Linkages Home | Sustainable Consumption Home | Search | Send Feedback

Toward a Working Definition of Consumption tor Environmental Research and Policy - DRAFT
Paul C. Stern
National Research Council

The concept of "environmental impacts of consumption" is rooted partly in environmental high politics. These roots can be discerned in the 1994 Presidential Decision Directive that mobilized the U.S. government to pay attention to the issue, so named, for the first time. The directive was issued in preparation for the Intemational 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 population growth woul d be met by criticism directed at American levels of consumption. The directive stated that the US and other developed countries must maintain an awareness or their disproportionate impacts on the global environment through consumption patterns that are a t several times the level ot developing countries. To effectively achieve the goal of marshalling an international response to population growth trends, the US must also demonstrate leadership by example in addressing the implications ot these consumption patterns, with an aim toward reducing the negative global environmental impacts of consumption of goods and services in the United States. The Environmental Protection Agency, in coordination with the Departments of Energy and Transportation and other ap propriate agencies, shall develop a statement articulating US strategies for reducing such negative impacts.

The EPA was also given responsibility to develop a research agenda to guide future policy in this area, and it called on the National Research Council to organize this workshop to help it with that task. So in this political usage, "environmental impacts of consumption" appears to refer to everything people do, aside from increasing their numbers, that may harm the environment. It especially emphasizes what people in rich countries do. Treating the subject scientifically, however, requires a more precise definition of "consumption" that is acceptable across disciplines and is useful for analyzing the environmental impacts of human activity. This paper discusses some meanings of "consumption" and tentatively proposes a working definition for use in environ mental research and policy.

Specialists' Meanings ot "Consumption"

"Consumption" has fairly precise meanings in several scientific communities that are likely to address the "global enviromnental impacts of consumption." Unfortunately, none of these meanings corresponds to the one in the phrase. A good way to begin to cl arify thinking is to specify some ot the things "consumption" in the quoted phrase seems *not* to mean.

The physicists' meaning. According to the First Law of Thermodynamics, consumption is impossible: Matter/energy can be neither produced nor consumed. So for physicists, "consumption" must be translated as transformations of matter/energy. Accordin g 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 "environmental impacts of consumption."

The economists' meaning. Economists distinguish between consumption and production and between consumption and savings, or investment. Economic output (production) can be meaningfully divided into consumer goods, consumer services, and investment . There are also waste products of production and distribution, as well as post-consumer waste, which are not part of economic output (though the industries that process waste are part of GNP). In these terms, the environmental impacts of "consumption" in clude the waste products of economic production and distribution as well as of economic consumption. They also include the effects of economic production processes on natural resources, which degrade those resources or make them less available (or in more precise economic terms, increasingly costly to transform) for productive purposes. Environmental impacts may come from investment as well as consumption: investment in different things may have different enviromnental impacts, and the environmental impac t of the average dollar invested may be different trom that ot the average dollar spent. To an economist, the environmental impact of "consumption" might be translated as the environmental impact of econoJnic activity. This formulation suggests that there may be differential impacts of different kinds of economic activity, and of different uses of income (e.g., consuming vs. investing).

The ecologists' meaning. To ecologists, green plants are (primary) producers, and humans and other animals are consumers. (Humans also "consume" minerals.) In these terms, human consumption corresponds to what humanity does with the estimated 40% of global terrestrial net primary productivity (NPP) that we "appropriate" (Vitousek et al., 1986). It is not obvious, however, that the 40% estimate is a valid measure of the global environmental impact of human consumption, because human appropriation of primary productivity is not simply an ecological negative. What humanity does is transform ecosystems, and not only destroy them. For example, agriculture provides additional habitat for alfalfa weevi1s, honeybees, aphids, and the like, and for their p redators and diseases. So the link between human consumption of global NPP and the "environmental impacts of consumption" is not 1:1. The two usages are not equivalent, and their relationship is yet to be determined.

Meanings in sociology. "Consumption" also has sociological meanings, not precisely defined, that are reflected in terms like "consumerism" and "conspicuous consumption." In this usage, "consumption" connotes what individuals and households do when they use their incomes to increase social status through certain kinds of purchases. It 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 subc ultures, one can increase status by building an all-solar house that conspicuously consumes money, but reduces environmental impact by consuming less fossil and nuclear energy. Similarly, a late-model luxury car may cost more money, provide more status, a nd yet consume less fuel and steel than an old pickup truck.

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 incomp lete basis for analyzing the issue. The definition can be distilled from some images that commonly appear in discourse about U.S. consumption and the environment: dumps filled with disposable products, plastics, and consumer packaging waste; freeways clog ged with traffic that pollutes the air but barely moves; automobiles 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 "need"; air-conditioned shopping malls surrounded by acres of asphalt; and trash-lined streets and highways. In some of these images, consumers appear as a cquiring 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 in order to work increasingly long hours to get money to buy things they feel they need bu t do not really want. The images portray excesses of resource use, waste, and material acquisitiveness and lives that are driven by, but ultimately unfulfilled by material things. They connote what participants in a recent U.S. study most often referred t o as "materialism"--a set of values that place prosperity, wealth, and material abundance ahead of all else (The Harwood Group, 1995).

It should not go without saying that these images embody a normative critique of consumerist culture, based on claims that it is destructive environmentally, and in some versions of the critique, socially and spiritually as well. Many writers on consumpti on 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 say they would like to earn and consume less, especially if doin g 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: Consumption consists of the transformations of materials and energy that result from the purchase decisions of househo lds and from what they do with their purchases. This definition embodies some assumptions about what causes the "environmental impact of consumption" and implies a research agenda. Both of these are worth stating explicitly.

Assumption #1: Individuals and households are the actors most directly responsible for the environmental impacts of consumption. This assumption points to the importance of research on individual and household behavior, particularly consumer behavi or.

Assumption #2: Affluence, or more specifically, affluence US-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 th an save) their incomes and on their pattems of spending, particularly the extent to which spending is on materials- and energy-intensive products and services.

Assunption #3: The driving forces of anthropogenic environmnental change, other than population growth, are economic growth and a set of forces acting on consumer "preferences." This assumption underlines the importance of research on the causes of economic growth, particularly growth in consumers' incomes. It also directs attention to such forces as individuals' values and worldviews as they concern material goods, social norms and interpersonal influences regarding material possessions, the socia lization of materialist values or "consumer culture," and market-related forces affecting consumer behavior, including pricing and advertising of materials- and energy-intensive products. This assumption would direct economists to study income as a driver of consumption. Psychologists would study factors within individuals, such as values, attitudes, knowledge, and purportedly fundamental human tendencies such as selfishness and the desire for status. Sociologists would study forces such as advertising, s tatus competition, and the ideology of mastery over nature that came to ascendance with the rise of capitalism.

This definition of consumption in terms ot household behavior provides 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? Wh at determines the energy- and materials- intensiveness of household spending and the use of household technologies? What policies can induce households to use their incomes in more environmentally benign ways? There is room for all the social and behavior al sciences and for specialists in technology to do useful work on these questions. Moreover, answers to them would be useful for modeling environmental change and perhaps for stopping or slowing undesirable change.

Nevertheless, the definition and its research agenda are too limited to do justice to the issue of "environmental impacts of consumption," because each of the assumptions is too constraining. The first assumption, that most consumption is directly caused by individuals and households, is simply incorrect for the United States and many 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 r esults directly from organizational behavior rather than individual behavior--specifically, t'rom the acts of corporations and governments (Gardner and Stern, 1996). The most environmentally significant choices are not those that householders make, such a s to purchase and then use consumer technologies, but the purchase and use choices of organizations, and organizational choices about how technologies that affect the environment are designed, produced, distributed, and marketed. To presume that consumers are primarily responsible for the environmental impacts of consumption is to overlook most of the phenomenon. (One might argue that household behavior is the ultimate driver because of consumer sovereignty, but that effect is indirect and incomplete. (it would be an analytical mistake to overlook organizational decisions that more directly affect the energy- and materials-intensity of the economy and do so somewhat independently of consumer choice, for example, by determining which products are available for purchase or which industrial processes are employed to manufacture them.)

The second assumption, that US-style affluence is the cause of environmental degradation, is better treated as a hypothesis. The focus on affluence suggests that researchers should classify patterns of living--styles of affluence, frugality, poverty, and so forth--and compare their environmental effects. It is particularly important to learn how these patterns or styles are shaped by people's social, economic, and geographic contexts; how they change; and whether they can be materially influenced by acts of individual will or by policy.

The third assumption, about driving forces, is too constraining because it focuses 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, eco nomic growth, technological change, political-economic institutions, and attitudes and beliefs (National Research Council, 1992). By omitting technology and institutions, the third assumption rules out important lines of investigation. Both of these profo undly affect human transformations of materials and energy, and both provide important strategies for controlling the environmental impacts of human activity.

The household-based definition of consumption is also too constraining in policy terms, because it unnecessarily narrows vision concerning the strategies available for discussing consumption's environmental impacts. It focuses attention on households and on affluence, suggesting that to solve enviromnental problems. ordinary people must spend less.

Aside from the fact that this conclusion is unpopular and therefore unlikely to lead to acceptable policy options, it has major substantive problems.

One is that it may be overly pessimistic. There are effective 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 practicable policy directed at households could have done. A broader definition of consumption can h elp us to identify such strategies and explore 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 Carter's call in 1979 to lower home heating levels in winter. Some are too limited in scope because households are not the main cause of the targeted environmental problems. And others are likely to tail because household behavior is multiply determined and the interventions target only a single element of it, such as co nsumerist values or a presumed lack of information 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 addressing 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 impacts of consumption, a working definition of consumption should not foreclose research on significant actors, major driving forces, their interrelationships, or the various possible ways to control consumption's impacts. I propose the following definition for consideration: Consumption consists of human transformations of materials and energy. Consumption is environmentally important to the extent that it makes materials or energy less a vailable for future use, moves dynamically stable biophysical systems toward a different state or, through its effects on those systems, threatens human health, welfare, or other things people value. 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 humanenvironment transaction. Its causes (driving forces) are economic and social but its effects are biophysical. The study of consumption therefore lies at the interface of t he 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 fields, rather than by social categories such as money or status. It follows that the appropriate units for measuring consumption are physical and biological. rather than economic or social.

3. What people do as consumers is not necessarily consumption in the present sense. These activities may be more or less consumptive, and some may not consume at all. The pure case is the purchase of information, and some economic transactions, such as th e purchase of computer software or time on the Internet, approach this ideal. Producers' economic activity also may or may not be consumptive. Waste clean-up is undertaken to reduce the environmental impact of other economic activity, and some economic ac tivity can even reduces net consumption. This can happen when a firm finds ways to use its own or another firm's wastes as an input to production: the environmental damage from waste disposal and the extraction and processing of virgin materials decreases , and output increases.

4. "Consumption" is not done only by those who are consumers in the economists' sense. Producers and distributors also transform materials and energy.

5. Consumer goods are not the only things that consume resources and have environmental impacts. Services and even investment are environmentally important consumption if they have major environmental consequences. Noneconomic activity (e.g., religious ri tual) can also transform materials and energy.

6. All human beings and societies, not just the affluent ones, consume. The drastically different quantities and qualities ot consumption around the world are a matter for empirical investigation rather than tor polemic.

The broad definiton of consumption allows for an appropriately broad research agenda and does not foreclose the study of promising strategies for reducing the environmental impacts of human activity. The following list suggests some of the environmentally important social phenomena that tend to be overlooked under the narrow, popular definition of consumption but are included in the broader one.

• Changes in the structure of productien and work. The environmental impact of human activity can decrease without change in households' preferences, incomes, or wellbeing if products are manufactured in less environmentally destructive ways (e .g., encyclopedias on line instead of on paper, fiber optic telephone lines instead of copper), and perhaps also if working conditions put less stress on the environment (e.g., telecommuting). What trends move the economy in these directions? Which struct ural changes are environmentally beneficial? What policies promote or hinder these changes?

• 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 of it from passive solar housing des ign; they buy automobiles to travel but might get some of this service from well-designed mass transit. However, people also substitute restaurant food for home cooking, a change that may not, on balance, be good for the environment. What drives such tren ds? What are their environmental implications?

• Changes in household composition and patterns of life. Recent sociodemographic trends such as decreasing household size, the aging of the population, and increasing labor force participation among women of childbearing ages have significant e nvironmental implications. They affect demand for travel, space conditioning, and various other consumer services independently of any change in basic values or attitudes about the environment. Even though there may be no viable policies to change these t rends, there may be ways to change their environmental implications if their effects on consumer behavior were better understood.

• Change in nonenvironmental policies. It is commonly asserted that the home mortgage tax credit and the interstate highway system have indirectly increased U.S. energy demand and disrupted ecological systems. International trade policies are a lso claimed to have environmental impacts. To what extent are such claims accurate? What other nonenvironmental policies affect the environmental impact of human activity, and how can such effects be estimated in advance?

Adopting the broad definition of consumption thus to 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 environmental impli cations of income growth, consumerist ideology, personal values and preferences, and the like would provide the kinds of knowledge needed to understand and reduce the "environmental impacts of U.S. consumption."

However, the broad definition of consumption may not be entirely satisfactory because under it, "environmental impacts of consumption" seems to be a redundancy. If consumption consists of materials and energy transformations, it automatically has environm ental implications, even if the effects are not necessarily undesirable ("impacts"). The definition makes it necessary to speak of the "environmental impact of human activity" (rather than of consumption) as the object of research, and to see the study of the determinants (and biophysical effects) of consumption as the scientific field that can illuminate that object.

Gardner, G.T. and Stern, P.C.
1996 Environmental problems and human behavior. Boston: Allyn and Bacon. Harwood Group

1995 Yearning for balance. July 1995: Views of American on consumption. materialism. and the environment. Takoma Park, MD: Merck Family Fund. National Research Council

1992 Global environmental change: Understanding the human dimensions. (P.C. Stern, O.R. Young, and D. Druckman, eds.). Washington: National Academy Press.

Schor, J.
1991 The overworked American: The unexpected decline of leisure. New York: Basic Books. Vitousek et al., 1986