The Greening of Industrial Ecosystems. 1994.
Pp. 228-240. Washington, DC:
National Academy Press.
Industrial Ecology and Design for Environment: The Role of Universities
JOHN R. EHRENFELD
In the wake of the United Nations Conference on Environment and Development (UNCED) in Rio de Janeiro in June 1992, most press notices painted a gloomy picture of the outcome, especially the lack of implementable agreements. Unquestionably, however, UNCED represents a watershed in the continuing move to establish sustainability as the basic underlying theme for environmental policy and management strategies. Sustainable development has been defined in several ways, but all are similar to the definition appearing in the report Our Common Future (World Commission on Environment and Development, 1987), which set in motion the process leading to the Rio conference. That report defines the notion simply as "development that meets the needs of the present generation without compromising the ability of future generations to meet their own needs."
The very simplicity of this watchword belies the complexity and difficulty in achieving such a state, particularly given present societal activities that can hardly be called sustainable, whether they be in the industrialized North or the developing South. Arguments about the ability of the planet to support a growing human population have been with us since Malthus and continue to be debated at length.1 The economic determinists argue that the system is self-correcting. As soon as natural resources appear to be insufficient to satisfy current needs, new technologies will emerge, driven by economic interests and the inventive spirit of man. This approach, which assumes that the "right" set of technologies will simply show up at each historical juncture, is problematic because the environment that sustains life has certainly become impoverished along the way. Many issues are raised by this failure, but one conclusion seems clear—we, as a global society or a set of sovereign states, must invent a better way of designing and making choices
among technologies. I limit my comments here to technological change, not to the need for a concomitant fundamental change in human values and our stance toward nature and the environment.
In the past several decades, societal concerns about environmental protection have led to the creation and implementation of environmental policies, largely in the form of end-of-pipe, technical regulations designed to limit waste flows into the environment. The cost of implementation, and of large penalties imposed on firms that used inappropriate disposal practices in the past, has led to technical innovations and practices that are more effective, but not in any systematic, sustainable way. Relatively systematic, holistic frameworks for analysis and choice, such as product life cycle analysis, are finding their way into practical applications, but as research at the Massachusetts Institute of Technology indicates, quite slowly and with uncertain results (Sullivan, 1992). Other authors in this volume address a new, broader idea called industrial ecology or industrial metabolism, which includes such concepts as dematerialization, design for environment, clean technology, environmentally clean or environmentally conscious manufacturing, and life cycle analysis. All have a technological context; the need is to develop a framework that will lead to sustainable choices of technology, regardless of what it is called.
Today all such concepts are embryonic. For the sake of simplicity, I will use the term industrial ecology. It is not yet clear what we mean by industrial ecology or the related concepts, nor how we can apply them to the design and implementation of sustainable development. One view expressed by several authors is that of a largely analytic framework that serves mostly to identify and enumerate the myriad flows of materials and technological artifacts within a web of producers and consumers (Ayres, 1989; Frosch and Gallopoulos, 1989). This aspect of industrial ecology has been called industrial metabolism (Ayres, 1989), but is only one possible way of thinking about this new framework. The idea of an industrial ecology can be expanded to include the institutions that are involved in the technological evolution. If the framework does not incorporate such an institutional aspect, it is not likely to be useful in serving as a practical guide toward sustainable development. Analysis is a necessary part of the overall calculus but not a sufficient framework to guide real decisions and implementation. Elements bearing on economic, legal, political, managerial, and other social processes must also be included. Allenby (1992) refers to this as a metasystem (see Figure 1; see also Tibbs, 1992.)
Other authors in this volume expand the context for developing an industrial ecology framework. The key points include the following:
The problems appear at several scales that are often coupled: global warming, ozone depletion, urban-smog/lead poisoning, and indoor air pollution.
Potential solutions appear at different scales and are often closely coupled:
macropolicy choices such as priorities and strategies, sectoral/infrastructural (interfirm) choices, and product/process (intrafirm) choices.
The system is knowledge-intensive.
Many values are conflicting.
Existing rules and structures are conflicting and uncoordinated.
Technological change is central.
New forms of learning are necessary.
The development of such a framework is in itself a "daunting task." (Allenby, 1992, p. 63). This is particularly true if the framework is to be created out of whole cloth and by the players most involved in the very decisions that must be made. The key institutional players have become blind to the systems character of the industrial ecological world. This blindness to the complex, interlinked relationships among social units and the natural world, as well as our Western penchant for acting only after creating singular problems to be solved, has produced an atmosphere of blame and adversariness as well as policies that are shortsighted and limited in effectiveness.2 Two factors, however, support an optimistic outlook toward the development and application of industrial ecology as a means to deal with many of the contextual complexities and challenges.
First, many of the pieces that could be woven together to form an industrial ecology framework are already available in some stage of development. They need, however, to be understood and aggregated. Second, the university can and must play a central role in developing the concept of industrial ecology and institutionalizing its practice. Only the university offers the possibility of a competent institution that has not become blinded or coopted by the current policy and management decision-making system. This is not to say that the university is an ideal nursery, as its own institutional perspective is clouded by strong disciplinary barriers and jealousies and by its own political dynamics. Nevertheless, a practical industrial ecological way of thinking about policy, design, and the world is most likely to arise out of an academic setting, but only if universities are willing to enter into the broad discourse among all the players and to reconstruct the disciplines in a way that mimics the seamless web of the very world that we are attempting to understand. "Seamless web" has become a key metaphor in analyzing the development of large-scale sociotechnological systems, which systems are similar to the larger industrial ecological systems.3
The next sections of this paper present elements of an industrial ecology research program and a discussion of the roles that a university might play in such a program.
A PUTATIVE INDUSTRIAL ECOLOGY FRAMEWORK
Figure 1 shows the systems-like character of industrial ecology. Figure 2 presents another view depicting how the concept of industrial ecology might be put into effect to guide technological and other policy and management decisions. A similar concept is being developed at the Massachusetts Institute of Technology (MIT) as part of a current research program to examine chlorine-containing chemicals that form a large family of some 1:5,000 products in commerce today. The objective of the project is to develop a process to lump the chemicals into a small number of families that can be examined for potential policy options that would phaseout, restrict, mitigate, or otherwise change present patterns of production and use. In particular, the project is intended to identify the most promising opportunities for the innovation and application of clean technologies. There is no claim that this is the "right" way to capture the notion of industrial ecology in a practical framework, but it can serve as a model with which to develop a research agenda.
Making sustainable technological choices must combine the best knowledge we can get about the technical and environmental consequences of the option being considered, and our understanding of the workings of social institutions and value systems. Both domains contribute to the processes that create the future. By starting with a more or less traditional model in which the analysis comes first,
it will be necessary to map the results into a framework that can guide social choices among policy and managerial options. Given the extreme complexity of such choices and the uncertainty that is always present under these circumstances, a framework that can quickly limit the number of strategic options has much practical value.4 If such a framework can be developed, then the industrial ecological system can be divided into a set of data-gathering and analytical steps. These are followed, first, by a mapping that transforms the complex information set into a relatively parsimonious set of strategic or policy options and second, by the development of particular implementing mechanisms. Implementation, even if it proceeds according to the agreed upon strategies and mechanisms, is certainly going to introduce unanticipated consequences. Given the seriousness and potential or practical irreversibility of some of the unwanted consequences, some sort of monitoring and feedback should be included as part of the overall industrial ecology framework. Such elements have become part of modern notions of productivity and competitiveness. For example, monitoring, continuous adaptation, and change are fundamental elements of the emerging concepts of total quality management and its environmental counterpart, total quality environmental management.
The following paragraphs further define each of these steps, or stages, and indicate some of the challenges that can shape future academic and other research. I have been guided in this discussion by the work we have been doing to build an industrial ecology model of chlorine.
Data Acquisition—The concept of industrial ecology requires data on the production and consumption of materials and how these are interconnected so that a food-web-like understanding can be established. Economic data at each node in the web are important to evaluate the social impact of change. Jobs are likely to be lost or shifted along with other economic flows. Consumer satisfaction may be forgone or shifted to other technologies. Alternative production and user practices must be specified. It is important to have data that describe the industrial and other institutional organizations involved. Existing policies that constrain choice, such as applicable regulations, and that influence flows in the web, such as subsidies, must be carefully catalogued.
With only a partial discussion of the data that will be needed to construct an industrial ecology, it should be evident that this is a daunting task. Our experience with the chlorine project indicates that publicly available data are limited. Cooperation among private and public institutions will be essential to produce an adequate data base. One important research project might be to develop linkages among separate data bases using the rapidly emerging open architecture and networking capabilities of computer systems.
Analytic Transformation—The data must be manipulated and combined into useful measures for guiding choice. Input/output models such as suggested by Ayres (1992) or Duchin (1992) could be used to collapse data into a smaller set of
policy-analytic metrics. Forms of life cycle analyses that can identify and link environmental consequences of technological choice are needed. Alternative economic accounting methods that include environmental cost surrogates reflecting long-term effects and changes in inventories and quality are now being developed. These are but a few of the kinds of analytic systems employing new forms of environmental calculus that will be needed to put the concept of industrial ecology into a practical context.
Policy (Sustainability) Accounts—In a manner perhaps comparable to the use of macroeconomic accounts such as gross national product or gross domestic product in shaping macropolicy choices in most countries today, a new set of sustainability accounts might be developed. The development of such accounts is essential to clarify the idea of sustainability and to develop performance measures by which we can monitor our progress toward achieving a sustainable world. No such standards, agreed upon by more than a handful of people, exist today. Such accounts, which would comprise a much richer set of metrics than the standard economic accounts, might include the following elements:
Ecological summary, including characteristics of the production/consumer web such as total flows among sectors; materials "value in use" assessments that can assist trade-off comparisons; and others.
Environmental performance indicators, including estimates of impacts on the environment with measures of the certainty of the assessments. In all of the accounts, estimates of uncertainty are critical, particularly if the so-called precautionary principle is to become a standard part of the industrial ecological policy setting algorithm.
Technology balance sheet, indicating the maturity of existing products and processes and some measure of potential for innovation.
Infrastructure balance sheet, indicating the degree to which social and technical infrastructure is adequate to support, or can adapt to, changes.
Economic indicators, including conventional as well as environmentally enhanced measures.
Political balance sheet, including the various interests affected by the set of issues under consideration with some assessment of the distribution of power.
Strategy Mapping—This element would consist of some way of using the accounts to locate the set of issues on a policy or strategic grid. For example, Charles Perrow, in his study of complex systems (1984), developed a simple mapping process to point to social choices. He set up a matrix with catastrophic potential and cost of alternative technologies as the two axes and located technologies such as nuclear power and biotechnology on the grid. In an analogy to the precautionary principle, he selected out those technologies he considered likely to fail at some time with consequent unacceptable results. No matter what form of mapping is developed, the process will be strongly influenced by political inter-
ests. The design of this part of the overall scheme must incorporate mechanisms to resolve the differences among the many players.
Implementation Design—For each major strategic option, such as phaseout, restriction, mitigation, or substitution of clean technologies, a set of implementing mechanisms must be selected and designed. Mechanisms would include conventional policy instruments such as regulation, taxes, and market mechanisms. Given the strong emphasis on technology, new mechanisms that enhance technological change in the directions indicated by the process, without actually defining the technology, will be needed. Current experience with technology-forcing policies and managerial principles indicates that there is much room for innovation and improvement. Again, the choice among alternatives will be strongly influenced by political and other interests, and this step in the overall decision process must incorporate means for bringing in these parties and producing agreements.
Moreover, sustainability calls fundamentally for a new approach to the design of technology and for new public and private institutions. Establishing such institutions and practices must be considered in designing implementation systems.
Implementation—Implementation is largely the responsibility of public-and private-sector institutions other than the university. One activity, research and development of clean technological substitutes for current processes and products, is clearly an area in which academia can play a key and continuing role. Education is another activity that fits into the implementation phase as well: all students must be educated to be environmentally aware, and professionals must be educated to carry out the overall analysis and policymaking process. Shifting curricula toward a new paradigm will not be an easy transition for schools deeply steeped in old ways of thinking, but it is a necessary shift all the same.
Monitoring—Given the complexity of the industrial ecological system and the incremental impact of many potential polices, it is important to monitor progress toward sustainability, learning more about the notion in the process. Continuous learning will be necessary. Learning in the systems sense will require some sort of closely coupled feedback between the acting and designing parties and the world as it is affected by the actions taken. Environmental monitoring has been grossly ignored historically. New technological systems and institutions to develop and handle the data on a global basis must be developed. Both will require considerable research.
SELECTED RESEARCH EXAMPLES
The foregoing discussion points to general areas of academic research and includes a number of suggested topics for universities (see the right-hand column in Figure 2). The set of categories does not easily identify other research areas that may be cross-cutting or subjects of particular concern. The following short list is included to provide additional examples: it is not exhaustive, but only
representative of the large family of research that must be done to develop such an industrial ecology framework. The point being made is that there are unanswered questions that fit well into the academic research structure.
Data Acquisition—Amassing the quantities of data required to support a large industrial ecology system is a huge task. Many key data exist, but in many forms and places. Other important data do not exist and must be developed. A key research project might be to develop data networks that can access and combine such disparate sources of information. New forms of data bases must be designed and built to accommodate the demanding analytic needs of an industrial ecology framework. The availability of large relational data bases should avoid the task of creating data systems de novo.
Analytic Transformation—The work begun by several researchers on industrial metabolism models requires considerable further development. Ayres and Duchin (in this volume) provide examples of different approaches that can be developed to indicate and quantify relationships between material flows across industrial sectors. Research on life cycle analysis also provides a systematic framework for analyzing product and process implications.
Policy Accounts—Little progress will be made over time unless sustainable performance measures are developed. The current economic system of accounts now widely used to guide policy decisions does not capture factors deemed important for long-term decision making. Accounting and management information systems include expansions of standard macroaccounting systems used at national and international policy levels and microsystems used within the firm for management decisions (see Todd, in this volume). In addition, life cycle frameworks that can provide information on the environmental future of contemplated actions are critically important.
Strategy Mapping—New and improved processes for negotiating and agreeing on values and criteria are needed. The earlier discussion should have made clear the political context of industrial ecology. No progress is likely to be made without some consensus among the many interests involved that the policies and strategies are necessary and ultimately serve their interests. Better means and institutional frameworks for introducing science into policy dialogues and for dealing with uncertainly are high priority areas. As the ecological web grows larger and the time frame becomes longer, more and more uncertainty will show itself to the players. Finding ways to avoid stagnation and vacillation under these conditions is critical.
Implementation Design—Institutional studies of firm behavior are very important. Many alternative clean technologies exist today but are not being used by firms in spite of potential cost savings and product quality gains. Research on the reasons for such apparently irrational behavior is needed, together with other fundamental studies of innovation and design practices. Although this field is large and well established, current understanding is inadequate to suggest changes that
should produce a more sustainable result. Organizational learning is another key area, as new demands will require large shifts in farm culture and competence. A recent column in Environmental Science and Technology (Aim, 1992) notes that social science, particularly that leading to an ''understanding of how industrial firms make decisions," has been notably absent from the research program of the U.S. Environmental Protection Agency. The field of historical and sociological studies of technology offers great opportunity for understanding the broad systems dynamics of technological change and its relationship to social institutions. Recent work on large-scale systems is of special interest.
Implementation—New "clean technologies"—processes and products to replace the particularly "dirty" current technologies—are clearly needed. Chemical, fiber, and metals production technologies, for example, have evolved under a set of constraints that have created large quantities of waste. New (and, perhaps, some old) ideas are needed. Institutional forms that are more cooperative and integrate players across sectors and national boundaries need to be designed and studied.
Monitoring—Given the complex and often highly uncertain nature of the problems that will fall under an industrial ecology rubric, new forms of monitoring progress are critical. The present approach to environmental decision making often assumes relatively certain answers to the problems posed and proceeds with little provision for monitoring the actual consequences. To the extent that it occurs, monitoring frequently targets environmental proxies, such as compliance with regulations, but fails to follow the environment itself. New technologies and monitoring strategies are needed to address the daunting questions raised by many current and prospective concerns.
These are but a few elements of an academic research agenda. The research needed is substantial, calling on new alliances among research disciplines and perhaps creating new fields.
CLIMBING DOWN FROM THE TOWER
The policy and management context of industrial ecology, as presented in this paper, suggests that the university must adopt a more active stance than its traditional detached, scholarly perspective. Even the preparation of a carefully developed set of research papers presenting new analytic frameworks and implementation policies and strategies is not likely to have a significant immediate effect. Traditional academic audiences are narrow, self-centered, and organized by discipline, whereas the study of industrial ecology should be broad and involved in practice. To be effective in moving toward sustainability, the research agenda must be linked to activities designed specifically to intervene in current practices and to change institutional forms. Industrial ecology is, at least in part, a new way of thinking; it is not merely the expansion of an existing set of theories about the
world. Thus, if change is to occur, it must include cultural and institutional transformations. The university can contribute to such shifts by exercising its historical role in education and by encouraging and becoming more directly involved in public policy dialogues, a less traditional and often perilous role.
Education is one way that the new ideas and ways of thinking can be diffused into the world. It is important that universities introduce the industrial ecology notion as a fundamental framework in teaching subjects of all kinds. This notion should be part of the general thrust toward improving environmental literacy. Much of the focus on literacy has been on learning how the environment works and how important it is to human activity. I am suggesting that it is also important to include a large dose of industrial ecological context so that students begin to understand how intimately virtually everything we do is fled to the natural world.
The need to educate professionals in fields, such as business administration, where environment has been virtually ignored is now being recognized. Although such professionals will have to spend much of their time coping with existing regulatory demands, it is important to instill the broader context of industrial ecology. Otherwise, we are likely to continue to seek solutions to problems rather than to avoid them in the first place. The technological consequences of societal activities should become a more explicit part of the education of professionals heading for planning, policy, managerial, and design careers. Education of professionals already in practice is also important, as this group is most directly involved in the decisions that influence technological choice.
INTERVENING IN THE PUBLIC DIALOGUE
Industrial ecology is a manifestation of the recognition that human activities and the natural world are inextricably linked. This linkage joins the so-called objective scientific set of disciplines with the social, practical world. In that practical world, individual and societal values enter into the decision and design processes that continuously shape the future. It is impossible to conceive of a purely objective, detached way of thinking about sustainability, far from the reality of everyday activities. The most important decisions are those made by the players. The university needs to join that circle to be able to bring its knowledge and systems purview to full fruition and effectiveness. The road to sustainability, as the preparations for UNCED 1992 demonstrated, is obstructed by disagreements among the major interests. The university, building on its objective research base, can play a key role in creating an action-producing discourse among the players. Much theory points to the essential need for trust and a sense of legitimacy in consensus-building processes. The university, as an institution, appears to have
largely maintained its legitimacy as an objective and competent institution while public and private bodies have been battered of late.
By convening and facilitating dialogue on key issues, universities can forward the action, far beyond the normal path of publishing papers. Individual faculty members have long taken stands on issues. I speak of a different process in which the whole institution acts as a neutral party. Its role is not to lake a stand on the issues out of some interest beyond its legitimate bounds, but to point out the state of knowledge of the complicated world and to compare the many options always available to decision makers. We, at MIT, have had some success in this kind of endeavor (Ehrenfeld et al., 1989) and are planning to continue our attempts at helping parties now at odds to come to agreement, even if very slowly. This may well offer a model for future emulation.
Allenby, B. R. 1992. Achieving sustainable development through industrial ecology. International Environmental Affairs 4(1):62.
Aim, A. 1992. Science, social science, and the new paradigm. Environmental Science and Technology 26(6):1123.
Ayres, R. U. 1989. Industrial metabolism. Pp. 23-49 in Technology and Environment, J. H. Ausubel and H. E. Sladovich, eds. Washington, D.C.: National Academy Press.
Ayres, R. U. 1992. Toxic heavy metals: Materials cycle optimization. Proceedings of the National Academy of Sciences 89:812-820.
Callon, M. 1980. The state and technical innovation: A case study of the electric vehicle in France. Research Policy 9:358-376.
Callon, M., and J. Law. 1982. On interests and their transformation: Enrolment and counter-enrolment . Social Studies in Science 12:615-625.
Duchin, F. 1992. Industrial input-output analysis: Implications for industrial ecology. Proceedings of the National Academy of Sciences 89:851-855.
Ehrenfeld, J. R. 1990. Technology and the environment: A map or a mobius strip? Paper prepared for Toward 2000: Environment, Technology, and the New Century, a World Resources Institute symposium, Annapolis, Md., June 13-1:5, 1990.
Ehrenfeld, J. R., E. P. Craig, and J. Nash. 1989. Waste incineration: Confronting the sources of disagreement. Environmental Impact Assessment Review 9:305-315.
Frosch, R. A., and N. E. Gallopoulos. 1989. Strategies for manufacturing. Scientific American 261(3):144-152.
Hughes, T. P. 1988. The seamless web: Technology, science, et cetera, et cetera. In Technology and Social Process, B. Elliott, ed. Edinburgh, Scotland: Edinburgh University Press.
Lindblom, C. E. 1988. Democracy and the Marketplace. New York: Oxford University Press.
Meadows, D. H., D. L. Meadows, J. Randers, and W. W. Behrens III. 1972. The Limits to Growth. New York: Universe Books.
Meadows, D. H., D. L. Meadows, J. Randers, and W. W. Behrens III. 1992. Beyond the Limits. Post Mills, Vt.: Chelsea Green Publishing.
Perrow, C. 1984. Normal Accidents. New York: Basic Books.
Senge, P. 1990. The Fifth Discipline. New York: Doubleday.
Simon, J. L., and H. Kahn, eds. 1984. The Resourceful Earth: A Response to Global 2000. Oxford: Basil Blackwell.
Sullivan, M. 1992. Environmental Life Cycle Frameworks: Industry Management of Product Innovation and Environmental Impact. M.S. thesis. Technology and Policy Program, Massachusetts Institute of Technology.
Tibbs, H. B. C. 1992. Industrial ecology—An agenda for environment management. Pollution Prevention Review Spring:167-180.
World Commission on Environment and Development. 1987. Our Common Future. New York: Oxford University Press.