یکپارچه سازی توسعه پایدار در زنجیره تامین: مورد ارزیابی چرخه زندگی در نفت و گاز و بیوتکنولوژی کشاورزی

It is widely accepted that firms play an important stewardship role in addressing sustainable development concerns. A key challenge in this role is to balance the often conflicting pressures created by sustainable development—firm-level economic performance versus environmental degradation and social disruption. Drawing on complexity theory, risk management, stakeholder theory and the innovation dynamics literature, we discuss the problems of integrating sustainable development concerns in the supply chain, specifically the applicability of life cycle assessment (LCA). Many authors have emphasized the importance of the “cradle to grave” approach of LCA in optimizing closed-loop supply chains, improving product design and stewardship. Based on two case studies (an agricultural biotechnology and an oil and gas company) with supporting data collected from key stakeholders, we argue that sustainable development pressures have increased complexities and presented ambiguous challenges that many current environmental management techniques cannot adequately address. We provide a framework that addresses these deficiencies and discuss implications for practitioners and management theory.

It is widely recognized that firms play an important stewardship role in addressing sustainable development pressures, and such concerns have become part of many companies’ operational and competitive strategies (Angell and Klassen, 1999, Bansal and Roth, 2000, Hart, 1995, Hart, 1997, Hart and Milstein, 1999, Porter and Van der Linde, 1995, Shrivastava, 1995 and Sharma and Vredenburg, 1998). A number of authors have emphasized the importance of such tools as life cycle assessment (LCA) to optimize closed-loop supply chains as well as improve product design and stewardship (e.g. Krikke et al., 2004, Sarkis, 2001 and Sroufe et al., 2000). The “cradle to grave” approach of LCA that extends throughout the supply chain represents an evolution over environmental assessments focused on firm-specific impacts and end-of-pipe analyses, and is now part of many organizations’ broader sustainable development efforts (Mihelcic et al., 2003). Such an approach is theoretically elegant when key interacting variables and boundaries of responsibilities are well understood. Unfortunately, such situations are rare, while the benefits from sustainability efforts have been elusive (Bowen et al., 2001, Hall and Vredenburg, 2003 and Walley and Whitehead, 1994). In reality, practitioners continue to grapple with how and when LCA should be applied, due to the complexities and uncertainties of environmental systems involved, imperfections of human reasoning and impossibility of ideal societal decisions (Funtowicz and Ravetz, 1992, Hertwich et al., 2000 and Allenby, 2000). In the operations management area, specific challenges of closed-loop supply chains may be intensified by complexities associated with product, remanufacturing, testing, evaluation, returns volume, timing and quality (Guide et al., 2003). Because of complexity, decision-makers are limited in what they can know (bounded rationality) and thus rational calculations cannot guarantee optional solutions (Simon, 1962 and Simon, 1969). These difficulties are exacerbated when dealing with novel and complex technologies such as genetic technology, because they involve science that has yet to be established within an accepted paradigm (Kuhn, 1970). New technologies may also create new industry structures, require new regulatory frameworks, generate consumer uncertainty (Ansoff, 1957, Martin, 1994, Nelson and Winter, 1982, Rogers, 1994 and Utterback, 1994) and suffer from ’liabilities of newness’ (Stinchcombe, 1965). In these situations, uncertainties about possible environmental, health and social impacts are more salient (Hall and Martin, 2005). More complexities can be expected when dealing with sustainable development because it involves a higher number of interacting parameters (i.e. economic, environmental and social, the popular ‘triple bottom line’ (Elkington, 1998) diagram of three overlapping circles). Indeed, the seminal definition of sustainable development (WCED, 1987, p. 43), “meeting the needs of the present generation without compromising the ability of future generations to meet their own needs” emphasized the temporal and dynamic aspect of sustainability, thus exacerbating complexity. According to Hall and Vredenburg, 2003 and Hall and Vredenburg, 2005, innovating for sustainable development is also usually more ambiguous, i.e. when it is not possible to identify key parameters or when conflicting pressures are difficult to reconcile. Such ambiguities make traditional risk assessment techniques unsuitable, as the estimation of probabilities through for example actuarial sciences, surveys, simulations and cost–benefit analysis would be based on unacceptably high degrees of imperfect information. They further argue that sustainability concerns frequently involve a wider range of stakeholders, many of whom are not directly involved with the organization. Decision-makers are thus likely to have significant difficulties in dealing with sustainable development. A better understanding of complexity and ambiguity may allow practitioners to determine the appropriateness of LCA in the extended supply chain. According to Simon (1962), a complex system is characterized by a large number of interacting parameters, and it is difficult to infer properties of the entire system. Interdependence (positive or negative) and intensity (interaction strength) alternate with time, as well as parameter importance (Ethiraj and Levinthal, 2004). Related to complexity theory is the biological concept of fitness landscape (Wright, 1932), a distribution of possible genotypes (i.e. fitness values) mapped from an organism's structure to its fitness level. Kauffman (1993) argues that a landscape can be more or less rugged depending on the distribution of fitness values and interdependences among the parts—the more complex a system, the more rugged the landscape. A number of management studies have applied Kauffman's concepts, such as Frenken (2001) for product evolution, Rivkin (2000) for firm development, Rivkin and Siggelkow (2003) and Levinthal and Warglien (1999) for organizational design, Gavetti et al. (2005) for strategic analysis, Choi et al. (2001) and Choi and Krause (in press) for supply chain management and Wolter (2005) for industrial cluster coordination. In general these studies argue that smooth landscape designs (low interdependence) result in relatively stable and predictable behaviour. Conversely, rugged landscape designs (higher interdependence such as diversification of functional teams) lead to greater exploration of possibilities of actions, at the cost of increased difficulties in coordination. Here we expand on Choi et al. (2001) and Choi and Krause (in press) application of complexity theory to supply chain dynamics by analysing its integration with sustainable development parameters. We consider sustainable development an inherently rugged landscape that requires coordination of social, environmental and economic dimensions. Environmental tools such as LCA should thus be ’connected’ with social and economic dimensions, and is only meaningful if applied as part of a decision-making process and not a “disintegrated aggregation of facts” (Hertwich et al., 2000, p. 15). The use of LCA explored in this paper thus differs from other operations and general management literature, which discusses LCA disconnected from social issues (e.g. Bovea and Wang, 2003, Geyer and Jackson, 2004, Mehalik, 2000, Mattheus, 2004 and Sarkis, 2001). A key difficulty is the decision-maker's limitations to deal with uncertainties and ambiguities, i.e. the level of interdependence among the dimensions and the degree to which key input and output parameters cannot be determined. Generally, such abilities are inversely correlated to the degree of novelty and complexity of the product or process under analysis. One technique that helps decision-makers and designers deal with such complex systems is modularization, a process that consists of identifying parameters, their role in the completion of the design and the degree of interdependences. We start this paper by reviewing the literature on complexity theory, fitness landscapes and modularity to better understand how to deal with the difficulties of complexity. We then discuss the risks and complexities of sustainable development innovation, particularly in the context of stakeholder management. Stakeholder theory has been well established in the sustainable development discourse, while more recently the complex and sometimes ambiguous nature of certain stakeholder relationships have exacerbated sustainable development efforts (Hall and Vredenburg, 2003 and Hall and Vredenburg, 2005). Next we discuss our research methodology, followed by an examination of LCA for environmental management (a relatively restricted scope of analysis) and sustainable development (a relatively more complex scenario with interrelated parameters). We then present two polar cases (Eisenhardt, 1989) to better understand issues regarding the integration of sustainable development in the supply chain through LCA. The first concerns the application of LCA in oil production from bitumen in the Athabasca Oil Sands, Canada's largest carbon-based energy source. The second case discusses the applicability of LCA to agricultural biotechnology and a company's attempts to introduce their technology in a developing country. These cases show a clear contrast between an integrative and a myopic landscape model of LCA application. In the first case, LCA is applied using cross-functional teams to assure diversity of skills, addressing economic, environmental and social concerns, and their interactions. In the second case, although there are clearly links between environmental and social concerns, LCA is conducted addressing relatively well-known environmental parameters, disregarding cross integration with social and economic factors. Drawing on both environmental science and management perspectives, we present a preliminary framework for adaptive search towards sustainability of technologies that considers the appropriateness of LCA. We conclude with implications for managers, LCA practitioners and academic research in sustainable development.

Although it has been widely recognized that firms play an important stewardship role in addressing sustainability concerns, there remains considerable difficulties implementing this role. LCA is one tool initially developed to integrate environmental concerns throughout the supply chain into corporate decision-making, and more recently expanded to include sustainability concerns. This transition to the more complex and ambiguous nature of sustainable development has however created additional challenges in terms of applicability and resource requirements. In this paper, we have drawn on complexity theory, risk management, stakeholder theory and the innovation dynamics literature to illustrate these problems, and proposed a framework to help practitioners deal with them. We considered a rugged landscape as the most appropriate approach to search for high performance when dealing with sustainable development, which presents both obvious and not-so-obvious interdependences among environmental, social and economic parameters. The three main points of the framework are identifying parameters and uncertainties, searching for interdependences and adapting through cross-functional walks. We used the concept of matrix of interactions and tasks to identify parameters and interdependences among economic, environmental and social factors. These factors analytically framed the appropriate dimensions of a potential technology towards sustainable development. The framework also helps managers determine if LCA is appropriate. For example, if there are any potential environmental or social impacts that are unknown or require specific investigation, the assessor should consider undertaking more narrowly focused studies by applying site-specific techniques before conducting a LCA. Our two cases illustrate the appropriateness of LCA under different circumstances and approaches. Both were engaged in activities with wide-ranging social and environmental impacts. Company A was engaged in incremental technical change, where complexities were relatively easily reduced and manageable, allowing for satisfactory options. Due to operational maturity, key parameters, such as technological characteristics and secondary stakeholders could be identified. Their cross-integrative LCA approach was tailored and evolved to fit their sustainability policies. In contrast, Company B was exploiting relatively new science with many technical and social uncertainties, and most importantly a high degree of interactions amongst them. Such circumstances exacerbate the need to use cross-integrative approaches, yet the company failed to recognize the implications of these interactions, and any attempts at LCA would be heavily criticized. Note that Company A's relatively successful initiatives were provoked by a previously inadequate recognition of interactions among sustainability parameters. Their policies and ability to cope with these pressures changed considerably once these dynamics were recognized, emphasizing the importance of learning and path dependencies. With the maturing of ag-biotech, perhaps Company B's LCA initiatives will be more successful, providing they appreciate the importance of parameter interactions. Thus while technological differences matter, business factors, managerial approaches, a willingness to adapt and related path dependences are also crucial. Recent trends in the literature have emphasized the importance of taking a broader perspective when dealing with product stewardship, including environmental impacts created by extended supply chain members, and LCA was explicitly developed for this purpose. However, as societal expectations shift from environmental issues to broader sustainable development concerns, isolated attempts to reduce environmental impacts are destined to provide less than optimal solutions or even counter-productive outcomes. The managerial challenge is thus to explore the interdependences amongst parameters in an attempt to identify satisfactory solutions. Our findings suggest that it is perhaps better to understand these interconnections moderately well (Company A), rather than having deep, but disconnected expertise in each area (Company B), much like Bijker's ’heterogeneous engineer’. In other words, environmental management programs that are highly specialized but disconnected may not be as effective as managing for sustainable development, which in our view is based on understanding, seeking out and exploiting strong interdependencies. We further caution that the complexities, ambiguities and idiosyncrasies of sustainable development make deductive approaches used in environmental management inadequate. Under such circumstances, Gavetti et al. (2005) recommend analogical reasoning, which is based on accumulated experience and alertness. While we are not suggesting that the discipline of environmental management is obsolete, we would suggest that it is becoming a role for middle managers or technicians, whereas the complexities and ambiguities of sustainable development will increasingly be the focus of senior operational and strategic managerial attention and frustration, as well as a rich area of academic research.