Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
12 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION 2 Toward a Working Definition of Consumption for Environmental Research and Policy Paul C. Stern The concept of âenvironmental impacts of consumptionâ is rooted partly in environmental high politics. These roots can be discerned in a 1994 Presidential Decision Directive that first mobilized the U.S. govern- ment to pay attention to consumption as an environmental issue. The directive was issued in preparation for the International Conference on Population and Development that would be held in Cairo that October, at which it was widely expected that any U.S. initiative on controlling popu- lation growth would be met by criticism directed at American levels of consumption. The directive stated that the United States and other devel- oped countries must maintain an awareness of their disproportionate impacts on the global environment through consumption patterns that are at several times the level of developing countries. To effectively achieve the goal of marshalling an international response to population growth trends, it said that the United States must also demonstrate lead- ership by example in addressing the implications of these consumption patterns, with an aim toward reducing the negative global environmental impacts of consumption of goods and services in the United States. The U.S. Environmental Protection Agency (E.P.A.), in coordination with the Departments of Energy and Transportation and other appropriate agen- cies, was directed to develop a statement articulating U.S. strategies for reducing such negative impacts. The directive went on to give the E.P.A. responsibility for developing a research agenda to guide future policy in this area. In this political usage, âenvironmental impacts of consumptionâ ap- 12
TOWARD A WORKING DEFINITION OF CONSUMPTION 13 pears to refer to everything people do, aside from increasing their num- bers, that may harm the environment. This usage especially emphasizes what people in rich countries do. Treating the subject scientifically, how- ever, requires a more precise definition of âconsumptionâ that is accept- able across disciplines and is useful for analyzing the environmental im- pacts of human choices and actions. I discuss some specialized disciplinary meanings of consumption, the inadequate definition implicit in much recent discourse on the subject, and finally a working definition that I tentatively propose for use in environmental research and policy. SPECIALISTSâ MEANINGS OF CONSUMPTION Consumption has fairly precise meanings in several scientific com- munities that are likely to address the âglobal environmental impacts of consumption.â These meanings are in common use in their respective disciplines, whose adherents often have them in mind when discussing the environmental impacts of consumption. Unfortunately, none of these disciplinary meanings corresponds to the one in the phrase. A good way to begin to clarify thinking is to state these definitions, because they differ from what âconsumptionâ in the quoted phrase seems to mean. The Physicistsâ Meaning According to the First Law of Thermodynamics, consumption is im- possible: Matter/energy can be neither produced nor consumed. So for physicists, consumption must be translated as transformations of matter/ energy. According to the Second Law, such transformations increase entropy, and this increase in entropy, to the extent that it takes the form of pollution or of a decrease in the usefulness of the transformed resource, is part of what is meant by âenvironmental impacts of consumption.â The Economistsâ Meaning Economists define consumption as part of total economic activity: it is total spending on consumer goods and services (e.g., Samuelson and Nordhaus, 1989:969). The rest of economic activity consists of investment in capital goods. Economists also distinguish the consumption of goods and services from their production and distribution. In neither of these senses does the economistsâ usage conform to what is meant in the phrase âenvironmental impacts of consumption.â Investment has environmen- tal impacts just the way the purchase of consumer goods and services does. In fact, the activities may be physically identical, as when a truck or a computer is produced and purchasedâeither for use as capital equip-
14 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION ment in a firm or as a consumer good. And in economistsâ terms, the environmental impacts of âconsumptionâ result from production and dis- tribution as well as from economic consumption. All three processes have environmentally significant impacts, and production and distribu- tion may be more environmentally disruptive than consumption. Cer- tainly, production processes such as mining and agricultural tillage are responsible for significant pollution problems, and also significantly de- grade natural resources (in more precise economic terms, they make re- sources increasingly costly to transform for productive purposes). The economistsâ definition of consumption leads many economists to con- sider it analytically inappropriate to speak of âenvironmental impacts of consumptionâ because this phrasing artificially extracts consumption from the system of activities of which it is a part. These economists might prefer to translate the environmental impact of âconsumptionâ as the environmental impact of economic activity. This formulation reflects the systemic unity of economies and also suggests that there may be differen- tial impacts of different kinds of economic activity. For example, the impact of the average dollar invested may be different from that of the average dollar spent on consumer goods and services; different invest- ments may have different impacts; spending on goods may have a differ- ent impact from spending on services, and so forth. Confusion sometimes arises when people use economic statistics, which apply the economic definition of consumption, to analyze the envi- ronmental effects of âconsumption.â They may, for example, treat an increase in consumer spending as if it automatically indicates a propor- tional increase in environmental impact. This may or may not be so, depending on what changes in the size and types of economic consump- tion account for the increased spending. The Ecologistsâ Meaning To ecologists, green plants are (primary) producers, and humans and other animals are consumers. (Humans also âconsumeâ minerals.) Ecolo- gists define production, or net primary productivity (NPP), in terms of photosynthesis: âNPP is the amount of energy left after subtracting the respiration of primary producers (mostly plants) from the total amount of energy (mostly solar) that is fixed biologicallyâ (Vitousek et al., 1986:368). In this meaning, any organism that obtains its energy by eating is a con- sumer; human consumption corresponds to what humanity does with the estimated 40 percent of global terrestrial NPP that we âappropriateâ (Vitousek et al., 1986). It is not obvious, however, that the 40 percent estimate is a valid measure of the global environmental impact of human consumption, because human appropriation of primary productivity is
TOWARD A WORKING DEFINITION OF CONSUMPTION 15 not simply an ecological negative. Humanity transforms ecosystems, sub- stituting species that seem to meet our needs for those that do not. In the process, some species become more prevalent, and in some cases, produc- tivity increases. For example, agriculture adds nutrients to the soil and provides additional habitat for alfalfa weevils, honeybees, aphids, and the like, and for their predators and diseases. So the link between human consumption of global NPP and the âenvironmental impacts of consump- tionâ is not 1:1. The two usages are not equivalent, and their relationship is yet to be determined. Meanings in Sociology âConsumptionâ also has sociological meanings, not precisely defined, that are reflected in terms like âconsumerismâ and âconspicuous con- sumption.â In this usage, âconsumptionâ connotes what individuals and households do when they use their incomes to increase social status through certain kinds of purchases (see, e.g., Veblen, 1899; Campbell, 1987; Scitovsky, 1992; Institute for Philosophy and Public Policy, 1995). Consumption in this sense is not related in any straightforward way to environmental impact, as can be seen by looking at what may be included in conspicuous consumption. In some American subcultures, one can increase status by building an all-solar house that conspicuously con- sumes money (for architectural design, solar panels, and so forth) but that may reduce environmental impact if it decreases fossil and nuclear en- ergy consumption enough to compensate for the additional materials in the house. Similarly, a late-model luxury car may cost more money, provide more status, and yet consume less fuel and steel than an old pickup truck. The sociological definitions have different referents from those implied in the phrase âenvironmental impacts of consumptionâ because they do not distinguish environmentally benign from environ- mentally destructive consumption. It is analytically misleading to pre- sume that manifestations of consumerism are necessarily destructive to the environment. The recent phenomenon of âgreen consumerism,â which encompasses choices that are, or are believed to be, environmen- tally beneficial, illustrates the point. A POPULAR BUT INADEQUATE DEFINITION As a step toward a working definition of consumption, it may help to explicate a definition that is implicit in much popular discussion of consumption and the environment, that implies an interesting research agenda, but that ultimately provides an incomplete and misleading basis for analyzing the issue. I do not advocate accepting this definition.
16 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION Rather, I present it because it embodies some of the confusions that are common in many recent discussions of consumption and the environ- ment. The definition can be distilled from some images that commonly ap- pear in discourse about U.S. consumption and the environment: dumps filled with disposable products, plastics, and consumer packaging waste; freeways clogged with traffic that pollutes the air but barely moves; auto- mobiles and appliances junked when they might have been repaired; tracts of large, single-family homes with few occupants, but centrally air conditioned and with heated or cooled swimming pools; advertisements for products that no one seemed to want a few years ago but that soon everyone will âneedâ; air-conditioned shopping malls surrounded by acres of asphalt; and trash-lined streets and highways. In some of these images, consumers appear as acquiring and disposing of things they want but do not necessarily need; in others, they are running on a treadmill, sacrificing time with their families and friends to work increasingly long hours for money to buy things they feel they need but do not really want. The images portray excesses of resource use, waste, and material acquisi- tiveness and lives that are driven by, but ultimately unfulfilled by, mate- rial things. These images connote what participants in a recent U.S. study most often referred to as âmaterialismââa set of values that places mate- rial abundance ahead of all else (Harwood Group, 1995). It should not go without saying that these images embody a norma- tive critique of consumerist culture, based on claims that it is destructive environmentally, and in some versions of the critique, destructive socially and spiritually as well (see, e.g., Institute for Philosophy and Public Policy, 1995). Many writers on consumption and the environment believe there is too much consumption in the United States and that people should want (or should be influenced) to consume less. Interestingly, a substantial minority of Americans say they would like to earn and consume less, especially if doing so would free more time for family life and other nonmaterialist pursuits (e.g., Schor, 1991; Harwood Group, 1995). But leaving aside normative content, what does consumption mean in this discourse? The implicit definition might be stated this way: Con- sumption consists of the purchase decisions of households and what they do with their purchases. Its environmental impacts are the transformations of materials and energy that ultimately result from these activities. This definition embod- ies some assumptions about what causes the âenvironmental impact of consumptionââeach of them heuristically useful to a point, but analyti- cally flawedâand implies a research agenda. I first state the assumptions and then assess them and their research implications. Assumption 1: Individuals and households are the actors most responsible for the environmental impacts of consumption. This assumption points to the
TOWARD A WORKING DEFINITION OF CONSUMPTION 17 importance of research on individual and household behavior, particu- larly consumer behavior. It draws attention away from the activities of firms and governments, except as their activities serve or induce consum- ersâ desires. Assumption 2: Affluence, or more specifically, affluence U.S. style, is the pattern of living through which households cause environmental impacts. This assumption implies that research should focus on the extent to which households spend (rather than save) their incomes and on their patterns of spending, particularly the extent to which spending is on materials- and energy-intensive products and services. It also suggests research on consumer decisions to forego earning (and therefore consuming) in favor of other uses of their time. Assumption 3: The driving forces of anthropogenic environmental change, other than population growth, are economic growth and a set of forces acting on consumer âpreferences.â This assumption underlines the importance of research on the causes of growth in consumersâ incomes. It also directs attention to such forces as individualsâ values and worldviews as they concern material goods, social norms and interpersonal influences re- garding material possessions, the socialization of materialist values or âconsumer culture,â and market-related forces affecting consumer behav- ior, including pricing and advertising of materials- and energy-intensive products. This assumption would direct economists toward further study of how income drives consumption. Psychologists would study factors within individuals, such as values, attitudes, knowledge, and purport- edly fundamental human tendencies such as selfishness and the desire for status. Sociologists would study forces such as advertising, status com- petition, and the ideology of mastery over nature that came to ascendence in Western societies with the rise of capitalism. Assumption 4: The environmental impacts of consumption are more or less the same for all kinds of consumption. This assumption, unreasonable when made explicit, in fact underlies some popular writing on consumption and the environment. Those who define consumption in ways like the above often fail to distinguish between consumption that does and does not leave the transformed materials available for reuse (e.g., lead in auto- motive batteries vs. lead in paint) or between consumption of things of equal price that differ in the environmental consequences of producing and using them. This definition of consumption in terms of household behavior pro- vides a useful heuristic as far as it goes because it points to a coherent and pertinent set of empirical questions: What causes household income to increase? What drives rates of saving? What determines the energy- and materials-intensiveness of household spending and the use of household technologies? What policies can induce households to use their incomes
18 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION in more environmentally benign ways? There is room for all the social and behavioral sciences and for specialists in technology to do useful work on these questions. Moreover, answers to them would be valuable for modeling environmental change and perhaps for stopping or slowing undesirable change. Nevertheless, the research agenda does not do justice to the issue of âenvironmental impacts of consumption,â and the assumptions are ana- lytically flawed. The first assumption, that most consumption is directly caused by individuals and households, is simply incorrect for the United States and other affluent countries. The vast majority of energy use, releases of water and air pollutants, and many other environmentally destructive activities in the United States results directly from organiza- tional behavior rather than individual behaviorâspecifically, from the acts of corporations and governments (Gardner and Stern, 1996; Allen, Chapter 3). The most environmentally significant choices are not those that householders make, such as to purchase and then use consumer technologies, but the purchase and use choices of organizations, and or- ganizational choices about how technologies that affect the environment are designed, produced, distributed, and marketed. To presume that consumers are entirely responsible for the environmental impacts of con- sumption is to overlook most of the phenomenon. One might argue that household behavior is the ultimate driver be- cause of consumer sovereignty, but that effect is indirect and incomplete. Most people normally have weak preferences with regard to the technol- ogy used to produce what they purchase. Also, the environmental impact of production processes is typically hidden from consumers when they make choices. It would therefore be an analytical mistake to overlook organizational decisions that directly affect the energy- and materials- intensity of the economy and do so somewhat independently of con- sumer choice, for example, by determining which products are available for purchase or which industrial processes are employed to manufacture them. It might also be a practical mistake, if the goal is to reduce environ- mental impact. Systematic campaigns to help consumers understand the environmental impacts of production processes may be effective. The potential of such an approach in this area is suggested by the growing demand in the United States for âorganicâ food products, which are mar- keted as environmentally superior. The second assumption, that U.S.-style affluence is the source of envi- ronmental degradation, is better treated as a hypothesis to be analyzed than a conclusion. The focus on affluence suggests that researchers should classify patterns of living at different income levelsâstyles of affluence, frugality, poverty, and so forthâand compare their environmental ef- fects. It is particularly important to learn how these patterns or styles are
TOWARD A WORKING DEFINITION OF CONSUMPTION 19 shaped by peopleâs social, economic, and geographic contexts; how they change; and whether they can be materially influenced by acts of indi- vidual will or by policy. This assumption also suggests a sharp distinc- tion between spending and saving that may not, in fact, be environmen- tally significant. Whether consumer saving is better for the environment than consumer spending is an empirical question, the answer to which depends on what kinds of investments are made by those who hold the savings. A focus on consumer spending thus distracts attention from the environmental impacts of investment, which are intimately tied to those of economic consumption. The third assumption, about driving forces, is flawed because it fo- cuses on only a subset of the relevant driving forces of anthropogenic environmental change. Most of the important driving forces fit into five categories: population growth, economic growth, technological change, political-economic institutions, and attitudes and beliefs (National Re- search Council, 1992). By omitting technology and institutions and the forces that shape them in turn, the third assumption rules out important lines of investigation. These neglected driving forces profoundly affect human transformations of materials and energy, and altering them pro- vides ways of controlling the environmental impacts of human activity. Thus, the household-based definition of consumption is not only in- adequate for understanding but also inappropriate in policy terms: It unnecessarily narrows vision concerning the strategies available for changing consumptionâs environmental impacts. Such a definition fo- cuses attention on households and on affluence, suggesting that to solve environmental problems, individuals and households must spend (and perhaps earn) less. Aside from the fact that this conclusion is unlikely to lead to acceptable policy options, it has major substantive problems. One is that the conclusion may be overly pessimistic. There are effec- tive policy strategies that do not directly target individuals and that, often by focusing on technology and institutions, accomplish desired goals more effectively and in more acceptable ways. For example, improving emissions control technology in automobiles, a policy directed mainly at manufacturers, did more to reduce urban air pollution than any politi- cally practicable policy directed at households could have done. A broader definition of consumption might help identify such strategies and allow analysis of how much they can accomplish. Another problem is that the focus on directly changing household behavior suggests a panoply of interventions that do not work well. Some of them provoke public resistance, like President Carterâs call in 1979 to lower home heating levels in winter. Some interventions are too limited in scope because households are not the main cause of the targeted envi- ronmental problems. And others are likely to fail because household
20 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION behavior is multiply determined and the interventions target only a single element of it, such as consumerist values or a presumed lack of informa- tion on how to cut back. Changing consumer behavior directly is a viable policy strategy, but success depends on addressing the complexities of environmentally relevant household behavior and usually requires ad- dressing several barriers to change simultaneously (Gardner and Stern, 1996). TOWARD A WORKING DEFINITION OF CONSUMPTION At this stage of development of research on the environmental im- pacts of consumption, a working definition of consumption should not foreclose research on significant actors, major driving forces, their interre- lationships, or the various possible ways to control consumptionâs im- pacts. I propose the following definition for consideration: Consumption consists of human and human-induced transformations of materials and energy. Consumption is environmentally important to the extent that it makes materials or energy less available for future use, moves a biophysical system toward a different state or, through its effects on those systems, threatens human health, welfare, or other things people value.1 One might say that this is a definition of environmental consumption, as distinct from, for instance, economic con- sumption. A few points that are implicit in this definition should be stated explicitly: (1) Consumption in this sense is not solely a social or economic activity but a human-environment transaction. Its causes (driving forces) are largely economic and social, at least in advanced societies, but its effects are biophysical. The study of consumption therefore lies at the 1When consumption makes materials and energy less available for future use, it may affect the environment in various, sometimes contradictory, ways. First there is a straight- forward resource-depletion effect resulting from the fact that resources use requires energy and produces waste. Because easily accessible resources tend to be exploited first, each additional unit of the same resource tends to take more energy to extract and to produce more waste. Consumption of materials and energy can have some countervailing effects as well. For instance, a resource (e.g., natural gas) may be used in increasing amounts as a substitute for a more environmentally damaging one (e.g., oil or coal), resulting in a net environmental improvement. Also, resource depletion with its increasing economic and environmental costs may spur the development and adoption of more environmentally benign substitutes (e.g., passive solar building design), with the result that short-term envi- ronmental damage leads to long-term improvement. Because of these countervailing ef- fects, the net long-term environmental effect of materials and energy consumption is not easily determined: careful empirical analysis is required that looks at the larger social and economic system in which resource use is embedded. I am indebted to Joel Darmstadter for emphasizing these complexities.
TOWARD A WORKING DEFINITION OF CONSUMPTION 21 interface of the social and natural sciences, and seems to require their collaboration. (2) Consumption is defined by biophysical categories such as coal and carbon dioxide, forests and croplands, rather than by social catego- ries such as money or status. It follows that the appropriate units for measuring consumption are physical and biological, rather than economic or social (e.g., Odum, 1971; Cleveland et al., 1984; Wernick, Chapter 3). (3) Although consumption is defined by biophysical categories, its environmental impacts are seen through human eyes. Changes of state in biophysical systems are all environmental consequences, but not all are necessarily negative from human perspectives. The connection from en- vironmental âchangeâ to environmental âdisruptionâ or âharmâ is medi- ated by human values, and individuals may disagree about whether a particular environmental effect of consumption is one to be avoided. (4) Consumer behavior is environmentally significant consumption according to this definition only to the extent that it has environmental effects. Thus, consumer behavior may be more or less environmentally consumptive, and some of it comes close to not consuming at all. The purchase of automotive fuel is highly consumptive; by contrast, the pur- chase of computer software and time on the Internet are among the least consumptive of consumer activities, especially on a per-dollar basis. Simi- larly, producersâ economic activity may be highly consumptive, or it may not (e.g., National Research Council, 1994). Waste clean-up is undertaken to reduce the environmental impact of other economic activity, and some activities of economic producers can even reduce net consumption in the environmental sense. This can happen when a firm finds ways to use its own or another firmâs wastes as an input to production: the environmen- tal damage from waste disposal and the extraction and processing of virgin materials decreases, and economic output increases. (5) âConsumptionâ is not affected only by those who are consumers in the economistsâ sense. Producers and distributors transform materials and energy. Consumption is also affected by those whose actions indi- rectly shape the purchase of consumer goods and services, for example, by setting building codes or standards for the manufacture of equipment. Public officials are also responsible for large amounts of environmental consumption. Perhaps the most extensive public consumption is by mili- tary organizations, which use large amounts of fuel, metals, explosives, and the like and engage in large-scale transformation of ecosystems, espe- cially in wartime. The actions of military and civilian public officials can have major environmental effects. For instance, by changing their pur- chasing practices, they can affect the environment both directly and through their influence on the producers of what they purchase. (6) Consumer goods are not the only things that consume resources
22 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION and have environmental impacts. Public sector activities, services, and even investment are environmentally important consumption if they have major environmental consequences. Activities outside the market (e.g., religious ritual) can also transform materials and energy. (7) All human beings and societies, not just the affluent ones, con- sume. The drastically different quantities and qualities of consumption around the world are a matter for empirical investigation rather than for polemic. The broad definition of consumption has the advantage that it does not foreclose the study of human choices and activities that may hold keys to reducing the environmental impacts of human activity. The fol- lowing list suggests some of the social phenomena that tend to be over- looked under the narrow, popular definition of consumption but are in- cluded in the broader one, and that may be environmentally important. â¢ Changes in the structure of production and work. The environmental impact of human choices and actions can decrease without change in householdsâ preferences, incomes, or well-being if products are manufac- tured in less environmentally destructive ways (e.g., encyclopedias on line instead of on paper, fiber optic telephone lines instead of copper), and if working conditions put less stress on the environment (telecom- muting may be an example). What trends move the economy in these directions? Which structural changes are environmentally beneficial? Which policies promote or hinder these changes? â¢ Substitution of services for products. Most consumer purchases are motivated by a desire for a service or function rather than for a product in itself. People buy natural gas for heating but might get some home heat from passive solar housing design; they buy automobiles to travel but might get some of this service from well-designed mass transit. People also substitute restaurant food for home cooking, a change that may or may not, on balance, be good for the environment. What drives such trends? What are their environmental implications? â¢ Changes in household composition and patterns of life. Recent socio- demographic trends such as decreasing household size, the aging of popu- lations, and increasing labor force participation among women of child- bearing ages may have significant environmental implications. They affect demand for travel, space heating and cooling, and various other con- sumer services independently of any change in basic values or attitudes about the environment. Even though there may be no viable policies to change these trends, there may be ways to change their environmental implications if their effects on consumer behavior were better understood. â¢ Change in nonenvironmental policies. It is commonly asserted that
TOWARD A WORKING DEFINITION OF CONSUMPTION 23 the home mortgage tax credit and the interstate highway system have indirectly increased U.S. energy demand and disrupted ecological sys- tems. International trade policies have also been claimed to have envi- ronmental impacts. To what extent are such claims accurate? What other nonenvironmental policies shape the environmental consequences of hu- man choices and actions, and how can such effects be estimated in ad- vance? These examples suggest how adopting the broad definition of con- sumption may have significant advantages for guiding future research. A research agenda that illuminated the above phenomena and that also addressed the more household-focused questions about the environmen- tal implications of income growth, consumerist ideology, personal values and preferences, and the like would provide much of the knowledge needed to understand and reduce the âenvironmental impacts of con- sumptionâ in the United States and elsewhere. However, the broad definition of consumption may not be entirely satisfactory because under it, âenvironmental impacts of consumptionâ seems to be a redundancy. If consumption consists of materials and energy transformations, it automatically has environmental implications, even if the effects are not necessarily undesirable (i.e., perceived as im- pacts). The definition makes it necessary to speak of the âenvironmental impact of human choices and actionsâ (rather than of consumption) as the object of research. The relevant field of study, then, is human choices and activities that alter the biophysical environment, especially in ways widely considered undesirable. Many of these choices and activities are those of wealthy individuals and households, as the narrower definition of con- sumption presumes. But the broader definition may lead researchers in productive directions that might be missed if research looks mainly at âconsumers.â The broad definition may also be unsatisfying to some because it fails to point to policy goals for âreducing the negative global environmental impact of consumption of goods and services,â that is, for achieving one of the central objectives of sustainable development. In particular, it does not single out affluent consumers or consumerism as a source of environ- mental problems but instead leaves their roles an open question. This circumspection is actually an advantage for the purpose of informing public decisions for two reasons. First, it appropriately reflects the state of knowledge. For instance, evidence suggests that the relationship be- tween affluence and environmental degradation is not monotonic (see, e.g., Dietz and Rosa, Chapter 4). A definition that presumes that con- sumer behavior lies at the root of the environmental effects of human activity in the richer countries begs a question that is still open and draws
24 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION attention away from potentially effective strategies for addressing those effects without restricting aggregate consumer activity. The second rationale for a definition that does not presume the tar- gets of policy is that choosing such targets is inherently value laden. The definition directs research attention to human choices and actions that change biophysical systems but leaves open the question of which changes in which systems are most to be avoided. The definition emphasizes that environmental changes are more or less important depending on what people value. This formulation makes an appropriate distinction between analytical questions about the effects of human activities on biophysical systems and questions about the social meaning of those effects. Al- though the definition does not imply a policy strategy for sustainable development, it does point a way to get the understanding needed to inform policy debates. ACKNOWLEDGMENTS I express gratitude to those who offered helpful comments on an earlier draft, particularly Thomas Dietz, Emily Matthews, Eugene Rosa, Vernon Ruttan, Robert Socolow, James Sweeney, and Richard Wilk. REFERENCES Campbell, C. 1987 The Romantic Ethic and the Spirit of Modern Consumerism. London: Blackwell. Cleveland, C.J., R. Costanza, C.A.S. Hall, and R. Kaufman 1984 Energy and the U.S. economy: A biophysical perspective. Science 225:890-897. Gardner, G.T., and P.C. Stern 1996 Environmental Problems and Human Behavior. Boston: Allyn and Bacon. Harwood Group 1995 Yearning for Balance, July 1995: Views of Americans on Consumption, Materialism, and the Environment. Takoma Park, Md.: Merck Family Fund. Institute for Philosophy and Public Policy, University of Maryland 1995 The Ethics of Consumption. Special issue of Report from the Institute for Philosophy and Public Policy 15(4). National Research Council 1992 Global Environmental Change: Understanding the Human Dimensions. P.C. Stern, O.R. Young, and D. Druckman, eds. Committee on the Human Dimensions of Global Change. Washington, D.C.: National Academy Press. 1994 Assigning Economic Value to Natural Resources. Papers presented at the Workshop on Valuing Natural Capital for Sustainable Development, July 1993. Washing- ton, D.C.: National Academy Press. Odum, H.W. 1971 Environment, Power, and Society. New York: Wiley-Interscience. Samuelson, P.A., and W.D. Nordhaus 1989 Economics, 13th ed. New York: McGraw-Hill.
TOWARD A WORKING DEFINITION OF CONSUMPTION 25 Schor, J. 1991 The Overworked American: The Unexpected Decline of Leisure. New York: Basic Books. Scitovsky, T. 1992 The Joyless Economy: The Psychology of Human Satisfaction. New York: Oxford. Veblen, T. 1899 The Theory of the Leisure Class. New York: Macmillan. Vitousek, P.M., P.R. Ehrlich, A.H. Ehrlich, and P.A. Matson 1986 Human appropriation of the products of photosynthesis. BioScience 36:368-373.