainability standards are usually developed by Western buying firms or industry associations, using the laws and practices in industrialized countries as the template. The SSCM literature has reported several sustainability standards that are not always compatible with the socio-economic reality in emerging markets. For example, the use of child labor is a taboo in Western countries while it is oftentimes regarded as a supplementary schooling system in Southern Asian countries like India and Bangladesh (Lund-Thomsen & Lindgreen, 2014; Neilson & Pritchard, 2011). Likewise, overtime works are generally regarded as bad for the physical wellbeing of shop-floor workers, while it is widely preferred by Chinese workers as an important way of boosting their take-home incomes (Chung, 2015). As such, tensions can also exist at the individual level between short-term economic wellbeing and long-term physical wellbeing. As such, prior studies (e.g., Chung, 2015; Huq, Stevenson, & Zorzini, 2014) have shown that the social, economic, and institutional environment in emerging countries can shape the stakeholders’ interests and preferences in a way that is not always conducive for sustainability improvement as required by Western buying firms and other groups of stakeholders.
To summarize, sustainable supply chain management, especially in the emerging-country context, is a complex and complicated process that is impeded by the multiple tensions that are inherent in SSCM (Hahn et al., 2014; Hahn et al., 2015; Rivoli & Waddock, 2011). However, the SSCM literature has yet to systematically explore how the social, economic, and institutional context influences organizational experiences of and responses toward the tensions in developing sustainable supply chains. Moreover, related to the first gap, the way tensions are addressed in the inter- and intra-organizational processes of SSCM by those practitioners who experience them in the emerging-market context has been largely under-explored (Hahn et al., 2015; Matthews et al., 2016). Early studies (e.g., Margolis, Elfenbein, & Walsh, 2007; Orlitzky, Schmidt, & Rynes, 2003) in this field have intentionally or unintentionally overlooked the tensions in SSCM by subordinating social and/or environmental goals under economic goals, exploring questions like “does it pay to be green” and “does it pay to be good”. In other words, this stream of literature has mostly applied an instrumental perspective toward SSCM. Recently, there are increasing calls to go beyond the instrumental perspective and
5
to apply alternative integrative (Gao & Bansal, 2013) and/or paradoxical perspectives (Hahn et al., 2015; Jay, 2013; Matthews et al., 2016) to explore how practitioners working toward supply chain sustainability are coping with the enduring tensions in SSCM.
To fill these two main gaps, we apply three theoretical perspectives in this thesis to develop thorough understandings of SSCM. In Chapter 2, we re-apply the instrumental perspective to conduct a cross-country comparison on the relationship between corporate sustainability performance and corporate financial performance. Although this chapter does not directly explore how tensions are addressed in SSCM, it helps to develop some valuable insights into the time-contextual variations of sustainability-performance relationships in global supply chains. In Chapter 3, we apply the institutional theory to explore the effect of institutional pressures in driving firms to overcome the intertemporal tensions in SSCM. In Chapter 4, we delve into the buying firm and apply the paradox theory to explore how purchasing and sustainability managers within the buying firm are coping with the tensions in SSCM in an emerging-market context. Detailed introductions of these three empirical chapters are provided in the following section.
1.3 Instrumental, institutional, and paradoxical perspectives toward SSCM Chapter 2 analyzes the time-contextual dynamics of the relationship between corporate sustainability performance and corporate financial performance. We revisit a classic question in the SSCM literature: Does it pay to be green (good)? Drawing on instrumental stakeholder theory, we develop a focal hypothesis arguing that the financial effect of sustainability improvement is negatively impacted by country-level sustainability performance, because stakeholders will take a firm’s sustainability improvement for granted in countries with good social and environmental performance. We test this focal hypothesis in a cross-country setting drawing on the 6th International Manufacturing Strategy Survey. This chapter supplements these data with secondary data drawn from the Human Development Index and the Environmental Performance Index. The results support our hypothesis that firms in countries with higher levels of sustainability performance generally find it more difficult to capitalize on corporate sustainability performance than do their counterparts in countries with relatively low levels of sustainability performance. This chapter suggests that sustainability management can be a source of competitive advantage for firms located in emerging countries, where in general the level of sustainability performance is relatively low. Moreover, this chapter suggests that individual companies can experience increasing
6
levels of tensions in SSCM along with the institutionalization of sustainability within a broader socio-economic context. On the one hand, companies will have to go beyond low-hanging fruits and to make substantial and probably costly investments to meet higher sustainability standards. On the other hand, financial returns from such investments will gradually attenuate, partly due to changes in stakeholder responses.
In Chapter 3, we focus on the effect of institutional pressures in driving firms to overcome the tensions in SSCM. Drawing upon institutional theory and slack resources theory, we theorize that country-level sustainability performance can profoundly influence how firms allocate slack resources to address stakeholders’ requests on supply chain sustainability. Empirical results based on the data from 6th International Manufacturing Strategy Survey and secondary data of the Human Development Index and the Environmental Performance Index support our hypotheses. As hypothesized, in countries with low levels of sustainability performance, firms with considerable slack resources are more responsive to stakeholder pressures than their peers with limited slack resources. In contrast, in countries with high levels of sustainability performance, there are no significant differences between firms with and without considerable slack resources in their responsiveness to stakeholder pressures. Having used institutional theory to develop a boundary condition of slack resources theory (Waddock & Graves, 1997; Seifert, Morris, & Bartkus, 2004), this chapter contributes to a better understanding of organizational responses to stakeholder pressures. Moreover, this study suggests that institutional pressures are effective in driving firms to overcome temporal tensions in SSCM. More specifically, in countries with low levels of sustainability performance, firms with limited slack resources are more likely to focus on short-term goals and overlook stakeholder requests for long-term supply chain sustainability. In contrast, in countries with high levels of sustainability performance, firms are generally responsive toward stakeholder requests for long-term supply chain sustainability, irrespective of the levels of slack resources at hand.
In Chapter 4, we present a paradoxical perspective and argue that sustainability and other business aims are not always compatible, particularly in an emerging-market context. Often, tensions originate in conflicts between the socio-economic environment of emerging-market suppliers and their Western customers’ demands for both cost competitiveness and sustainability. We argue that Western buying firms can play a key role in moderating such tensions, as experienced by emerging-market suppliers.
7
Specifically, we explore how purchasing and sustainability managers within buying firms make sense of and respond to paradoxical tensions in SSCM. We conduct an in-depth case study of a Western multinational company that sources substantially from Chinese suppliers. While we find strong evidence for a persisting instrumental perspective in both the sensemaking and practices of purchasing and sustainability managers, we also observe an alternative response, taken primarily by sustainability managers that we labelled as “contextualizing”. Contextualizing can make sustainability standards more workable in an emerging-market context and help individual managers to move toward paradoxical sensemaking. Thereby, it can alleviate the tensions otherwise present in SSCM. We outline the value of paradoxical sensemaking in bringing about changes toward “true sustainability” (Montabon, Pagell, & Wu, 2016; Pagell & Shevchenko, 2014) in SSCM.
Figure 1.1 provides a visual summary of the three empirical chapters. To be specific, this figure highlights the theoretical perspective, the research focus, and the level of analysis of the three empirical chapters. Moreover, it provides a succinct summary of the theoretical implications for managing tensions in SSCM.
8
Figure 1.1. Instrumental, institutional, and paradoxical perspectives toward tensions in
sustainable supply chain management
TENSIONS IN SUSTAINABLE SUPPLY CHAIN MANAGEMENT
Chapter 4
Theoretical Perspective: paradoxical theory Research Focus: How are purchasing and sustainability managers coping with the tensions in SSCM?
Implications for Tensions: Bridging the gap between sustainability standards and the socio-economic reality of emerging countries can alleviate the tensions experienced by the managers from both the buying firm and the suppliers
Level of Analysis: functional
Chapter 3
Theoretical Perspective: Institutional theory Research Focus: The effect of institutional pressures and slack resources on
influencing organizational responses toward supply chain sustainability
Implications for Tensions: As sustainability becomes institutionalized in a broader socio-economic context, SSCM will be given higher priority when companies are faced with the temporal tensions between short-term and long-term goals.
Level of Analysis: Organizational
Chapter 2
Theoretical Perspective: Instrumental stakeholder theory
Research Focus: Time-contextual variations of sustainability-performance relationships Implications for Tensions: As sustainability becomes institutionalized in a broader socio-economic context, individual companies can experience increasing levels of tensions in SSCM.
9
Chapters 2 – 4 are the core of this thesis and are based on the following publications:
Chapter 2 --- Xiao, C., Wang, Q., Van der Vaart, T., Van Donk, D. P., 2018. When does corporate sustainability performance pay off? The impact of country-level sustainability performance. Ecological Economics, 146, 325 – 333.
Chapter 3 --- Xiao, C., Wang, Q., Van Donk, D. P., Van der Vaart, T., 2018. When are stakeholder pressures effective? An extension of slack resources theory. International Journal of Production Economics, 199, 138 – 149.
Chapter 4 --- Xiao, C., Wilhelm, M., Van der Vaart, T., Van Donk, D. P., 2018. Inside the buying firm: exploring responses toward paradoxical tensions in sustainable supply chain management. Journal of Supply Chain Management, in press.
Chapter 5 concludes this thesis. First, we shortly summarize the three empirical chapters. Second, we elaborate on the practical implications of this thesis on managing tensions in SSCM. As developing sustainable supply chains entails addressing tensions, we provide several suggestions that can help supply chain practitioners to cope with tensions in more constructive ways. Third, as tensions are yet to be subject to extensive empirical explorations in the SSCM literature, we provide several promising avenues of further research into tensions.
10
Chapter 2
When does corporate sustainability performance pay off? The impact of
country-level sustainability performance
2.1 Introduction
Managers and researchers around the globe acknowledge that it is crucial for firms to improve their corporate sustainability performance (CSP), which indicates a firm’s contribution to environmental protection and social development (Wagner, 2010). However, there is less unanimity regarding the financial benefits that can be expected from such improvements. This chapter argues that stakeholder responses to CSP improvements are influenced by the country-level sustainability performance, a measure of how well a country performs in terms of high-priority social and environmental issues (Siche et al., 2008; Wagner, 2010). Specifically, we submit that CSP improvements can be financially effective in countries with relatively low levels of sustainability performance, whereas similar investments may have very limited financial effect in countries with high levels of sustainability performance.
Empirical studies exploring the relationship between CSP and corporate financial performance (CFP) have variously reported positive, non-significant, and even negative results (Margolis, Elfenbein, & Walsh, 2007; Flammer, 2015). This indicates that even after more than thirty years of research, there is no clear answer to the question as to whether CSP pays off (Barnett, 2007). In their attempts to explain these mixed results, researchers have explored how firm-level factors, such as engagement strategy (Tang, Hull, & Rothenberg, 2012) and stakeholder influence capacity (Barnett & Salomon, 2012), may impact on the CSP-CFP relationship. In contrast, systematic reviews of the relevant literature (Lee, 2008; Aguinis & Glavas, 2012) show that researchers have barely explored the potential impact of societal factors on this relationship, even though it is widely accepted that the financial effect of CSP is primarily rooted in firm-society interactions and positive stakeholder responses (Jones, 1995; Barnett, 2007). As Arya and Zhang (2009) and Aguinis and Glavas (2012) put it, without a clear understanding of societal impact on the financial effect of CSP, our knowledge regarding the CSP-CFP relationship can, at best, be described as partial. This chapter aims to close this gap and explores societal impact on the financial effect of CSP.
11
Our focus on societal impact is further inspired by several pairs of empirical studies on the CSP-CFP relationship in developed countries, e.g., Shane and Spicer (1983) versus Diltz (1995), and Spicer (1978) versus Pava and Krausz (1996). The earlier study in both pairs found positive CSP-CFP relationships, whereas the more recent studies did not find support for such relationships in similar contexts using similar measurements. It is plausible that country-level sustainability performance could explain this apparent inconsistency in the CSP-CFP relationship. In the 1970s and 1980s the sustainability performance of developed countries was generally low. Although there was little external pressure to improve CSP, firms that did so could gain significant financial benefits because these improvements could advance firm-stakeholder relationships (Jones, 1995; Barnett, 2007). Since then, developed countries have enforced laws and regulations on environmental integrity and social equity, and there have been substantial improvements in sustainability performance over the past two decades. As a consequence, stakeholders are no longer that sensitive and responsive to firms improving their social and environmental performance, making it difficult for firms to gain financial benefits from such improvements (Barnett, 2007). Based on these observations and reasoning, we hypothesize that country-level sustainability performance can negatively influence the financial effect of CSP. To test this, this chapter evaluates the CSP-CFP relationship in a cross-country setting drawing on the data of the 6th International Manufacturing Strategy Survey (IMSS VI), which are gathered in 22 countries characterized by considerable differences in country-level sustainability performance.
This chapter makes an important contribution to the CSP literature by extending it to include country-level sustainability performance in the CSP-CFP relationship. So far, researchers have mainly focused on firm-level factors that may impact on the financial effect of CSP. In reality, CSP is an area of extensive and enduring firm-society interactions (Matten & Moon, 2008; Lee, 2008) and the financial effect of CSP largely stems from these firm-society interactions (Jones, 1995; Barnett, 2007). As such, this chapter provides novel insights into the contingent relationship between CSP and CFP, and it contributes to a more nuanced understanding of the financial effect of CSP. Moreover, this chapter contributes to the literature by testing the CSP-CFP relationship in a sample of manufacturers located in 22 countries. Our findings show that the financial effect of corporate sustainability performance differs across countries. As such, we
12
suggest that a global approach towards sustainable development should also take into account country differences.
The remainder of this chapter is structured as follows. The second section introduces the theoretical foundation of the CSP-CFP relationship and various studies that have addressed the mixed financial effects of CSP. Further, this section also introduces country-level sustainability performance and develops the hypotheses of this study. The characteristics of the dataset used and the statistical methods applied are discussed in the third section. The fourth section reports the results. The fifth section discusses the implications of the findings for CSP theories and practices, the limitations of this study, and research avenues that are worthy of further exploration. The final section draws conclusions.
2.2 Theoretical foundation and hypotheses development
This section consists of two subsections. In the first, we briefly introduce corporate sustainability performance and the mechanism through which CSP improvements can contribute to corporate financial performance, and discuss the firm-level factors that have been found to influence the CSP-CFP relationship. The second subsection introduces country-level sustainability performance and discusses how this factor can impact on the financial effect of CSP.
2.2.1 The CSP-CFP relationship
CSP reflects a firm’s impact on society, including employees, customers, suppliers, and local communities, and on the natural environment (Hillman & Keim, 2001; Matten & Moon, 2008). As such, CSP is conceptualized as a broad construct consisting of social/human and environmental dimensions (Perrini et al., 2011). The environmental aspect of CSP is relatively well understood and covers the issues of resource preservation, energy consumption, waste minimization, and emission/pollution abatement (Krause et al., 2009; Wagner, 2010). The social or human aspect of CSP has a broader scope and concerns poverty alleviation, health and safety of employees, protection of human rights, and participation in diverse social initiatives (Krause et al., 2009; Perrini et al., 2011).
13
Several theories have argued for a positive CSP-CFP relationship, such as resource productivity theory (Porter & Van der Linde, 1995), the natural resource-based view of the firm (Hart, 1995; Russo & Fouts, 1997), and instrumental stakeholder theory (Jones, 1995; Barnett, 2007). Among them, instrumental stakeholder theory has been more intensively used and empirically tested. More importantly, given its focus on stakeholders it fits with the notion that CSP is shaped within a societal context and in interaction with the firm’s environment. Instrumental stakeholder theory argues that CSP improvements can contribute to CFP by advancing a firm’s relationships with its internal and external stakeholders (Barnett, 2007; Jones, 1995). In more detail, it is argued that internal stakeholders such as shop-floor workers and managers will respond positively to a firm’s CSP improvements, such as initiatives to manage occupational health/safety risks and programs on work-life balance (Jones, 1995; Wicks et al., 1999). These favorable responses lead to improved human capital and innovation capability for firms, which in turn can result in reduced costs and better financial performance (Perrini et al., 2011). Similarly, CSP improvements can deliver positive signals to external stakeholders, such as customers, non-governmental organizations (NGOs), the general public, and governmental agencies, about a firm’s commitment to social and environmental wellbeing (Perrini et al., 2011). These stakeholders will take these improvements (e.g., environmental protection and charitable giving) as evidence that the firm is trustworthy and reliable. Such firms can benefit from an improved reputation and customer satisfaction through charging price premiums and the expanded marketing opportunities (Luo & Bhattacharya, 2006; Perrini et al., 2011; Surroca et al., 2010). This is summarized in our first hypothesis of this chapter:
H2.1: There is a positive relationship between corporate sustainability
performance and corporate financial performance.
2.2.2 The impact of country-level sustainability performance on the CSP-CFP relationship
Although the theoretical framework underpinning the CSP-CFP relationship outlined above is widely accepted (Barnett, 2007), systematic reviews (e.g., Margolis et al., 2007; Orlitzky et al., 2003) show that the large number of empirical studies testing this relationship have produced mixed results. The mixed findings, summarized by Barnett
14
(2007) and Margolis and Walsh (2003), indicate that CSP can have varying financial effects across firms and contexts. As a result, researchers have started to explore the potential contingencies that might influence the financial effect of CSP (e.g., Barnett & Salomon, 2012; Servaes & Tamayo, 2013; Tang et al., 2012; Wang and Choi, 2013). Several firm-level factors have been identified, including stakeholder influence capability (Barnett & Salomon, 2012), customer awareness (Servaes & Tamayo, 2013), and engagement strategy (Tang et al., 2012). For instance, Barnett and Salomon (2012) found that the CSP-CFP relationship is positively moderated by stakeholder influence capacity, defined as the ability of a firm to “identify, act on, and profit from opportunities to improve stakeholder relationships through corporate social responsibility” (p. 1306).
In addition to these firm-level factors, factors and developments related to sustainability performance at the country level can also influence the financial effect of CSP. It is notable that the CSP literature presents a pattern where the CSP-CFP relationship seems to be weakening over time. For instance, Luck and Pilotte (1993) and Hillman and Keim (2001) retested the CSP-CFP relationship in the large US publicly traded companies included in the Kinder, Lydenberg, and Domini (KLD) database using the same measurements. However, whereas Luck and Pilotte (1993) found a significant positive relationship, Hillman and Keim (2001) found a less clear relationship. Similar contrasting findings are to be found in Shane and Spicer (1983) versus Diltz (1995), McGuire et al. (1988) versus Brown (1997), and Spicer (1978) versus Pava and Krausz (1996). This observation can be related to what has been labelled as country-level sustainability performance, which reflects the extent to which the tenets, principles, and practices of environmental integrity and social equity are institutionalized and embedded in a specific country domain (Campbell, 2006; Moon, 2014).
Seeing country-level sustainability performance as a factor in the CSP-CFP relationship provides a plausible explanation for the variation in this relationship presented above. Country-level sustainability performance reflects how well high-priority social issues (e.g., alleviation of poverty, education promotion, protecting human rights, and improving living standards) and environmental issues (e.g., resource conservation, pollution abatement, and eco-efficiency) have been addressed in a country (Siche et al., 2008; Wagner, 2010). As the overall social and environmental performance of a country improves, internal stakeholders such as employees, are likely to respond less positively to a firm’s CSP improvements, because internal stakeholders will increasingly see it as a
15
firm’s fiduciary duty to contribute to environmental integrity and social equity. Consequently, improving a firm’s social and environmental performance will gradually lose the potential to stimulate internal stakeholders’ participation, collaboration, and knowledge creation/sharing. Similarly, as the overall social and environmental performance is improved in a country, improvements in a firm’s CSP will gradually lose the potential to help the firm to expand marketing opportunities and charge price premiums. The main reason is that external stakeholders, including customers, the general public, NGOs, and governmental agencies, will become less sensitive and responsive to a firm’s contribution to environmental integrity and social equity as these become progressively institutionalized and taken-for-granted in a country.
We use environmental protection, a crucial aspect of sustainable development (Delmas & Toffel, 2008; Sarkis et al., 2010), to illustrate this effect. When a country is suffering from problems caused by environmental degradation, firms contributing to environmental protection will be rewarded with positive responses from a wide range of stakeholders, including their employees, customers, NGOs, and the general public. These positive responses from internal and external stakeholders help to generate financial returns (Barnett, 2007; Flammer, 2015; Jones, 1995; Oikonomou et al., 2014). As the government enforces the laws on environmental protection, and environmentally friendly practices are widely adopted by firms, various groups of stakeholders will gradually come to see environmental protection as the norm. In other words, the social expectations of environmental contributions from firms will grow. Internal and external stakeholders will become less sensitive and responsive to firms’ efforts to reduce their environmental impact. Consequently, firms will find it increasingly difficult to advance the firm-stakeholder relationships through improving environmental performance. Firms that try to outperform the social expectations on environmental performance may well incur additional costs that outweigh the potential benefits from these improvements. Hence the financial effects will at best be neutral.
To summarize, if the overall sustainability performance of a country improves, firms located in that country will generally find it increasingly difficult to advance firm-stakeholder relationships through contributing to social and environmental wellbeing. That is, CSP improvements will lose the value of contributing to corporate reputation, customer satisfaction, innovation capabilities, and human capital, i.e., the intangible
16
assets that can benefit financial performance (Perrini et al., 2011; Surroca et al., 2010). Consequently, the financial effect of CSP will decrease as a country’s sustainability performance improves. These arguments are summarized in the following hypothesis.
H2.2: Country-level sustainability performance negatively moderates the positive
relationship between CSP and CFP. In other words: In countries whose sustainability performance is relatively poor, the CSP-CFP relationship will be significantly positive, while this relationship will be less positive in countries that have achieved high levels of sustainability performance.
2.3 Methods
2.3 1 Sample and data
The sample used in this study is drawn from the 6th International Manufacturing Strategy Survey (IMSS VI). The IMSS is a research project carried out every four years by a global network of researchers that provides cross-sectional data on manufacturing strategy within the discrete manufacturing industry (ISIC 25 - 30) through a detailed questionnaire administered in multiple countries by local research groups. The original questionnaire was developed in English and the same questionnaire was translated into local languages to facilitate data collection. The most recent round of survey (IMSS VI) was administered in 2013 and contained 931 valid responses from 22 countries, including developed countries such as Germany and the Netherlands, and developing countries such as China and India (See Appendix 2A for an overview of the countries and number of cases from each country). As responses with missing values were not included in the analysis of this study, 811 out of the 931 cases were used in this chapter. These 22 countries differ substantially in terms of their economic, social, and institutional development. Moreover, the more-developed countries have relatively good social and environmental performance (Schultz & Wehmeier, 2010), while developing countries such as China (Yin & Zhang, 2012) and India (Chapple & Moon, 2005) are still struggling with many social and environmental problems. As such, IMSS VI provides us with a cross-country setting characterized by considerable differences in country-level sustainability performance (See Appendix 2A for an overview of the social/environmental performance of these 22 countries), and it can serve as an appropriate empirical setting for testing our research hypotheses.
17 2.3 2 Measurements
Corporate sustainability performance (CSP)
Corporate sustainability performance (CSP) indicates a firm’s contribution to environmental protection and social development and consists of social and environmental dimensions (Wagner, 2010). We have used four IMSS VI items to form a measure of a firm’s CSP. These four items consist of two social indicators, i.e., workers’ motivation and satisfaction (Gimenez et al., 2012), and health and safety condition (Pagell & Gobeli, 2009), and two environmental indicators, i.e., materials, water, and/or energy consumption (Rao, 2002), and pollution emission and waste production levels (Zhu & Sarkis, 2004). We see these four items as adequately reflecting a firm’s impact on society and on the natural environment.
In the IMSS VI, respondents (mainly directors of operations or manufacturing) were requested to evaluate their firm’s improvements in terms of these four aspects over the previous three years (between 2009 and 2012). The two social indicators are measured on a five-point scale with a score of 1 representing a decrease in performance (by at least 5%) and 5 representing a strong improvement (by at least 25%). With the two environmental indicators, a score of 1 represents a decrease in environmental performance (an increase in consumption or pollution of at least 5%) and 5 represents a strong improvement (at least a 25% reduction in consumption or pollution). An explorative factor analysis indicated that these four items loaded satisfactorily onto one underlying factor. The Cronbach’s alpha was 0.72, indicating acceptable within-scale reliability. We consequently aggregated and averaged these four indicators to provide a net score for CSP improvement for each firm.
Corporate financial performance (CFP)
Corporate financial performance (CFP) indicates the overall financial wellbeing of a firm over a certain period of time, and it can be used to compare similar firms in the same industry or to compare industries in different sectors. We measured CFP using return-on-sales (ROS), which is defined as earnings before interest and taxes divided by total return-on-sales (Venkatraman & Ramanujam, 1986). This approach is commonly found in the literature (Julian & Ofori‐Dankwa, 2013; Peloza, 2009). It is argued that CSP can contribute to CFP through reducing transaction costs and increasing price premiums (Perrini et al., 2011).
18
Therefore, we believe that the financial effect of CSP should be reflected in a change of the ROS. In IMSS VI, respondents were asked to evaluate the improvement of their firm’s ROS over the past three years using a five-point Likert scale, ranging from 1 (much lower) to 5 (much higher). Such a single-item metric can be used when the measure is clear and directly interpretable by respondents (Bergkvist & Rossiter, 2007; Power et al., 2015), as is the case of this chapter.
Country-level sustainability performance (CLSP)
Country-level sustainability performance (CLSP) measures how well a country performs in terms of high-priority social and environmental issues (Siche et al., 2008; Wagner, 2010). In this chapter, we used the Human Development Index (HDI) developed by the United Nations Development Program and the Environmental Performance Index (EPI) developed by Yale University to reflect country-level sustainability performance (See the Appendix 2A for more detail). The HDI captures a country’s overall performance in terms of three important social issues: per capita income as a proxy for living quality, life expectancy at birth as a proxy for health achievement, and adult literacy together with educational enrolment as a proxy for educational attainment (Neumayer, 2001). The HDI is seen as concisely reflecting the overall social sustainability performance of a country (Sagar & Najam, 1998). Similarly, the EPI ranks how well countries perform in terms of high-priority environmental issues in two areas: protection of human health from environmental harm (including health impacts, air quality, water, and sanitation) and protection of the ecosystem (including water resources, agriculture, forests, fisheries, biodiversity and habitat, and climate and energy). As such, the EPI reflects the overall environmental sustainability performance of a country. As sustainability covers both social and environmental issues, we combined HDI and EPI to reflect a country’s overall sustainability performance. The HDI scale ranges from 0 to 1, and the EPI scale ranges from 0 to 100. We rescaled the EPI to range from 0 to 1 and added these two indexes to reflect the country-level sustainability performance with a scale ranging from 0 to 2.
19
Our aim was to explore how country-level sustainability performance moderates the CSP-CFP relationship, with CSP-CFP as the focal dependent variable in this study. Therefore, we needed to control for other variables that have been argued to exert significant impact on CFP, as is measured by ROS in this study. Following this principle, we identified three control variables: firm size (Orlitzky, 2001), market size dynamics (Campbell & Hopenhayn, 2005), and technological dynamics (Porter, 2008) that have all been argued or found to influence the profitability of firms. Firm size is measured as the number of employees. Respondents were asked to evaluate market size dynamics (the rate of market growth or decline in the industry) using a five-point Likert scale with 1 representing “declining rapidly” and 5 representing “growing rapidly”. Similarly, technological dynamics (the rate of technological change in the industry) was measured with a five-point Likert scale, ranging from 1 (a very low rate of technological change) to 5 (a very high rate of technological change).
It is often recommended to include country and industry dummy variables in a multi-country and multi-industry study. First, HDI (Human Development Index) and EPI (Environmental Performance Index) comprehensively cover country level factors such as the level of economic development (reflected by GDP per capita) and environmental regulations. As such, including HDI and EPI in the model as a moderator entails including them as additional control variables in the model, and they can adequately account for the cross-country variations in the dependent variable. Adding additional country dummy variables as control variables would just cause confusion. Following the suggestions of Becker (2005), we decide not to include additional country dummy variables. Second, the industries included in the IMSS VI are quite homogeneous (discrete manufacturing, ISIC code 25-30). Still, as an additional safeguard, we did analyze the model by including industry as control variables, but identified non-significant effects on the dependent variable. Therefore, as adding industry dummy variables cannot improve the explanatory power of the model, we decided to keep the model parsimonious.
2.3.3 Common method bias
The data used in this study comes from two different sources and this can substantially reduce the risk of common method bias. Although IMSS VI is cross-sectional in nature and relies on a single respondent from each firm, the risk of common method bias is
20
relatively low for the purpose of the present study. First, this survey involves a questionnaire with multiple scales. As such, it would be quite difficult for the respondents to link questions. Further, the items measuring CSP and the item measuring CFP are drawn from different scales, which can substantially reduce self-report biases (cf. Podsakoff et al., 2003). Second, the exploratory factor analysis shows that the five items taken from IMSS VI load on two factors that together account for 65.05% of the total variance. Further, the largest factor accounts for less than half (36.15%) of the total variance. Although these results do not preclude the possibility of common method bias, their magnitude indicates that common method bias is unlikely to seriously confound the relationships studied (cf. Podsakoff et al., 2003).
2.3.4 Data analysis
Because the key independent variable of this study, CSP, was measured via respondent self-reports, we assessed the construct validity of this measure by examining its dimensionality, face validity, and content validity, using Cronbach alpha and factor loadings.
The CSP-CFP model to be tested was hierarchical, with the dependent variable, CFP, being a firm-level construct, and the predicting variables spanning the firm and country levels. Therefore, we adopted the hierarchical linear modeling (HLM; Raudenbush & Bryk, 2002) method. To obtain consistent estimations of the coefficients, we opted for the fixed effects model (see below), as the random effects assumption is generally too strong to be met (Clarke et al., 2010).
Formally, our model can be described as follows:
Level 1 (firm level)
𝐶𝐹𝑃𝑖𝑗 = 𝛽0𝑗+ 𝛽1𝑗𝐶𝑆𝑃𝑖𝑗+ 𝑐𝑜𝑛𝑡𝑟𝑜𝑙𝑠 + 𝑒𝑖𝑗 (𝑙𝑒𝑣𝑒𝑙 1) Level 2 (country level)
𝛽0𝑗 = 𝛾00+ 𝛾01𝐶𝐿𝑆𝑃𝑗+ 𝜇0𝑗(𝑙𝑒𝑣𝑒𝑙 2) 𝛽1𝑗 = 𝛾10+ 𝛾11𝐶𝐿𝑆𝑃𝑗+ 𝜇1𝑗 (𝑙𝑒𝑣𝑒𝑙 2)
CFPij and CSPij stand for the CFP and CSP of firm i in country j, respectively. CLSPj
stands for the country-level sustainability performance of country j.
We tested the model in five steps, following the suggestions of Davison et al. (2002) and Liao and Rupp (2005). First, we estimated a null model which has no
21
predictors at either level 1 (firm level) or level 2 (country level). We partitioned the variance of CFP into within- and between-countries components. Second (model 1), in a level 1 analysis, within each country, CFP is regressed on the control variables (firm size, market dynamics, and technological dynamics). Third (model 2), we included the focal independent variable of this study, CSP, as an additional firm-level predictor of CFP. A regression line was estimated for each of the 22 countries in this step. In the fourth step (model 3), we used the intercept estimates obtained from level 1 as outcome variables and regressed these on the country-level variable, CLSP. In the last step (model 4), we regressed the slope estimates obtained from level 1 analysis on the country-level variable, CLSP, to detect cross-level interaction effects.
HLM includes a measure of model fit, the deviance statistic that can be used to compare the fit of two models if one is hierarchically embedded in the other. The difference in the deviance statistics for two nested models asymmetrically follows a chi-square distribution (Davison et al., 2002). Therefore, at each step, we computed the deviance statistic of the model, calculated the difference of this deviance statistic, and tested the significance of the deviance difference. Moreover, we also calculated the Akaike Information Criterion (AIC; Akaike, 1987) to evaluate the statistical fitness of the models.
Finally, in this study we aimed to understand the extent to which country-level sustainability performance moderates the CSP-CFP relationship. Based on the results of HLM analysis, we conducted simple slope analysis (Dawson & Richter, 2006; Preacher et al., 2006) to enable a sound interpretation of the cross-level interaction effect.
2.4 Results
2.4.1 Construct validity of corporate sustainability performance
We conducted the following analyses to demonstrate the validity of the measure of CSP in this data. We first checked the face validity of this measure by discussing it with several experts in the field of sustainable supply chain management. Positive responses from these experts confirmed the face validity of corporate sustainability performance. We examined the dimensionality of this measure by conducting a principle components factor analysis with “varimax” rotation and obtained a one-factor solution in which all the items
22
had high “loadings” (average loading = 0.737). The content validity of this measure was checked by referring the literature that has explored different aspects of corporate sustainability performance (Bourlakis et al., 2014; Gimenez et al., 2012; Golini et al., 2014; Longoni et al., 2014; Luzzini et al., 2015; Pagell & Gobeli, 2009; Rao, 2002; Wong et al., 2012; Zhu & Sarkis, 2004). The environmental and social dimensions included in this study have been widely used in these studies. Table 2.1 summarizes the descriptive statistics and product moment correlations of all the variables in this chapter.
Table 2.1. Descriptive statistics and correlations
Mean S.D. 1 2 3 4 5
1. Log. (Firm size) 6.024 1.720 1
2. Market dynamics 3.280 0.852 0.100** 1
3. Technological change 3.300 0.992 0.119** 0.339** 1
4. CSP 2.877 0.700 0.054 0.155** 0.215** 1
5. CLSP 1.461 0.263 －0.036 －0.235** －0.225** －0.218** 1
6. CFP 2.970 0.971 0.088* 0.257** 0.126** 0.167** －0.077*
**. Correlation is significant at the 0.01 level (two-tailed) *. Correlation is significant at the 0.05 level (two-tailed)
N=811 (Responses with missing values were not included in the analysis)
2.4.2 Hypothesis testing with HLM
Null model: Using HLM, we estimated a null model in which no predictors were specified for either the level 1 or level 2 function to test the significance level of the level 2 residual variance of the intercept (τ00 = 1.46%, p > 0.05). The intra-class correlation coefficient (ICC1) was 1.39%, indicating that only 1.39% of the variance in corporate financial performance can be attributed to country-level factors. In this chapter, we do not hypothesize that country-level sustainability performance is a predictor of CFP. Instead, country-level sustainability performance is hypothesized to moderate the financial effect of CSP. Therefore, even though ICC1 of the null model is rather small, it does not mean that we should not include country-level variables in further analysis.
Model 1 (adding firm-level control variables): At this step, we included three firm-level control variables as predictors of corporate financial performance. One of the control variables (i.e., market dynamics) contributed significantly to corporate financial performance (β = 0.254, p < 0.001). This result is consistent with prior studies showing that an expanding market creates room for firms to improve their financial performance (Campbell & Hopenhayn, 2005).
Model 2 (adding corporate sustainability performance): At this step, the focal independent variable of this study, corporate sustainability performance, is included as an additional predictor of corporate financial performance. The result shows that corporate
