Global supply chain risk management strategies Ila Manuj Department of Marketing and Logistics, University of North Texas, Denton, Texas, USA, and John T. Mentzer The University of Tennessee, Knoxville, Tennessee, USA Abstract Purpose ��� Global supply chains are more risky than domestic supply chains due to numerous links interconnecting a wide network of firms. These links are prone to disruptions, bankruptcies, breakdowns, macroeconomic and political changes, and disasters leading to higher risks and making risk management difficult. The purpose of this paper is to explore the phenomenon of risk management and risk management strategies in global supply chains. Design/methodology/approach ��� This paper is based on an extensive literature review and a qualitative study comprising 14 in-depth interviews and a focus group meeting with senior supply chain executives. Findings ��� The study provides insights into the applicability of six risk management strategies with respect to environmental conditions and the role of three moderators. Research limitations/implications ��� The model is developed in a global manufacturing supply chain context. It should be tested in other contexts and with other methods to provide generalizability. The study takes a much needed step toward building a theory of risk management in global supply chains, which opens important future research directions. Practical implications ��� This research provides direction to managers for choosing risk management strategies based on the global supply chain environment. Moderators have practical implications for global supply chain managers. Originality/value ��� The paper addresses an identified gap in the literature for selecting risk management strategies in global supply chains. It employs grounded theory, a methodology appropriate for theory-building, to explore a phenomenon with an inadequate theoretical base. Keywords Risk management, Supply chain management, International business Paper type Research paper Introduction Global supply chains are a source of competitive advantage. Global configurations of firms provide access to cheap labor and raw materials, better financing opportunities, larger product markets, arbitrage opportunities, and additional inducements offered by host governments to attract foreign capital (AlHashim, 1980 Kogut and Kulatilaka, 1994). However, coupled with these benefits that entice firms to go global are the uncertainties and consequent risks that managers face in global supply chains. As Barry (2004) argues, ���An enterprise may have lowest over-all costs in a stable world environment, but may also have the highest level of risk ��� if any one of the multiple gating factors kink up an elongated global supply chain!��� There is wide acknowledgement in the literature of the risks and uncertainties in global supply chains. Although risk management in multinational enterprises has been The current issue and full text archive of this journal is available at www.emeraldinsight.com/0960-0035.htm IJPDLM 38,3 192 Received 6 August 2007 Revised 7 January 2008 Accepted 28 January 2008 International Journal of Physical Distribution & Logistics Management Vol. 38 No. 3, 2008 pp. 192-223 q Emerald Group Publishing Limited 0960-0035 DOI 10.1108/09600030810866986

examined (Baird and Thomas, 1985 Baird and Thomas, 1991 Carter and Vickery, 1989 Ghoshal, 1987 Kogut, 1985 Lessard and Lightstone, 1986 Miller, 1992), risk management was relegated to the background until recently when several researchers (Barry, 2004 Cavinato, 2004 Christopher and Lee, 2004 Giunipero and Eltantawy, 2004 Juttner, �� 2005 Manuj and Mentzer, 2008 Norrman and Jansson, 2004 Spekman and Davis, 2004 Zsidisin, 2003a Zsidisin et al., 2004) revived an interest in risk management, particularly in global supply chains. Chopra and Sodhi (2004) contend that most companies develop plans to protect against recurrent low-impact risks in their supply chains. Many, however, ignore high-impact, low-likelihood risks. By understanding the variety and interconnectedness of supply chain risks, managers can tailor balanced, effective risk-reduction strategies for their companies. Hauser (2003) suggests that in today���s increasingly complex environment, risk adjusted supply chain management can translate into improved financial performance and competitive advantage. In sum, understanding global supply chain risk management is important and a top priority for both academics and practitioners. An initial review of the literature on risk management in supply chains led to the identification of three major gaps. First, no definition exists that adequately takes into account the unique dimensions of risk and risk management in a global supply chain. Rather, there are a multitude of definitions and conceptualizations, and therefore, confusion between terms such as risks, uncertainties, vulnerabilities, and sources of risks. Second, strategies to address risks warrant more attention ( Juttner, �� 2005). Although several studies provide a list of risk management strategies ( Juttner �� et al., 2003), these studies do not address how managers select among them, i.e. the antecedents to global supply chain risk management strategies, and their consequences. Juttner �� et al. (2003) suggest investigating risk management in different supply chains and developing strategies based on their environments. Third, there is limited research on moderators of the risk management process (Manuj and Mentzer, 2008). In light of these gaps, the purpose of this paper is to take a step toward building a theory of global supply chain risk management strategies. A qualitative research design was chosen, as not much is known about the phenomenon. Based on the analysis of in-depth qualitative interviews using grounded theory methodology, a model of risk management strategies in global supply chains is developed. It is important to mention that the context in this qualitative study was manufacturing firms. Manufacturing firms can be considered the focus of product supply chains, and thus, a good starting point for initial development of supply chain theory. Expanding the theory to other contexts (such as other levels in the supply chain or for service providers) is left to future research. This research makes three major contributions. A definition of supply chain risk is developed and four types of global supply chain risks are defined. In addition, the interaction of different risks in the global environment is explored. Second, a definition of risk management in a global supply chain context is developed. Rich descriptions of risk management strategies are provided and important antecedents to strategy selection are discussed. Third, three moderators in the process of risk management are explored, namely team composition, supply chain complexity management, and inter-organizational learning. Supply chain risk management strategies 193

Methodology The nature of the research problem should drive the choice of research strategy (Creswell, 1998 Denzin and Lincoln, 1998). Consistent with this philosophy and because of the lack of extant knowledge on the phenomenon, we employed grounded theory methodology. Grounded theory is a qualitative research methodology that allows the exploration of concepts, identification of relationships in raw data, and organization of concepts and the relationships into a theoretical scheme (Strauss and Corbin, 1998). Another advantage of grounded theory is the ability to handle complex phenomenon such as risk management because the methodology emphasizes the need for developing multiple concepts and their linkages in order to capture the central phenomenon. Insights from the grounded theory study and existing literature were used to develop the model presented in this paper. Selecting participants who can provide meaningful data on multiple incidents is critical for grounded theory. To maximize the variations in the phenomenon, we interviewed managers involved in making and executing global supply chain decisions from a variety of manufacturing companies, including home appliances, electronic component suppliers, pharmaceuticals and over-the-counter products, office products, heavy equipment, and consumer goods. A varied group of managers holding different positions from different industries and firms with different sizes were recruited to look for similar incidents across multiple contexts. We included managers who had worked in several different companies and industries, as well as those who had worked with one organization over an extended period of time and witnessed the company move through several transformations. The data in this paper come primarily from fourteen in-depth qualitative interviews with senior supply chain executives across eight companies. Apart from the interviews, we also conducted a focus group meeting involving seven senior executives of a global manufacturing firm. Five of these seven executives were later interviewed separately for this study and are a part of the 14 in-depth interviews. In total, the study had 16 unique participants. The number and content of in-depth interviews was based on the concept of ���theoretical sampling,��� i.e. successive respondents were chosen based on the emerging theory. The initial participant sample was based on a participant���s interest in and experience with the phenomenon, job profiles, articulation skills, and willingness to participate in the research. Theoretical sampling ensures adequate representation of important themes. It is a process of data collection for generating theory whereby the analyst jointly collects, codes, and analyzes the data and decides what data to collect next and where to find them, in order to develop the theory as it emerges (Glaser, 1978). We started with broad open-ended questions which were followed by focused and directed questions as the interviews progressed within interviews and between successive interviews (Morrison et al., 2002 Strauss and Corbin, 1998). Interviews continued until ���theoretical saturation��� was reached ��� i.e. further interviews did not reveal any new information related to antecedents, risk management strategies, dimensions of risk, or moderators in the risk management process. In theory building, the objective is deep understanding from a few key informants (McCracken, 1988). Generalizing the findings is the purpose of theory-testing research (Mentzer and Flint, 1997), such as surveys, experiments, or simulation. Fourteen interviews are in line with the qualitative research IJPDLM 38,3 194

guidelines that eight or fewer informants often provide theoretical saturation (McCracken, 1988 Strauss and Corbin, 1998). We reached theoretical saturation in the 13th interview, which did not reveal anything significantly new. We conducted one more interview to confirm theoretical saturation. We found that the 13th and 14th interviews did not add much to our understanding of the phenomenon though the interviews did provide additional support for our constructs. Discovering commonalties between different respondents helped in getting to the core of the phenomenon. The constructs included in the model were discussed by multiple interviewees. Also, as suggested by the concept of ���fit��� (Table I), only those concepts mentioned by a participant were included that fit with the substantive area under investigation (Strauss and Corbin, 1990). All constructs included in the model were deliberated by the authors following repeated readings of the transcripts. Furthermore, the feedback from respondents on summaries of their interviews provided insights into whether a construct should be included. All interviews but one were recorded and transcribed verbatim. Detailed notes were taken for one interview participant who refused recording his/her interview. The interview transcriptions were analyzed using the ATLAS.Ti software. The grounded theory methodology of systematic organization and constant comparison of data within and between interviews using open, axial, selective coding was followed. In view of the fact that the authors have a good understanding of the extant literature on risk management, a conscious attempt was made to keep this knowledge away from the ongoing research to prevent the interference of this knowledge. The primary researchers wrote down all they knew or believed to be true about the phenomenon of risk management. This made the researchers aware of their Criteria (and explanation) Step(s) taken Credibility (extent to which the results appear to be acceptable representations of the data) Participants reviewed the findings and provided feedback Transferability (extent to which the findings from one study in one context will apply to other contexts) Participants chosen based on theoretical sampling Diverse sample representing variations in type of industry, responsibilities, position level, and company size Dependability (extent to which the findings are unique to time and place the stability or consistency of explanations) Participants reflected as far back as 15-20 years Core categories existed across industries Confirmability (extent to which interpretations are the result of the participants and the phenomenon as opposed to researcher biases) Documenting all personal knowledge of phenomenon to keep presuppositions away and journal-keeping to record personal viewpoints Findings supported by quotes All findings reviewed by both researchers Integrity (extent to which interpretations are influenced by misinformation or evasions by participants) Confidentiality assurance to participants Non-threatening interviews by researcher trained in qualitative research Sources: Adapted from Flint and Mentzer (2000) and Flint et al. (2002) Table I. Evaluation criteria Supply chain risk management strategies 195

preconceived theoretical notions about the phenomena. This list was used during the drafting of the interview protocol to make sure the presuppositions were not forced into the interview questions. This list was also referred to repeatedly during data analysis to ensure that insights were coming from the data and not artifacts of preconceived notions. Wherever appropriate, Glaser���s (1978) advice was followed to use the literature as another source of data and was incorporated into the findings from the qualitative study. The interview protocol is presented in the Appendix. A list of criteria for evaluating qualitative research and our efforts to maintain rigor are presented in Table I. The criterion for ���credibility��� needs particular mention. Each interviewee was provided a copy of the transcript of his/her interview. All significant comments from the interviewees were incorporated into the data analysis. At the end of each interview, the interviewer summarized the major points from the interview and gave the interviewee an opportunity to provide feedback. All interviewees had some comments at this stage that were duly noted/recorded and incorporated during data analysis. Lastly, all interviewees received a summary of the overall findings across all interviews and provided feedback. This feedback was used to fine-tune the findings. Refining the concept of risk The interviews revealed multiple interpretations of risk. One senior executive in a leading home appliance manufacturing firm said: Risks are all those things that keep you away from the perfect path and perfect outcomes and (you) got to be able to translate (risks) into dollars somehow. Another manager remarked, ���Unintended outcomes, you know, is the risk.��� These and several other definitions provided by supply chain managers are consistent with the literature that suggests two components of risk: (1) potential losses (if the risk is realized, what losses will result and what is the significance of the consequences of the losses (Harland et al., 2003 Manuj and Mentzer, 2008 Mitchell, 1995)) and (2) likelihood of those losses (the probability of the occurrence of an event that leads to realization of the risk). Therefore, risk is the expected outcome of an uncertain event, i.e. uncertain events lead to the existence of risks. We call these uncertain events ���risk events.��� While probability and impact of losses are the most commonly discussed dimensions of risk, two more risk dimensions (speed and frequency) are important in global supply chains. The quote below illustrates the consequences of slow discovery of a risk event (defective capacitors) coupled with a three-month pipeline: We���re three levels down into the supply chain here and we design the (circuit) board, we get it contract manufactured (in China), and sometimes we���re buying the components, sometimes the contract manufacturer is buying the components. But a component supplier, their process for making capacitors went out of control. Capacitors got integrated into our boards and you know, months later, unfortunately, in this case, you���re finding field failures because, it wasn���t immediate failure, it was a failure over time. So, even though all the reliability work had been IJPDLM 38,3 196