Environmentally Significant Consumption: Research Directions (1997)

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.

EXAMINING THE DRIVING FORCES 73 4 Examining the Driving Forces INTRODUCTION Any informed effort to address the environmental impacts of con- sumption must begin with an understanding of what causes, or drives, environmentally important consumption activities. Economics has made major contributions to understanding consumption by considering prices, budgetary constraints on choice, the costs of information about alterna- tive actions, the ability to externalize costs, and so forth. It has also emphasized the fact that consumption and production are elements of a dynamic system in which all the elements respond together to external events, so that the environmental impacts of consumption are intimately tied to those of production. These economic insights are essential for understanding the dynamics of consumption. Understanding environmentally significant consumption also re- quires the use of concepts not normally included in economic analyses. For example, economics normally treats preferences as exogenous to analyses, presuming that during the time frame of interest, preferences are constant. This assumption may not be reasonable when the analysis concerns human responses to long-delayed environmental changes such as in climate or the ozone layer, because the responses may occur over a period of several decades. In conducting such analyses, it is important to examine the possibility of change in preferences for at least two reasons. One is that preferences often change on time scales of a human generation or longer: it has been argued, for instance, that cohorts raised in an 73 

74 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION environment of affluence have different values and personal and policy preferences from cohorts raised with scarcity (Inglehart, 1990). Another reason to treat preferences as endogenous in environmental research is that information about impending environmental threats may be the sort of stimulus that causes people to reconsider their preferences. Thus, for studies of consumption and the environment, it may be important to consider processes such as preference construction and cultural change that may mediate the effects of standard economic variables. This chapter presents five brief reports from the workshop that exam- ine driving forces of consumption other than those usually addressed in economic analyses or that consider the relationships between economic forces and other factors. As in Chapter 3, the reports raise some intrigu- ing questions for research and, through their bibliographies, direct read- ers to broader related literatures. Loren Lutzenhiser’s research examines residential energy use in northern California. The analysis includes several physical and economic explanatory variables typically used in this field, such as climate, dwell- ing size and type, appliance ownership, and household size and income. It also includes some factors not usually included in energy analysis, such as race, ethnicity, and cultural assimilation among relatively recent immi- grant populations. Lutzenhiser finds that after taking climate, housing characteristics, household technology, and income into account, Hispanic and Asian households use less energy than whites, that African-Ameri- cans use more, and that the immigrant populations studied move toward the white American pattern as a function of acculturation, reflected by the language spoken in the household. The findings help address the ques- tion of how adoption of an American lifestyle alters household energy use and, through it, affects the environment. They suggest that immigrants may adopt patterns of energy use that are typically American over a generation or two. The report by Thomas Dietz and Eugene Rosa uses a multivariate analytic approach to examine the effects of two driving forces on an indi- cator of environmental impact and reveal variations that can be attributed to other forces. They analyze national-level data on carbon dioxide emis- sions and estimate the effects of levels of population and affluence. They find a nearly linear effect of population and an effect of affluence (GNP per capita) that reaches a maximum at about U.S. $10,000 and then begins to decline. When the effects of population and affluence are estimated by regression, the residual variations cover more than a 20-fold range, prob- ably attributable to national differences in technology, institutions, and other factors. Further study of the residual variation is one approach to clarifying the importance of driving forces other than population size and economic activity. 

EXAMINING THE DRIVING FORCES 75 Eugene Rosa’s analysis distinguishes measures of gross economic ac- tivity from other indicators of material well-being, analyzes the relation- ships among these other measures, and considers how they relate to an indicator of environmental change. He identifies four distinct composite indicators of nonfinancial material well-being and finds that all the afflu- ent economies studied continued to change montonically on these indica- tors through 1985, even though in some of them the oil-market events of the 1970s altered the direction of the trend in carbon emissions per capita. He concludes that the transitions in these countries reflect a shift to more service-based, postmodern economies, in which both gross domestic product and nonmonetary indicators of welfare became less tightly coupled to carbon emissions during that period. Rosa suggests that fur- ther reductions in resource consumption can be made with only limited impacts on welfare. Richard Wilk’s report considers the hypothesis that Western styles of consumption have global environmental effects because people in devel- oping countries emulate this consumption. Some scholars have inferred that exposure to Western cultural influence, through such media as ex- ported films and television programs, drives consumption patterns in developing countries where per capita income is increasing. Such emula- tion matters for environmental policy because if increasingly affluent populations in developing countries mimic affluent Western lifestyles, there would be very serious global environmental impacts. If they adopt less resource-intensive and polluting styles of affluence, however, there might be great environmental benefits. Wilk identifies several indicators of emulation, notes their serious limitations to date, and presents his ten- tative reading of the data: that Western-style consumption is not a single package that consumers everywhere accept but, rather, that people of increased means in developing countries may pursue a variety of con- sumption aspirations and lifestyles. Despite Western mass media pen- etration of developing countries, Wilk finds only weak evidence that American middle-class consumer aspirations have been uniformly ac- cepted. The question of whether there is emulation of the most environ- mentally damaging types of Western-style consumption has barely begun to be examined. The report by Willett Kempton and Christopher Payne considers major social transformations in human history and prehistory as influ- ences on both consumption (of energy and materials) and quality of life. They suggest that in the sweep of human history, increases in consump- tion have been driven by grand transformations of social structures but that these transformations, at least on some indicators such as health and leisure time, have not been associated with monotonic increases in quality of life. This analysis raises the question of whether forms of social organi- 

76 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION zation might be adopted that provide an acceptable quality of life at much lower levels of materials and energy consumption than now exist in the industrialized world. These five reports suggest some of the possibilities for investigating the effects of social and cultural phenomena on environmentally relevant consumption, either independently of standard economic variables or in interaction with them. There are, of course, many other such investiga- tions that could be conducted. In Chapter 5, we discuss some strategies for setting priorities among the vast range of possible research questions linking consumption and the environment. REFERENCE Inglehart, R. 1990 Culture Shift in Advanced Industrial Society. Princeton, N.J.: Princeton University Press. 

EXAMINING THE DRIVING FORCES 77 SOCIAL STRUCTURE, CULTURE, AND TECHNOLOGY: MODELING THE DRIVING FORCES OF HOUSEHOLD ENERGY CONSUMPTION Loren Lutzenhiser This paper reviews some alternative conceptions of household en- ergy consumption and uses an analysis of patterns of energy use in a California sample to demonstrate the joint influence of social status, eth- nicity, and material culture in the structuring of energy flows. These findings suggest that conventional models of consumption obscure the workings of sociotechnical systems, seriously limiting our ability to un- derstand the dynamics of energy consumption. Implications for scientific research and policy modeling, cross-cultural analysis, and environmental justice are also considered. SOCIAL CONSUMPTION AND ENVIRONMENTAL CHANGE Despite the current hiatus in public and policy concern about energy, the environmental impacts of energy use are increasingly clear. In fact, efforts to empirically examine, theorize, and model the dynamics and consequences of societal energy use have been pursued for more than 20 years. But understanding energy consumption is a far from straightfor- ward matter. Although it is fairly obvious that energy flows are pro- duced and shaped by human action, this consumption only occurs via a complex of fuel flows, energy-conversion technologies, and loosely coupled economic marketing/regulatory systems. And, as energy is con- sumed at many different end-use sites, and under fluctuating environ- mental conditions, the flow is determined by a fairly complex interplay of sociocultural, geographic, technological, and institutional factors. Because we lack an overarching interdisciplinary approach to such human-envi- ronment interactions (Stern 1993), efforts to understand this system have too often been narrowly focused—resulting in partial views of the system and its environmental impacts. The social sciences have produced a fairly rich body of work on the role of energy and energy technology in society (e.g., see Cottrell, 1955; Mazur and Rosa, 1974; White, 1975; Adams, 1975; Buttel, 1979; Duncan, 1978; Olsen, 1991; Humphrey and Buttel, 1982; see Rosa et al., 1988, and Lutzenhiser, 1994, for reviews). A large literature also focuses on the connections between social status and consumption in general (Veblen, 1899; Weber, 1978; Lynes, 1955; Packard, 1959; Douglas and Isherwood, 1979; Mason, 1981; Mukerji, 1983; Fussell, 1983; Bourdieu, 1984; Forty, 

78 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION 1986; Miller, 1987; McCracken, 1988; Otnes, 1988; Saunders, 1990; Warde, 1990; Burrows and Marsh, 1992), including the construction of status via the stylized consumption of food, clothing, music, language, automo- biles, housing, and appliances (Ewen, 1976, 1988; Cowan, 1989; Featherstone, 1990, 1991; Gartman, 1991). We know much less about how technology-shaping processes work in the institutional environment— e.g., how devices and machines come to have the energetic and stylistic features that they do, and how producers and consumers interact in the negotiation of design (Bijker et al., 1989; Bijker and Law, 1992). And, with a few exceptions (e.g., Uusitalo, 1983), little attention has been paid until quite recently to the linkages between culture, consumption, and the natu- ral environment [see Durning, 1992; Brown, 1989; Schnaiberg’s (1991) critique of Brown; Lutzenhiser and Hackett, 1993]. The interdisciplinary literature concerned directly with the consump- tion of energy suggests, however, that social structure and cultural prac- tice are indeed central to the structuring of energy consumption (Lutzenhiser, 1992a; Lutzenhiser and Hackett, 1993), for significant en- ergy use differences are observed between income groups (Newman and Day, 1975; Lacy, 1985; Skumatz, 1988), across life cycle stages (Frey and LaBay, 1983), and among ethnic subcultures (Kohno, 1984; Throgmorton and Bernard, 1986; Hackett and Lutzenhiser, 1991). Conservation behav- ior is also quite socially variable (Heberlein and Warriner, 1982; Dillman et al., 1983; Stern et al., 1986; Schwartz and True, 1990; Hackett and Lutzenhiser, 1991). Unfortunately, many of these studies have overlooked important housing and technology differences between social groups— “technical” variables that influence consumption. Conventional energy policy models do little better, however, often glossing over the sociocultural aspects of energy use and choosing in- stead to treat “stocks” of buildings and equipment as the molar elements of a thoroughly technical analysis. Although the weaknesses in such approaches are well known (Stern, 1984, 1986; Stern and Aronson, 1984; Archer et al., 1984; Cramer et al., 1985; Baumgartner and Midtunn, 1987; Lutzenhiser, 1992b, 1993, 1994), these models continue to dominate policy discourse and the generation of energy system inputs for environmental systems modeling. This disconnect between approaches focused exclusively on either the “social” or the “technical” aspects of energy consumption can, in fact, be overcome through a fairly straightforward synthesis. The following empirical case shows that consumption can, at once, be seen as shaped by the social allocation of buildings and equipment with energetic character- istics and by the cultural expression of energy-using behaviors. 

EXAMINING THE DRIVING FORCES 79 A SOCIOTECHNICAL ANALYSIS OF HOUSEHOLD ENERGY CONSUMPTION The data used in the analysis are from a major survey of housing, energy use, and household technology in northern California (California Energy Commission, 1986). Rather than consumption being homo- genous—as many of the simplest conventional models assume—these data show considerable variation in energy consumption across sample households (Figure 4-1), with distinct differences in consumption by sub- groups defined on the basis of both “social” (e.g., life cycle stage, wealth, ethnicity) and “ technical ” (e.g., age, type, and size of housing and appli- ances) characteristics (Table 4-1). Because the social and technical aspects of consumption are correlated in these sorts of data, a series of multivari- ate models were estimated, one of which is reported in Table 4-2. This sociotechnical model offers a good fit to the data and suggests that both the behavior of social groups and their material conditions contribute to the structuring of consumption in a variety of ways. A second-stage analysis using regression estimates and subgroup characteristics shows that various combinations of behavior, housing, and technology are re- sponsible for shaping consumption quite differently across social groups (Table 4-3). Rates of input energy waste and carbon dioxide pollution FIGURE 4-1 Annual Household Energy Consumption. Data from California Energy Commission (1986). 

80 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION TABLE 4-1 Physical and Social Variation in Energy Consumption, Northern California Households Mean consumption Cases (mBtu) SD n (%) Entire sample 129 (68) 4127 100 Building size (sq ft) < 400 69 (43) 160 4 400-599 71 (51) 311 8 600-999 83 (40) 877 21 1,000-1,499 123 (49) 1216 29 1,500-1,999 150 (52) 883 21 2,000-2,699 182 (63) 486 12 2,700-3,499 214 (91) 139 3 > 3,500 250 (149) 55 1 Housing type Single family detached 156 (65) 2357 57 Multi-family 86 (48) 1770 43 Year dwelling built 1979-84 112 (63) 651 16 1970-78 128 (69) 909 22 1960-69 138 (70) 771 19 1950-59 134 (64) 730 18 1940-49 124 (63) 376 9 pre-1940 116 (72) 690 17 Number of persons in household 1 78 (45) 796 21 2 127 (63) 1454 39 3 144 (63) 630 17 4 160 (67) 511 14 5 165 (66) 197 5 >6 175 (94) 118 3 Annual income (1986 dollars) < $10,000 93 (48) 635 15 $10,000-19,999 106 (55) 771 19 $20,000-29,999 118 (58) 772 19 $30,000-39,999 131 (60) 666 16 $40,000-49,999 139 (62) 453 11 $50,000-75,000 157 (71) 546 13 > $75,000 187 (105) 284 7 

EXAMINING THE DRIVING FORCES 81 TABLE 4-1 Continued Mean consumption Cases (mBtu) SD n (%) Race/ethnicity and language spoken at home White 130 (70) 3349 83 Black 119 (62) 154 4 Hispanic 117 (54) 143 4 Hispanic (Spanish) 95 (48) 124 3 Asian 110 (66) 138 3 Asian (other) 106 (54) 130 3 NOTE: SD = Standard deviation. were also found to be socially variable (Table 4-4). When conventional approaches focus on “typical” households and amorphous stocks of hous- ing, they fail to take these sorts of social variations in consumption into account. UNDERSTANDING THE SOCIAL NATURE OF MATERIALS SYSTEMS A fundamental reorientation of theory is needed. The material envi- ronment can usefully be seen as an evolving social system in which social status (accomplished through status-graded buildings, equipment, and behavior) is a primary determinate of energy consumption, waste, and pollution. In a system of status-graded lifestyles, volumes of energy flow provide rough measures of social standing—the poor being excluded from all but modest forms of consumption, the middle classes sustained by consumption centered largely in housing and technologies, and the wealthy empowered in a variety of ways by high levels of energy flow. Rather than the amorphous housing stock assumed in energy analysis, occupied structures actually compose an ordered artificial environment, elaborated over time, its present form reflecting the realities of topogra- phy and climate; historical access to materials; the costs of land, labor, and energy availability (a mirror of past political economy); as well as past technical knowledge and cultural preference. The built environment is a physical accretion of the products of sociotechnical change—literally embodying historical social arrangements (e.g., family size and class structure) in built forms—forms to which present occupants must behaviorally adapt. In treating buildings and 

82 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION TABLE 4-2 Regression of Annual Energy Consumption on Social, Housing, Technology, and Environmental Variables Energy Consumption (mBtu) b SE p Household characteristics N of children <18 yr (0–8) 7.5 (0.9) a N of adults (0–11) 6.8 (0.9) a African-American 14.4 (4.6) a Hispanic-English (spoken at home) –2.8 (4.6) Hispanic-Spanish –12.3 (5.1) b Asian-English –12.7 (4.9) b Asian-other –26.7 (5.1) a <$15,000 4.4 (2.2) c $15,000-34,999 –1.4 (2.4) >$50,000 15.0 (2.5) a Person(s) at home during day 7.0 (1.8) a Housing characteristics Dwelling size (1,000 sq ft) 27.6 (1.0) a Multi-family unit (attached) –17.8 (2.3) a Built after 1979 (energy building codes) –9.7 (2.6) a Building energy efficiency scale (1-6) –1.7 (0.6) b Air conditioning 13.3 (2.4) a Solar water heating –2.3 (4.9) Household technology Clothes washer 7.9 (3.6) c Clothes dryer 12.5 (3.3) a Dishwasher 10.3 (2.0) a Frost-free refrigerator 9.4 (2.2) a 2+ refrigerators 20.2 (2.4) a Freezer 12.4 (2.0) a Other appliances (0-7)d 2.4 (0.7) a Pool, hot tub, or spa 36.1 (3.5) a Environment CEC1 –3.5 (4.0) CEC2 –8.0 (3.0) b CEC4 –0.9 (2.9) CEC5 –3.7 (3.3) (Intercept) 22.0 (5.5) NOTE: b = slope of regression line, signifying mBtu consumed per unit of the independent variable; SE = standard error of b; CEC = dummy variable signifying climatic regions; mBtu = million British thermal units. ap < .001; bp < .01; cp < .05; dColor TV, computer, stereo, black and white television set, microwave, video, humidifier. 

EXAMINING THE DRIVING FORCES 83 technologies as the primary “actors” in society-environment relations, conventional models claim a fictive autonomy for physical objects—di- vorcing them from the social structures and cultural processes within which they are embedded and from which they necessarily derive. When used to inform policy, these approaches also import biases—masking important social differences in material conditions and behavior. IMPLICATIONS FOR RESEARCH AND POLICY ANALYSIS A number of basic scientific and policy research implications follow from these findings. A considerable amount of fruitful work might be done, for example, in examining empirical patterns of consumption and disaggregating their sources through time across the United States. Link- ages between energy-use patterns and the patterned consumption of other goods and services (automobiles, food, entertainment, travel, etc.) might also be explored. And, the influences of a wider range of lifestyle orienta- tions than can be captured by simple demographic categories should also be examined. Studies of consumption that compare U.S. patterns with those found elsewhere in the industrialized world would also be useful. These studies could extend to the consumption of energy “embodied” in goods and services (a significant fraction of overall consumption). It would also be valuable to inventory and compare other resource flows (water, food, paper, metal, plastic, packaging) and waste flows (garbage, sewage, at- mospheric emissions). And, a good deal of attention is overdue to the social patterning of transportation and gasoline consumption—a signifi- cant source of energy demand and environmental pollution. Policy implications also follow from the social variation in consump- tion, the persistence of some low-energy-use cultural patterns in the midst of affluence, and the failure of conventional models to capture these varia- tions. Policy-oriented research might focus on how conventional model- ing systems operate and persist, and how cultural and institutional fac- tors might be introduced to energy-policy modeling. Ethnographic work on cultural differences in consumption could shed light on the roots of persistence of low consumption levels and might suggest how durable and long-lived those patterns might be. Studies of “social traps” in hous- ing and technology—both for the poor and the relatively more affluent— might reveal policy openings and long-term problems with consumption rooted in settlement patterns and social institutions (e.g, property-owner- ship conventions, taxation, inheritance, and lending systems). The impli- cations for equity and community that follow from a more social model of built environment and energy use are also significant in a more populous, competitive, and highly engineered future. The growth of consumption 

84 TABLE 4-3 Mixture of Consumption Sources: Model Household Types White $ Older single Black $ Older single Hispanic(Spanish)$ Young/small family Persons 9 .12 Persons 9 .12 Persons 38 .44 Lifestyle 22 .29 Lifestyle 35 .47 Lifestyle 11 .13 Building 19 .25 Building 10 .13 Building 16 .18 Technology 26 .34 Technology 20 .27 Technology 22 .25 TOTAL 76 1.00 TOTAL 74 1.00 TOTAL 86 1.00 White $$ Young/small family Black $ Young/small family Hispanic(English) $ Young/small family Persons 34 .25 Persons 34 .29 Persons 38 .37 Lifestyle 18 .13 Lifestyle 36 .32 Lifestyle 20 .19 Building 37 .27 Building 16 .14 Building 20 .19 Technology 49 .35 Technology 29 .25 Technology 25 .24 TOTAL 138 1.00 TOTAL 115 1.00 TOTAL 103 1.00 White $$ Older couple Black $$ Young/small family Hispanic(Spanish) $$ Young/small family Persons 19 .14 Persons 30 .26 Persons 36 .35 Lifestyle 18 .13 Lifestyle 31 .27 Lifestyle 6 .06 Building 45 .34 Building 27 .23 Building 28 .27 Technology 51 .38 Technology 28 .24 Technology 33 .32 TOTAL 133 1.00 TOTAL 115 1.00 TOTAL 103 1.00 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION 

Hispanic(English) $$ White $$$ Young couple Young/small family Asian(Other) $$ Older/extended family Persons 19 .15 Persons 37 .32 Persons 44 .37 Lifestyle 18 .14 Lifestyle 14 .12 Lifestyle 0 .00 Building 41 .31 Building 31 .26 Building 35 .29 Technology 53 .40 Technology 34 .30 Technology 40 .34 TOTAL 130 1.00 TOTAL 116 1.00 TOTAL 119 1.00 White $$$ Young/small family Asian(English) $$ Young couple Asian(English) $$$ Young/small family Persons 36 .23 Persons 18 .25 Persons 37 .26 Lifestyle 18 .11 Lifestyle 6 .08 Lifestyle 6 .04 Building 46 .29 Building 24 .32 Building 45 .31 Technology 57 .36 Technology 27 .36 Technology 57 .39 EXAMINING THE DRIVING FORCES TOTAL 157 1.00 TOTAL 74 1.00 TOTAL 145 1.00 White $$$$ Young couple White $$$$ Young/small family White $$$$ Older/extended family Persons 19 .12 Persons 37 .19 Persons 39 .20 Lifestyle 32 .21 Lifestyle 33 .17 Lifestyle 33 .17 Building 47 .30 Building 58 .30 Building 59 .30 Technology 58 .37 Technology 66 .34 Technology 67 .34 TOTAL 156 1.00 TOTAL 193 1.00 TOTAL 197 1.00 NOTE: Units are million British thermal units and fractions of total. $ = low income; $$ = lower middle income; $$$ = upper middle income; $$$$ = high income. 85 

86 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION TABLE 4-4 Race/Ethnicity and Class Patterns of Energy Intensity, Waste, and Pollution Energy Intensity Total Energy Energy Living space energy per capita per sq ft (sq ft/person) (mBtu) (mBtu) (mBtu) Entire sample 650 126 58 100 White 685 130 62 99 < $15,000 648 99 61 105 $15,000-50,000 668 126 60 99 > $50,000 771 173 68 95 Black 555 119 55 125 < $15,000 550 103 59 131 $15,000-50,000 547 126 52 123 > $50,000 615 164 56 101 Hispanic (English) 433 117 44 116 < $15,000 361 99 44 136 $15.000-50,000 444 118 42 107 > $50,000 532 150 50 103 Hispanic (Spanish) 366 95 33 107 < $15,000 361 93 35 113 $15,000-50,000 364 98 29 99 > $50,000 569 110 49 79 Asian (English) 596 110 42 86 < $15,000 448 69 31 84 $15,000-50,000 579 109 39 87 > $50,000 727 136 56 85 China, Japan, S. Asia 422 106 30 92 < $15,000 448 87 30 99 $15,000-50,000 372 108 30 95 > $50,000 521 123 29 74 NOTE: mBtu = million British thermal units. atons of carbon. bpounds of carbon. 

EXAMINING THE DRIVING FORCES 87 Waste and Pollution Waste Waste Waste energy CO2 energy CO2 CO2 per capita per capita (mBtu) (tons)a (tons)a (mBtu) (lbs)b 68 2.8 1.5 27 2,232 70 2.9 1.5 29 2,383 54 2.3 1.2 28 2,379 68 2.8 1.5 27 2,208 93 3.8 2.0 33 2,693 65 2.8 1.4 26 2,216 57 2.4 1.2 27 2,305 69 2.9 1.5 25 2,086 86 3.6 1.9 26 2,194 62 2.5 1.3 20 1,619 51 2.2 1.1 16 1,387 63 2.5 1.3 20 1,613 79 3.1 1.7 25 1,956 50 2.1 1.1 14 1,211 49 2.1 1.1 14 1,235 52 2.1 1.1 14 1,116 54 2.6 1.3 20 1,889 62 2.5 1.3 23 1,859 41 1.7 0.9 16 1,384 61 2.4 1.3 22 1,729 74 3.1 1.6 27 2,304 60 2.4 1.3 16 1,253 48 1.9 1.0 13 1,011 60 2.4 1.3 16 1,274 72 2.9 1.5 18 1,455 

88 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION and the decline of ecosystem resources—even with the benefits of ad- vanced environmental technologies—are likely to pose serious problems for even the most prosperous peoples and places in the industrialized world. REFERENCES Adams, R. 1975 Energy and Structure: A Theory of Social Power. Austin, Tex.: University of Texas Press. Archer, D., M. Costavnzo, B. Iritani, T.F. Pettigrew, I. Walker, L.T. White 1984 Energy conservation and public policy: The mediation of individual behavior. Pp. 69-92 in W. Kempton, M. Neiman, eds., Energy Efficiency: Perspectives on Individual Behavior. Washington, D.C: ACEEE Press. Baumgartner, T., and A. Midtunn, eds. 1987 The Politics of Energy Forecasting: A Comparative Study of Energy Forecasting in West- ern Europe and North America. Oxford, England: Clarendon Press. Bijker, W., T.P. Hughes, and T. Pinch, eds. 1989 The Social Construction of Technological Systems. Cambridge, Mass.: M.I.T. Press. Bijker, W., and J. Law, eds. 1992 Shaping Technology/Building Society: Studies in Sociotechnical Change. Cambridge, Mass.: M.I.T. Press. Bourdieu, P. 1984 Distinction: A Social Critique of the Judgment of Taste. Cambridge, Mass.: Harvard University Press. Brown, L. 1989 State of the World. New York: W.W. Norton. Burrows, R., and C. Marsh, eds. 1992 Consumption and Class: Divisions and Change. London: Macmillan. Buttel, F.H. 1979 Social welfare and energy intensity: A comparative analysis of the developed market economies. In C. Unseld, D. Morrison, D. Sills, and C. Wolf, eds., Sociopolitical Efficient Energy Use Policy. Washington D.C.: National Academy of Sciences. California Energy Commission 1986 Residential Appliance Saturation Survey: Pacific Gas and Electric Company. Sacra- mento: California Energy Commission. Cottrell, F. 1955 Energy and Society: The Relation Between Energy, Social Change and Economic Devel- opment. New York: McGraw-Hill. Cowan, R.S. 1984 More Work For Mother: The Ironies of Household Technology from the Open Hearth to the Microwave. New York: Basic Books. 1989 The consumption junction: A proposal for research strategies in the sociology of technology. In W. Bijker, T.P. Hughes, and T. Pinch, eds., The Social Construction of Technological Systems. Cambridge, Mass.: M.I.T. Press. Cramer, J.C., N. Miller, P. Craig, B. Hackett, T.M. Dietz, M. Leoine, and E. Vine 1985 Social and engineering determinants and their equity implications in residential electricity use. Energy 10(12):1283-91. Dillman, D.A., E.A. Rosa, and J.J. Dillman 

EXAMINING THE DRIVING FORCES 89 1983 Lifestyle and home energy conservation in the U.S. Journal of Economic Psychology 3:299-315. Douglas, M., and B. Isherwood 1979 The World of Goods. New York: W.W. Norton. Duncan, O.D. 1978 Sociologists should reconsider nuclear energy. Social Forces 57:1-22. Durning, A. 1992 How Much is Enough? The Consumer Society and the Future of the Earth. New York: W.W. Norton. Ewen, S. 1976 Captains of Consciousness: Advertising and the Social Roots of the Consumer Culture. New York: McGraw-Hill. 1988 All Consuming Images: The Politics of Style in Contemporary Culture. New York: Basic Books. Featherstone, M., ed. 1990 Global Culture: Nationalism, Globalization and Modernity. London: Sage. 1991 Consumer Culture and Postmodernism. London: Sage. Forty, A. 1986 Objects of Desire: Design and Society 1750-1980. London: Thames and Hudson. Frey, C.J., and D.G. LaBay 1983 A comparative study of energy consumption and conservation across the life cycle. In R. Bagozzi and A. Tybout, eds., Advances in Consumer Research 10:641- 646. Fussell, P. 1983 Class. New York: Ballantine Books. Gartman, D. 1991 Culture as class symbolization or mass reification? A critique of Bourdieu’s dis- tinction. American Journal of Sociology 97:421-447. Hackett, B., and L. Lutzenhiser 1991 Social structures and economic conduct: Interpreting variations in household en- ergy consumption. Sociological Forum 6:449-470. Heberlein, T.A., and G.K. Warriner 1982 The influence of price and attitude on shifting residential electricity consumption from on to off-peak periods. Journal of Economic Psychology 4:107-130. Humphrey, C.R., and F.H. Buttel 1982 Environment, Energy and Society. Belmont, Calif.: Wadsworth. Kohno, R. 1984 Energy attitudes and behavior: A cultural comparison of Japanese and American families in the United States. Pp. 423-436 in B. Morrison and W. Kempton, eds., Families and Energy: Coping with Uncertainty. East Lansing: Michigan State Uni- versity. Lacy, M. 1985 Apparent and genuine affluence: Their relation to energy consumption. Sociologi- cal Perspectives 28:117-143. Lutzenhiser, L., and B. Hackett 1993 Social stratification and environmental degradation: Understanding household CO2 production. Social Problems 40:50-73. Lutzenhiser, L. 1992a A cultural model of household energy consumption. Energy—The International Journal 17:47-60. 

90 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION 1992b Modeling Energy Consumption: Social Contexts and Social Myths. Paper pre- sented at annual meeting of the American Sociological Association, Pittsburgh, Pa. 1993 Social and behavioral aspects of energy use. Annual Review of Energy and the Envi- ronment 18:247-289. 1994 Sociology, energy and interdisciplinary environmental science. The American So- ciologist 25:57-78. Lynes, R. 1955 The Taste-Makers. New York: Harper and Brothers. McCracken, G. 1988 Culture and Consumption: New Approaches to the Symbolic Character of Consumer Goods and Activities. Bloomington: Indiana University Press. Mason, R.S. 1981 Conspicuous Consumption: A Study of Exceptional Consumer Behavior. New York: St. Martin’s Press. Mazur, A., and E.A. Rosa 1974 Energy and lifestyle. Science 607-610. Miller, D. 1987 Material Culture and Mass Consumption. Oxford, England: Basil Blackwell. Mukerji, C. 1983 From Graven Images: Patterns of Modern Materialism. New York: Columbia Univer- sity Press. Newman, D.K., and D. Day 1975 The American Energy Consumer. Cambridge, Mass.: Ballinger. Olsen, M.E. 1991 The energy consumption turnaround and socioeconomic well-being in industrial societies in the 1980s. In L. Freese, ed., Advances in Human Ecology. Greenwich, Conn.: JAI Press. Otnes, P., ed. 1988 The Sociology of Consumption: An Anthology. New Jersey: Humanities Press. Packard, V. 1959 The Status Seekers. New York: D. McKay Co. Rosa, E.A., G. Machlis, and K. Keating 1988 Energy. Annual Review of Sociology 14:149-172. Saunders, P. 1990 A Nation of Home Owners. London: Unwin Hyman. Schnaiberg, A. 1991 The Political Economy of Consumption: Ecological Policy Limits. Paper presented at annual meeting of American Association for the Advancement of Science, Washington, D.C. Schwartz, D., and B. True 1990 What households do when electricity prices go up: An econometric analysis with policy implications. Pp. 2.121-2.130 in Proceedings of the American Council for an Energy-Efficient Economy. Washington, D.C.: ACEEE Press. Skumatz, L.A. 1988 Energy-related differences in residential target-group customers: Analysis of en- ergy usage, appliance holdings, housing, and demographic characteristics of resi- dential customers. Pp. 11.131-11.143 in Proceedings of the American Council for an Energy Efficient Economy. Washington, D.C.: ACEEE Press. 

EXAMINING THE DRIVING FORCES 91 Stern, P.C., ed. 1984 Improving Energy Demand Analysis. Panel on Energy Demand Analysis, Commit- tee on Behavioral and Social Aspects of Energy Consumption and Production, National Research Council. Washington. D.C.: National Academy Press. Stern, P.C. 1986 Blind spots in policy analysis: What economics doesn’t say about energy use. Journal of Policy Analysis and Management 5:200-227. 1993 A second environmental science: Human-environment interactions. Science 260:1897-1899. Stern, P.C., E. Aronson, J.M. Darley, D.H. Hill, E. Hirst, W. Kempton, and T.J. Wilbanks 1986 The effectiveness of incentives for residential energy-conservation. Evaluation Re- view 10:147-76. Stern, P.C., and E. Aronson, eds. 1984 Energy Use: The Human Dimension. Committee on Behavioral and Social Aspects of Energy Consumption and Production, National Research Council. New York: Freeman. Throgmorton, J.A., and M.J. Bernard, III 1986 Minorities and energy: A review of recent findings and a guide to future re- search. Pp. 7.259-7.280 in Proceedings of the American Council for an Energy Efficient Economy. Washington, D.C.: ACEEE Press. Uusitalo, L., ed. 1983 Consumer Behavior and Environmental Quality: Trends and Prospects. New York: St. Martin’s Press. Veblen, T. 1979/ The Theory of the Leisure Class. New York: Penguin Books. 1899 Warde, A. 1990 Introduction to the sociology of consumption. Sociology 24:1. Weber, M. 1978 Economy and Society. C. Wittich and G. Roth, eds. Berkeley, Calif.: University of California Press. White, L. 1975 The Concept of Cultural Systems: A Key to Understanding Tribes and Nations. New York: Columbia University Press. 

92 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION ENVIRONMENTAL IMPACTS OF POPULATION AND CONSUMPTION Thomas Dietz and Eugene A. Rosa How can we investigate the driving forces of environmentally signifi- cant consumption? The brief essays in this chapter offer a variety of theoretical and methodological answers to this question. As with most problems in science, multiple strategies will yield a more robust under- standing than any single approach. In this essay, we propose the use of statistical models of the driving forces of environmental change to under- stand impact over time and across nations. This approach is based on a venerable and robust tradition in the social sciences—that of macro-com- parative research (Bollen et al., 1993). To link this social science tradition to work in the environmental sciences, we begin with the IPAT (Impact = Population × Affluence × Technology) equation. The IPAT framework is useful for thinking about the human actions, including consumption, that drive environmental change. The formulation was first offered by Ehrlich and Holdren (Ehrlich and Holdren, 1971, 1972; Holdren and Ehrlich, 1974) in a debate with Commoner (1972a, 1972b) and has seen broad usage since then (Dietz and Rosa, 1994). IPAT is easily understood, frequently used for illustrative purposes and can discipline our thinking. It also serves as a good starting point for a statistical model of the driving forces of environmental change (Dietz and Rosa, 1994, 1997; Preston, 1995). The classical IPAT formula- tion is an accounting identity and thus must assume rather than test the effects of driving forces on environmental change. In contrast, the statisti- cal model can be used to test hypotheses about driving forces. The IPAT formulation is simply: I=P×A×T (1) where I is environmental impact, P is population, and A is affluence. The typical measure of A is per capita economic activity, so PA becomes ag- gregate, or total, economic activity. Measures for I, P, and A are used to calculate T, which is by definition I/(PA) or environmental impact per unit economic activity. Thus T represents not only technology per se, but also culture, social organization, and all facets of human life other than population and economic activity. In previous papers (Dietz and Rosa, 1994, 1997), we suggest that the IPAT idea can be reformulated into a stochastic model: I = aPbAcTde, (2) 

EXAMINING THE DRIVING FORCES 93 where a, b, c and d can be either parameters or more complex functions. In either case, they can be estimated using standard statistical procedures. The key change from the traditional IPAT approach is that an indepen- dent measure of T must be used—the researcher must specify what is meant by technology rather than solving for T as I/PA. I,P,A, and T can represent either single measured variables or vectors of measured vari- ables. The residual term e represents all variables not explicitly included in the model. This makes the residual both interesting and interpretable. It is the multiplier that represents all effects other than those specified in the model. The model is simple, systematic, and robust: simple because it incorporates key anthropogenic driving forces with parsimony, system- atic because it specifies the mathematical relationship between the driv- ing forces and their impacts, and robust because it is applicable to a wide variety of impacts. One can easily think of more complex formulations. For example, a long tradition of economic and environmental models attempts to explore the complex feedbacks among variables considered in a model. Follow- ing that tradition, a reasonable first step in elaborating our model would be to draw on the vast literature on population and development to specify equations linking population and economic growth. The realism of such models can be heightened by adding further equations to describe other causal feedbacks. Models with many equations, parameters and variables are commonplace in the econometric literature. But more elabo- rate models quickly become opaque and strain the limits of available data. Thus at this early stage in modeling the driving forces of environ- mental change we believe that the best strategy is one that begins with relatively simple models that are easy to understand and that can be disciplined by existing data sources. We also note that our formulation represents an advance over the existing literature examining the relationship between affluence and en- vironmental impacts. Most of that literature uses only one independent variable, affluence (Grossman and Krueger, 1995; Selden and Song, 1994; Shafik, 1994). In those models, environmental impact is considered a nonlinear function of affluence, as is the case in our model. But the impact of population on the environment usually is assumed to be di- rectly proportional to population size, as in the IPAT accounting equa- tion. Any nonproportional effects of population (increasing or decreasing returns to scale) are ignored. Since most debates about the driving forces of environmental change have focused on the impacts of population, we believe models that do not explore population effects will prove of lim- ited value. Thus our model, despite its admitted limitations, represents a step forward. Our goal here is to show how this model can easily be 

94 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION elaborated to answer interesting questions about consumption and the environment.1 APPLYING THE MODEL OF CONSUMPTION Different ideas about consumption can be accommodated by chang- ing the operational definitions of various terms in the model. A common, if controversial, argument about consumption suggests that humans en- gage in too much economic activity—use up too much ecological “space” (Daly and Cobb, 1989) or have too large an ecological “footprint” (Wackernagel and Rees, 1996). This definition seems to match what Stern (Chapter 2) proposes as the “ecological” or “physics” definition of con- sumption. Measures of per capita economic activity, such as per capita gross domestic product, used as A in the model, would be an appropriate specification for this interpretation of consumption. Then PA becomes aggregate economic activity—consumption defined in terms of ecological space.2 Consumption can also be taken to mean affluence per se. As societies become more affluent they not only consume more, but, in addition, their patterns of consumption may shift. This corresponds roughly to the “so- ciological” definition of consumption. As noted above, a number of recent papers examine the link between environmental impacts and affluence measured as gdp, that is, gross domestic product per capita (Grossman and Krueger, 1995; Holtz-Eakin and Selden, 1995; Selden and Song, 1994; Shafik, 1994). They find a “Kuznets” curve in which increasing affluence (consumption in this definition) leads to increasing impact until a turn- over point is reached (usually around $5000-$10,000 in per capita gross domestic product) at which point impact levels out or declines.3 Economics offers a more restricted definition of household consump- tion, defining it as the total spending on consumer goods (Samuelson and Nordhaus, 1989:969). This definition closely matches popular concerns with consumerism and implies that as consumer spending goes up, so 1Dietz and Rosa (1994) offer a more extended discussion of modeling strategies in the study of environmental impacts. 2The models we are specifying here do not take account of the spatial intensity of human activity. But when spatial intensity is an important consideration, as it is when land use is changed, it is simple to add a term to the model representing land area. For an example that applies our formulation to the problem of tropical deforestation, see Dietz et al. (1991). 3Unfortunately, as noted above, these analyses all presume directly proportional effects of population (in terms of our model they assume that b = 1). As a result, they may mis- estimate the effects of both population and affluence. 

EXAMINING THE DRIVING FORCES 95 does environmental impact. In our formulation, consumption defined as consumer spending would be modeled as a term that represents the per- cent of gross domestic product spent on consumer goods (C): I = aPbAcCfTde (3) Then the function represented by f would indicate the importance of con- sumer spending in generating environmental impact while holding over- all affluence constant. We also note, following Cramer (1995), that house- holds rather than individuals are often the key consumption units. This distinction would suggest breaking P into two terms, one for the number of households and one for average household size: I = aSgHhAcCfTde (4) where H is the number of households and S is the average household size. An example may illustrate the utility of the model. We have exam- ined CO2 emissions (in millions of metric tons of carbon per year) as a measure of I. The analysis is based on 1989 data for 111 nations. We use population size for P and gross domestic product per capita (gdp) for A. In this formulation, T is combined with e, a unique term for each nation in the analysis that combines the effects of culture, institutions, and technol- ogy per se. That is, e represents “everything else.” Thus the equation estimated is the following: I = aPbAce (5) It disaggregates consumption into three components: population size (the number of people consuming), affluence (per capita consumption in the sociological sense), and everything else. Details of our analysis are described in Dietz and Rosa (1997). The effects of population are displayed in Figure 4-2. Population has a strong impact, and there is some evidence of diseconomies of scale in that there are disproportionately large effects for the most populous na- tions. These results embarrass the argument that population has little effect, or even a beneficial effect, on the environment and lend support to ongoing concern with population growth as a driving force of environ- mental impacts. Of course, these conclusions are conditional on the cases used in the analysis. Figure 4-3 indicates that the effects of affluence on CO2 emissions level off and even decline somewhat at the very highest levels of gross domestic product per capita (gdp). We suspect that this shift is the result of structural changes in both consumption and production, including a 

96 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION FIGURE 4-2 Effects of population on CO2 emissions. Solid line represents the effect of population size relative to the geometric mean (12.3 million). Population effects are calculated at the geometric mean of gross domestic product ($1476). The curve reflects the best-fitting log-polynomial model, which has a linear re- gression coefficient of 1.123 (SE = 0.058), and a quadratic coefficient of 0.063 (SE = 0.026). The numbers in the body of the figure represent the countries used in the analysis; a key is provided in Dietz and Rosa (1997). SOURCE: Dietz and Rosa (1997). shift to a service-based economy and the ability of the more affluent econo- mies to invest in energy efficiency. (These are hypotheses that can be tested by adding the appropriate indicators to the model.) This decline in impact only occurs when per capita affluence is above $10,000.4 Seventy- five percent of the 111 nations in our sample have gdps below $5000. Thus our results suggest that for the overwhelming majority of nations, economic growth that can be anticipated for the next quarter century or so will produce production and consumption patterns that lead to increas- ing rather than declining CO2 emissions per unit GDP. Reductions in CO2 4As noted above, most economic analyses of economic growth and environment, which do not allow for nonproportional effects of population, suggest that impact declines some- where between $1,000 and $10,000 in per capita GDP (Grossman and Krueger, 1995; Shafik, 1994). The exception is Holtz-Eakin and Selden (1995) whose analysis of CO2 emissions implies a turning point of over $35,000 per capita. 

EXAMINING THE DRIVING FORCES 97 FIGURE 4-3 Effects of affluence on CO2 emissions. Solid line represents the effect of population size relative to the geometric mean ($1476). Affluence effects are calculated at the geometric mean of population (12.3 million). The curve reflects the best-fitting log-polynomial model, which has a linear regression coef- ficient of 1.484 (SE = 0.105), a quadratic coefficient of –0.152 (SE = 0.026), and a cubic coefficient of –0.070 (SE = 0.020). The numbers in the body of the figure represent the countries used in the analysis; a key is provided in Dietz and Rosa (1997). SOURCE: Dietz and Rosa (1997). emissions will not occur in the normal course of development and will have to come from targeted efforts to shift toward less carbon-intensive technologies and activities. One advantage of our approach is that it is easy to use the model to make projections under alternative scenarios. This allows comparisons with more complex models and can be used to assess policy options. Here, as an example, we use the estimated coefficients of our model to project global CO2 emissions for the year 2025. In one scenario, we as- sume that national technological multipliers (the e term) will not change over time. In the second scenario, we assume an increase in efficiency, and thus a decrease in the technology multiplier of 1 percent per year. In this assumption consumption, defined as affluence, increases but changes in production processes and the bundle of goods and services consumed lead to less impact per unit consumption. In both cases, we use the United Nations medium-case scenario for population projections and assume a 2 percent annual real growth in gdp (World Resources Institute, 1992). The 

98 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION first scenario, with no technological progress, implies global CO2 emis- sions in 2025 of 4.3 × 1010 metric tons, a 95 percent increase over 1991 emissions. A 1 percent per year increase in carbon efficiency would mean an increase of only 36 percent, to 3.0 × 1010 metric tons. To achieve a goal of stable emissions at 1991 levels in the face of growth in consumption and population, our model suggests efficiency increases would need to average about 1.8 percent per year from 1990 to 2025. While such in- creases are feasible, they will not occur without strenuous efforts. FURTHER DIRECTIONS In the interest of clarity, we have avoided complexity in this analysis. Of course, considering only two candidate driving forces, population and affluence, is not adequate for understanding environmental change. As noted above, our model can easily be expanded to assess ever more subtle and detailed hypotheses about the effects of consumption and other driv- ing forces on the environment. Equation 4 above, for example, allows an examination of the relative contribution of number of households, aver- age household size (a disaggregation of population into two components), gross domestic production (gdp) per capita, and percent of GDP spent on consumption (a disaggregation of overall GDP into consumption as de- fined by economists). As we have noted elsewhere (Dietz and Rosa, 1994), the method allows work on driving forces of global change to link directly with the substantial body of methodological, theoretical, and empirical work on macro-comparative data in the social sciences. In do- ing so, it provides a macro complement to analyses such as those of Schipper (Chapter 3) that disaggregate by usage or those of Duchin (Chap- ter 3) or Lutzenhiser (Chapter 4) that focus on the micro-level of house- holds and individuals. ACKNOWLEDGMENT We thank W. Catton, R. Dunlap, A. Ford, E. Franz, L. Hamilton, L. Kalof, and P. Stern for their comments. This work was supported in part by National Science Foundation Grants SES-9109928 and SES-9311593, by the Dean of the College of Liberal Arts at Washington State University, and by the International Institute of George Mason University. Figures 4- 2 and 4-3 were originally published in Dietz and Rosa (1997). REFERENCES Bollen, K.A., B. Entwistle, and A.S. Alderson 1993 Macrocomparative research methods. Annual Review of Sociology 19: 321-351. 

EXAMINING THE DRIVING FORCES 99 Commoner, B. 1972a A bulletin dialogue on “The Closing Circle”: Response. Bulletin of the Atomic Scientists 28(5):17, 42-56. 1972b The environmental cost of economic growth. Pp. 339-363 in R.G. Ridker, ed., Population, Resources and the Environment. Washington, D.C.: Government Print- ing Office. Cramer, J.C. 1995 Population Growth and Air Quality in California. Paper presented at the Eighth Conference of the Society for Human Ecology, South Lake Tahoe, Calif., October 13. Daly, H.E., and J.B. Cobb, Jr. 1989 For the Common Good: Redirecting the Economy Toward Community, the Environment and a Sustainable Future. Boston: Beacon Press. Dietz, T., L. Kalof, and R.S. Frey 1991 On the utility of robust and resampling estimators. Rural Sociology 56:461-474. Dietz, T., and E.A. Rosa 1994 Rethinking the environmental impacts of population, affluence and technology. Human Ecology Review 1:277-300. 1997 Effects of population and affluence on CO2 emissions. Proceedings of the National Academy of Sciences 94(1):175-179. Ehrlich, P.R., and J.P. Holdren 1971 Impact of population growth. Science 171:1212-1217. 1972 Impact of population growth. Pp. 365-377 in R.G. Ridker, ed,. Population, Re- sources and the Environment. Washington, D.C.: U.S. Government Printing Office. Grossman, G., and A. Krueger 1995 Economic growth and the environment. Quarterly Journal of Economics :353- 377. Holdren, John P., and Paul R. Ehrlich 1974 Human population and the global environment. American Scientist 62:282-292. Holtz-Eakin, D., and T.M. Selden 1995 Stoking the fires? CO2 emissions and economic growth. Journal of Public Econom- ics 57:85-101. Preston, S.H. 1995 The effect of population growth on environmental quality. Population Research and Policy Review. Samuelson, P.A., and W.D. Nordhaus 1989 Economics. New York: McGraw-Hill Book Company. Selden, T.M., and D. Song. 1994 Environmental quality and development: Is there a Kuznets curve for air pollu- tion emissions? Journal of Environmental Economics and Management 27:147-162. Shafik, N. 1994 Economic development and environmental quality: An econometric analysis. Oxford Economic Papers 46:757-773. Wackernagel, M., and W. Rees 1996 Our Ecological Footprint. Gabriola Island, B.C., Canada: New Society Publishers. World Resources Institute 1992 World Resource, 1992-1993. New York: Oxford University Press. 

100 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION CROSS-NATIONAL TRENDS IN FOSSIL FUEL CONSUMPTION, SOCIETAL WELL-BEING, AND CARBON RELEASES Eugene A. Rosa RESEARCH QUESTIONS This volume (see Chapter 2) provides a comprehensive, environmen- tally sensitive definition of consumption as consisting “. . . of human and human-induced transformations of materials and energy. Consumption is environmentally important to the extent that it makes materials or en- ergy less available for future use, moves a biophysical system toward a differ- ent state or, through its effects on those systems, threatens human health, welfare, or other things people value” (emphases added). The research re- ported here addresses two classes of threats contained in this definition: threats to atmospheric systems and threats to human well-being. Fossil fuel consumption has been the foundation of industrial pro- duction and modernity for well over a century. This consumption is pivotal in environmental importance because continued consumption of its finite stocks will make it less available for future use. But it is also pivotal because it is the principal anthropogenic source of the trace green- house gas, carbon dioxide (CO2). Carbon emissions are the smoking gun of fossil fuel consumption. Continued growth in the consumption of carbon-based fuels threatens to move the atmospheric system toward a warmer state. What have been the recent historical patterns of fossil fuel use and of the resulting CO21 loads for the leading industrial nations—the dominant consumers of fossil fuels and producers of CO2? Climate change due to continuously increasing CO2 loads could, in turn, threaten human health, well-being, or other features of social life. How can we measure well-being in a way that neither ignores its multiple domains nor relies on a single domain and single indicator, such as is the practice with aggregate economic measures? How can we assess whether the threats to well-being of increased CO2 are being realized? 1CO releases are due primarily to fossil fuel combustion and secondarily to the loss of 2 moist forests. In the case of the leading industrial nations, releases are due almost entirely to fossil fuel use. National estimates of CO2 emissions for the leading industrial nations, such as those presented here, are computed by the Carbon Dioxide Information and Analy- sis Center (CDIAC) at Oak Ridge National Laboratory (Boden et al., 1990) by applying a carbon conversion formula to national levels of fossil fuel consumption taken from United Nations compilations and then adding the minuscule amounts of CO2 due to cement pro- duction and gas flaring. 

EXAMINING THE DRIVING FORCES 101 Taken together the three foregoing questions converge to the central issue addressed here: to what extent is there an historical relationship between the two concomitants of industrialization, CO2 loads and well- being? More specifically, is there a coupling between fossil fuel con- sumption (or CO2 emissions) and well-being? Data from the 1970-1985 period clearly show that energy and economic activity, one domain of well-being, had decoupled in several wealthy industrialized countries. But what about other domains of well-being?2 Along with the carbon emissions it produces, industrialization has provided not only clear eco- nomic advantages to societies but also social and other advantages as well. The enjoyable features of modern lifestyles—characterized by the availability of a broad array of goods and services, by a remarkable geo- graphical mobility, by the power of personal climate control, as well as other things people value—depend largely upon industrial production and upon energy. Are these benefits coupled with fossil fuel use via industrialization? Knowing whether noneconomic well-being is tightly or loosely coupled to fossil fuel consumption and CO2 emissions can inform debates over broad carbon policy. If, unlike the economic domain, these other domains are tightly coupled, then policies calling for a reduction in CO2 emissions via reduced fossil fuel use will need to anticipate the costs to well-being of pursuing such a policy. On the other hand, if these do- mains, too, are decoupled from CO2, then policies to reduce fossil fuel consumption should anticipate no dramatic downturns in well-being. We address this question by examining 35-year trends in CO2 emissions and societal well-being. MEASURES OF WELFARE AND WELL-BEING Well-being is a principal inquiry of the economic sciences. The typi- cal way—indeed, virtually the only way—that aggregate welfare or well- being is measured is with summary measures of the economy derived from national accounts. Thus, measures such as gross national product (GNP) or gross domestic product (GDP) are used to assess welfare.3 Since 2These domains, such as health and lifestyle, do not always correlate highly with eco- nomic measures (see, for example, Sen, 1993). 3Economists have assessed CO impact costs using aggregate economic indicators to ad- 2 dress two questions: (i) What is the relationship between national levels of economic activ- ity and CO2 loads, and (ii) What would be the impacts to economic activity from alternative carbon policies? (See, for example, Nordhaus, 1991; Dowlatabadi and Morgan, 1993; Schelling, 1992; Peck and Teisberg, 1993; and Rothen, 1995, who provides a summary of European efforts along these same lines.) 

102 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION energy growth partially decoupled from economic growth in the 1970s and 1980s (see, for example, Alterman, 1985), it follows that energy also decoupled from well-being—at least as measured in terms of economic growth. There are significant, well-known limitations to using a single na- tional product indicator as a measure of welfare. Two inadequacies have been singled out for particular criticism: (i) that GNP and GDP summa- rize market transactions, not welfare (see for example, Sen, 1982, 1993); and (ii) that these measures ignore the external costs to “natural capital” or “ecosystem services” (Coase, 1960; Nordhaus and Tobin, 1973; Daly and Cobb, 1989; Daly and Townsend, 1992; Costanza, 1995, and a variety of important critical chapters in Costanza, 1991). With the foregoing considerations in mind our objective was to de- velop a more comprehensive measure of well-being, one that captured important domains of social life left unaddressed with economic mea- sures. To meet this objective we first identified the world’s 23 leading industrial nations.4 We then subjected 26 readily available cross-national indicators representative of key domains of social life—health, nutrition, transportation, education, culture, communications and media, and gen- eral satisfaction—to a principal factor analysis.5 The result was a four- factor solution, comprising 23 of the original 26 indicators, and represent- ing four broad domains of social life: modern lifestyle, general well-being, health and safety, and life stress. Results of the factor analyses are pre- sented in Table 4-5. MONITORING TRENDS IN CO2 LOADS AND SOCIETAL WELL-BEING To determine the degree of parallelism, or coupling, between CO2 4We confined the analyses to the leading industrial nations because they produce the lion’s share of world CO2 loads, they have the greatest policy flexibility, and it is there that we find generally reliable data of the type needed for this analysis. 5An alternative approach would be to rely on “subjective” measures of social well-being, thereby tapping into peoples’ perceptions of their life experiences. We did not pursue this option on practical and substantive grounds. First, unlike social indicator data that are routinely collected by international agencies, subjective “quality-of-life” data are only col- lected episodically and not always for a consistent set of countries. Second, an earlier literature was consistent in showing little relationship between the “objective” conditions of life and subjective satisfaction (Campbell, 1981; Campbell et al., 1976). A more recent literature, though questioning this long-standing conclusion (see, for example, Veenhoven, 1991, 1988) is inconclusive about the relationship between objective and subjective indica- tors of well-being, an ambiguity even more pronounced when focusing exclusively on the leading industrial nations (Myers and Diener, 1995), as is the case here. 

EXAMINING THE DRIVING FORCES 103 TABLE 4-5 Four-Factor Model: Rotated Solution Factor Proportion of Loading Eigenvalue Variance Explained Modern Life Style 10.24 .58 Televisions .86 Divorce rate .83 Radio receivers .82 University students .79 Cars .76 Telephones .75 Commercial vehicles .71 Secondary students .58 Cinemas –.40 Hours worked per week –.46 General Well-Being 2.23 .13 Life expectancy—males .81 Life expectancy—females .76 Books published .67 Daily protein supply .57 Physicians .47 Infant mortality rate –.69 Health and Safety 1.75 .10 Daily food supply –.83 Daily fat supply –.81 Cancer deaths –.75 Diabetes deaths –.68 Accident deaths (autos) –.62 Life Stress 1.10 .06 Pharmacists –.59 Ulcer deaths .51 Total Model .86 loads (driven by fossil fuel consumption) and societal well-being, we ex- amined trends in CO2 loads and the social well-being measures for the period 1950-1985 for the 23 industrial nations in our data set. We first examined the trend in per capita yearly CO2 emissions for each of the 23 countries, expecting to find them all on a consistently increasing, mono- tonic trend until such time as economic decoupling took place. This was the case until 1970. We then found (verified by a formal nonparametric test using the Mann-Kendall statistic) the evolution of 3 distinct trends after 1970: an increasing trend for 6 nations (Australia, Greece, Italy, Nor- way, Portugal, and Spain), a stabilizing trend for 10 nations (Austria, 

104 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION FIGURE 4-4 Carbon emissions per capita: All groups. INCARB = increasing trend in per capita CO2 emissions; STACARB = stabilizing trend in per capita CO2 emissions; DECARB = decreasing trend in CO2 emissions. Canada, Denmark, Finland, Germany, Iceland, Ireland, Japan, the Neth- erlands, and New Zealand), and a decreasing trend for seven nations (Belgium, France, Luxembourg, Sweden, Switzerland, the United King- dom, and the United States). We label the three trends INCARB, STACARB, and DECARB, respec- tively. Plots of the mean carbon loads for the groups of countries within each trend are presented in Figure 4-4. To assess whether carbon loads parallel the social aspects of well- being, we plotted summated scores based upon our factor analysis for each of the four domains of the factor solution: modern lifestyle, general well-being, health and safety, and life stress (Figure 4-5). The summated scores, by carbon grouping, are all either monotonically increasing or monotonically decreasing. With some exception for the economic lag- gards, the INCARB countries, we find that the social domains of well- being do not closely parallel the three distinct carbon trends. It appears that the coupling of well-being, whether defined narrowly or broadly, to fossil fuel consumption and the carbon emissions it produces weakened considerably in the 1970-1985 period. We interpret these findings with a Threshold-Asymptote-Decoupling (TAD) hypothesis. A threshold level of energy consumption is a prereq- uisite for a nation to reach industrial status, but beyond that threshold there is considerable flexibility in the amounts of energy needed to sus- tain or improve standards of well-being. Noneconomic measures of well- 

a Modern Lifestyle b General Well-Being 12 9 6 4.5 0 0 –6 -4.5 –12 –9 1950 1955 1960 1965 1970 1975 1980 1985 1950 1955 1960 1965 1970 1975 1980 1985 INCARB STACARB EXAMINING THE DRIVING FORCES c d DECARB Health and Safety Life Stress 7 2 5 1 3 1 0 –1 –3 –1 –5 –7 –2 1950 1955 1960 1965 1970 1975 1980 1985 1950 1955 1960 1965 1970 1975 1980 1985 FIGURE 4-5 Trends in four dimensions of societal well-being: Summated scores, all groups of countries, 1950-1985. INCARB = increasing trend in per capita CO2 emissions; STACARB = stabilizing trend in per capita CO2 emissions; DECARB = decreasing 105 trend in CO2 emissions. 

106 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION being of the type examined here, such as life expectancy, food intake, and types of death, have inherent limitations, and therefore are expected to reach some asymptote. Thus, the decoupling of energy from well-being is, in part, due to the asymptotic limitations in the well-being measures. Figure 4-6 presents the relationships of carbon emissions to well-be- ing in more detail and suggests a complex picture. Generally, the correla- tions are of much larger magnitude in the INCARB group than in the other groups, particularly in the earlier years of the time series. This finding is consistent with the TAD hypothesis. However, the correlations do not diminish with time for all groups and all indicators of well-being or drop sharply after 1970 in the DECARB countries, as might be expected if well-being was decoupling from carbon emissions during this period. Rather, the trends seem to move in different directions across indicators and country groups. These results may be interpretable in terms of prop- erties of the particular indicators or of particular countries within groups (because the groups are small, a single anomaly can have a large effect on correlations). The proper interpretation awaits a finer-grained analysis. SUMMARY OF RESULTS We have used historical trends to ask whether there is a parallelism between carbon loads and societal well-being and how closely key indica- tors of well-being couple with carbon loads. We found, in general, that well-being after 1970 did not closely track trends in per capita carbon loads in 17 of the 23 societies examined (the DECARB and STACARB groups). Parallelisms did remain between carbon loads and our well- being indicators for the INCARB group of six countries—a positive paral- lelism for the economic measure6 and the modern lifestyle and general well-being measures, on the one hand, but an inverse parallelism for our health and safety and stress measures, on the other. Because the indicators examined are well-behaved over a sizable spell of recent history (35 years) and because the types of indicators examined are not prone to abrupt or precipitous change, we can derive reasonable substantive and policy conjectures from our results. The patterns we are observing, we hypothesize, reflect a structural transformation of the ad- vanced industrial societies: from an industrial-based modernity to a ser- vice- and communication-based postmodernity.7 The three separate pat- 6These data are included as part of our larger analyses but are not included here for lack of space. 7A number of European scholars (see, for example, Mol, 1995) are promoting a theory of “ecological modernization”—whose main features are consistent with our results—to de- scribe the next stage of development for industrial societies. 

a Modern Lifestyle b General Well-Being 1 0.8 0.6 0.8 0.4 0.6 0.2 0 0.4 –0.2 0.2 –0.4 –0.6 0 –0.8 –0.2 –1 1950 1955 1960 1965 1970 1975 1980 1985 1950 1955 1960 1965 1970 1975 1980 1985 INCARB EXAMINING THE DRIVING FORCES STACARB c d Life Stress DECARB Health and Safety 0.4 0.4 0.2 0.2 0 0 –0.2 –0.2 –0.4 –0.4 –0.6 –0.6 –0.8 –0.8 –1 –1 1950 1955 1960 1965 1970 1975 1980 1985 1950 1955 1960 1965 1970 1975 1980 1985 FIGURE 4-6 Correlations of carbon dioxide emissions with four dimensions of societal well-being, all groups of countries, 1950- 1985. INCARB = increasing trend in per capita CO2 emissions; STACARB = stabilizing trend in per capita CO2 emissions; 107 DECARB = decreasing trend in CO2 emissions. 

108 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION terns of carbon loads (INCARB, STACARB, DECARB), then, reflect differ- ent stages in this transformation. On the policy side, we conjecture that broad strategies devoted to the reduction of CO2 emissions via reduced fossil fuel use in the leading industrial nations will probably not be ac- companied by significant impacts to societal well-being—at least, not over the next several decades. This conjecture presumes that reductions in fossil fuel consumption can continue to come from technical improvements and other innovations that do not interfere with material well-being, as apparently occurred between 1970 and 1985. There is some evidence, however, that the pat- tern of energy use that produced decoupling during that period has been changing. For instance, Schipper (Chapter 3) identifies ways that energy use has been shifting “from production to pleasure,” a shift that may make it more difficult to achieve future energy and emissions reductions without reducing well-being. Detailed analysis of data on carbon and well-being since 1985 can help address this issue. STRENGTHS AND SHORTCOMINGS OF THIS APPROACH The principal strength of the approach outlined here is that it pro- vides a way of addressing an essential analytic and policy question through an aggregate conceptualization of social welfare and with readily available archival data. Furthermore, it represents an effort to broaden the concept of well-being in a way that more thoroughly reflects the threats consumption may pose to human health, welfare, and other things people value. By differentiating the concept of well-being, it raises ques- tions that suggest promising directions for further research. The approach has two principal weaknesses. First, as with all aggregate measures of human activity, the aggregate data mask considerable underlying detail and countervailing processes. Second, our approach is more empirically driven than theory driven with the result that it can identify relationships with considerably greater ease than it can explain them. REFERENCES Alterman, J. 1985 A Historical Perspective on Changes in the U.S. Energy-Output Ratios. Washington, D.C.: Resources for the Future. Boden, T.A., P. Kanciruk, and M.P. Farrell 1990 Trends ’90: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center. Oak Ridge, Tenn.: Oak Ridge National Laboratory. Campbell, A. 1981 The Sense of Well-Being in America: Recent Patterns and Trends. New York: McGraw- Hill. 

EXAMINING THE DRIVING FORCES 109 Campbell, A., P. Converse, and W. Rodgers 1976 The Quality of American Life: Perceptions, Evaluations, and Satisfactions. New York: Russell Sage Foundation. Coase, R. 1960 The problem of social costs. Journal of Economics and Law 3:1-44. Costanza, R. 1991 Ecological Economics: The Science and Management of Sustainability. New York: Co- lumbia University Press. 1995 Integrated Ecological Economic Modeling and Adaptive Management of Com- plex Systems. Lecture, annual meetings of the American Association for the Ad- vancement of Science, Atlanta, Ga. February 17. Daly, H.E., and J.B. Cobb, Jr. 1989 For the Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future. Boston: Beacon Press. Daly, H.E., and K.N. Townsend, eds. 1992 Valuing the Earth: Economics, Ecology, Ethics. Cambridge, Mass.: M.I.T. Press. Dowlatabadi, H., and M.G. Morgan 1993 Integrated assessment of climate change. Science 259: 1813,1932. Mol, A.R.J. 1995 The Refinement of Production: Ecological Modernization Theory and the Chemical In- dustry. The Hague: CIP-Data Konninklijke Bibliotheek. Myers, D.G., and E. Diener 1995 Who is happy? Psychological Science 6:10-19. Nordhaus, W.D. 1991 To slow or not to slow: The economics of the greenhouse effect. The Economic Journal 101:920-937. Nordhaus, W.D., and J. Tobin 1973 Is growth obsolete? Pp. 509-532 in The Measurement of Economic and Social Perfor- mance. New York: National Bureau of Economic Research. Peck, S.C., and T.J. Teisberg 1993 Optimal CO2 Emissions Control with Partial and Full World-Wide Cooperation: An Analysis Using CETA. Palo Alto, Calif.: Electric Power Research Institute. Rothen, S.M. 1995 The Greenhouse Effect in Economic Modeling: A Critical Survey. Dübendorf, Switzer- land: EWAG (Swiss Federal Institute for Environmental Science and Technol- ogy). Schelling, T.C. 1992 Some economics of global warming. American Economic Review 82:1-14. Sen, A. 1982 Choice, Welfare, and Measurement. Oxford, England: Basil Blackwell. 1993 The economics of life and death. Scientific American 208(5):40-47. Stern, P., O.R. Young, and D. Druckman, eds. 1992 Global Environmental Change: Understanding the Human Dimension. Committee on the Human Dimensions of Global Change, National Research Council. Washing- ton, D.C.: National Academy Press. Veenhoven, R. 1988 The utility of happiness. Social Indicators Research 20:333-354. 1991 Is happiness relative? Social Indicators Research 24:1-34. 

110 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION EMULATION AND GLOBAL CONSUMERISM Richard R. Wilk There are good reasons for concern about the environmental impacts of 5 billion people consuming at the level of the developed countries of Europe and North America. Given high economic growth rates in many parts of the developing world, as well as the rapid spread of electronic media, advertising, and consumer goods, we must ask what kind of con- suming future we can expect in areas that are now constrained by poverty and isolation. If everyone develops a desire for the Western high- consumption lifestyle, the relentless growth in consumption, energy use, waste, and emissions may be disastrous. It is also possible, however, that each country, region, or ethnic group may maintain different aspirations, definitions of living standards, and consumption goals. Then we could expect a high degree of diversity in consumer demand and, perhaps, much more moderate long-term levels of consumption, even with more equal distribution of income. Of course, it is also possible that a large part of the developing world will never achieve the threshold income levels necessary to consume large amounts of durables, luxuries, or services, whatever their aspirations. The choice of different scenarios for the consumption trajectories of the developing world hinges partially on the issue of emulation. Do people of other cultures find the Western high-consumption model at- tractive? Or do their own cultures offer strong alternative values that make foreign models less attractive? Are some groups or cultures more likely to emulate the West than others? What cultural, social, and eco- nomic forces promote high-consumption lifestyles? The strongest form of emulation is often labeled “cultural imperial- ism.” This theory contends that a combination of Western control of mass media and improved advertising, along with falling trade barriers and the spread of industrial capitalism, will inevitably lead the developing world into emulative forms of consumption (Tomlinson, 1991; Rassuli and Hollander, 1986). There are various moral positions on cultural im- perialism; some see it leading to economic freedom, while others consider it a malign form of brainwashing and false consciousness (Ewen, 1988; Horowitz, 1988). Many social scientists reject cultural imperialism and contend that instead of increasing centralization and homogenization, the next century will be dominated by new forms of nationalism, localism, and cultural fundamentalism that will challenge both the economic and cultural hegemony of the West (Foster, 1991). There is some empirical evidence for both processes; some forms of localization are concurrent 

EXAMINING THE DRIVING FORCES 111 with other kinds of globalization; heterogeneity and homogeneity both seem to be increasing in different sectors and at different scales (Fried- man, 1990; Hannerz, 1987; Featherstone, 1990; Wilk, 1995; Drummond and Patterson, 1988; Belk and Dholakia, 1995). Recent historical and social-scientific research on consumption has produced a great deal of empirical data and many excellent case studies that bear directly on the issue of consumer emulation. Studies of the growth of consumer culture in post-World War II Japan (Tobin, 1992), France (Kuisel, 1993), and Austria (Wagnleitner, 1994) all argue that emu- lation is not at all mechanical or inevitable but is, instead, a temporary product of specific political policies, trade practices, and cultural influ- ences. The explosion of consumer demand in China during the last decade has been used both as an example of cultural imperialism and of au- tonomy and increasing diversity in consumer culture. Consumer aspira- tions have changed several times during the last 20 years, and compari- sons show that the mixture of goods consumed in China at a particular level of income is quite different from that found in other Asian countries (Sklair, 1994). The case studies demonstrate that people, in general, tend to emulate local elites rather than following a single global generic “Western” con- sumer model. Sometimes people explicitly reject foreign models instead of emulating them. Periods of emulation may alternate with intervals where global or international standards are rejected in favor of local goods or styles (e.g., Andrae and Beckman, 1985; Appadurai, 1986, 1990; Fried- man, 1994) The influence of Western media and the advertising of global brands on actual consumer behavior and aspirations is still not clearly under- stood, and the “cultivation effect” of television is weak or highly variable in cross-national data (Ware and Dupagne, 1994; Liebes and Katz, 1990; Moore, 1993). The few comparative cross-cultural studies of consumer aspirations and values are difficult to interpret. Technocratic, urban, highly educated groups in many parts of the world show some increasing commonalities in aspirations and cultural beliefs. But are these sectors the leaders of a new wave of consumerism or small Western cultural enclaves of technocratic and academic “cosmopolitans” (Hannerz, 1990; Belk, 1988; James, 1993)? Methodological problems (translation, sam- pling) also make these survey results equivocal (Holt, 1994). The precise linkage between Westernized values and the consumption of Western goods is also not well established. Economists find that the determinants of national demand, and the amount saved and invested instead of spent on consumer goods, are complex. The share of increased income spent on major consumption categories, such as food, durables, and housing, var- 

112 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION ies widely at the same income level, as does the savings rate (Lluch et al., 1977; Gereffi and Korzeniewicz, 1994). Even the historical development and present fluorescence of mass-consumer society in the West is poorly understood, despite a wealth of new studies (e.g., Mintz, 1985; Brewer and Porter, 1993; Cross, 1993; Richards, 1991; Benson, 1994; Tierstin, 1993; Craik, 1994; McCracken, 1988; Csikszentimihaly and Rochberg-Halton, 1981). Historians tend to agree that the growth in Western consumer demand resulted from a break- down of rigid class hierarchies and a relaxation of religious inhibitions on conspicuous consumption. Some authors stress the emergence of a “ro- mantic ethic” (Campbell, 1987) or a trend toward cultivating health and self-improvement (Lears, 1989). Recent work has shifted away from “so- cial emulation” or class-competition models of consumer demand, to- ward a focus on communication, nationalism, advertising, and the growth of markets and retailing. Because all of these trends are occurring in developing countries, we can expect a similar growth of consumer cul- ture, even in the absence of any specific form of emulation. So, even without emulation, consumption levels in developing countries may dra- matically increase. Judging the emulation hypothesis is premature because social sci- ence, in general, still lacks a well-established general theory of consump- tion. Each discipline tends to focus on consumption from its own narrow perspective (though see Miller, 1995). More synthetic work is being done in hybrid fields like consumer research, cultural studies, and social his- tory (Sherry, 1995). Studies of household decision making in several disciplines have been especially promising, because most major invest- ment and consumer decisions are made at the household level. There is considerable constructive debate on models of intra-household bargain- ing, gender, and power that have direct implications for understanding consumption, spending, and savings behavior (e.g., Phipps and Burton, 1995; Folbre, 1994). Given the importance of predicting future global demand for con- sumer goods, energy, water, food, and other resources, we need to better establish the determinants of consumer behavior. While there is a good deal of empirical data available that bear on these issues, there has been little cross-disciplinary synthesis, and fundamental theoretical issues re- main unresolved. The major points that emerge from the literature in several disciplines include the following: (1) There is still no generally accepted model of consumer behavior. (2) The database for cross-cultural comparison of consumption is poor in quality. 

EXAMINING THE DRIVING FORCES 113 (3) No single academic discipline has adequate tools or data for study- ing cross-cultural consumer behavior. (4) The development of consumer culture in developing countries is following a different trajectory from the historical path of the West. (5) There is still every reason to think that consumption will increase as incomes rise, but we cannot yet predict how that increase will be ap- portioned to various goods or sectors. (6) Simple emulation remains an empirically weak model for predic- tion. REFERENCES Andrae, G., and B. Beckman 1985 The Wheat Trap: Bread and Underdevelopment in Nigeria. London: Zed Books. Appadurai, A., ed. 1986 The Social Life of Things. Cambridge, England: Cambridge University Press. 1990 Disjuncture and difference in the global cultural economy. Theory, Culture, and Society 7:295-310. Belk, R. 1988 Third world consumer culture. Research in Marketing. Supplement 4.:103-127. Belk, R., and N. Dholakia 1995 Consumption and Marketing: Macro Dimensions. Belmont, Calif.: Southwestern. Benson, J. 1994 The Rise of Consumer Society in Britain, 1880-1980. London: Longman. Brewer, J., and R. Porter, eds. 1993 Consumption and the World of Goods. New York: Routledge. Campbell, C. 1987 The Romantic Ethic and the Spirit of Modern Consumerism. Oxford, England: Basil Blackwell. Craik, J. 1994 The Face of Fashion. London: Routledge. Cross, G. 1993 Time and Money: The Making of Consumer Culture. London: Routledge. Csikszentimihalyi, M., and E. Rochberg-Halton 1981 The Meaning of Things: Domestic Symbols and the Self. Chicago: University of Chicago Press. Drummond, P., and R. Patterson, eds. 1988 Television and its Audience: International Research Perspectives. London: British Film Institute Publishing. Ewen, S. 1988 All Consuming Images. New York: Basic Books. Featherstone, M., ed. 1990 Global Culture: Nationalism, Globalization and Modernity. London: Sage Publica- tions. Folbre, N. 1994 Who Pays for the Kids? Gender and the Structure of Constraint. London: Routledge. Foster, R. 1991 Making national cultures in the global ecumene. Annual Review of Anthropology 20:235-260. 

114 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION Friedman, J., ed. 1994 Consumption and Identity. Chur, Switzerland: Harwood. 1990 Being in the world: Globalization and localization. Theory, Culture and Society 7:311-328. Gereffi, G., and M. Korzeniewicz, eds. 1994 Commodity Chains and Global Capitalism. Westport, Conn.: Praeger. Hannerz, U. 1987 The world in Creolization. Africa 57(4):546-559. 1990 Cosmopolitans and locals in world culture. Theory, Culture, and Society 7:237-251. Holt, D. 1994 Consumers’ cultural differences as local systems of tastes: A critique of the per- sonality/values approach and an alternative framework. Asia Pacific Advances in Consumer Research 1:178-184. Horowitz, D. 1988 The Morality of Spending. Baltimore, Md.: Johns Hopkins University Press. James, J. 1993 Consumption and Development. New York: St. Martin’s Press. Kuisel, R. 1993 Seducing the French: The Dilemma of Americanization. Berkeley: University of California Press. Lears, T.J.J. 1989 Beyond Veblen: Rethinking consumer culture in America. Pp. 73-97 in Simon Bronner, ed., Consuming Visions: Accumulation and Display of Goods in America, 1880-1920. New York: Norton. Liebes, T., and E. Katz 1990 The Export of Meaning: Cross-Cultural Readings of Dallas. New York: Oxford University Press. Lluch, C., A. Powell, and R. Williams 1977 Patterns in Household Demand and Saving. New York: Oxford University Press and the World Bank. McCracken, G. 1988 Culture and Consumption. Bloomington: Indiana University Press. Miller, D. 1995 Acknowledging Consumption. London: Routledge. Mintz, S. 1985 Sweetness and Power: The Place of Sugar in Modern History. New York: Penguin Books. Moore, S. 1993 Interpreting Audiences: The Ethnography of Media Consumption. London: Sage Publications. Phipps, S., and P. Burton 1995 Social/institutional variables and behavior within households: An empirical test using the Luxembourg income study. Feminist Economics 1(1):151-174. Rassuli, K. and S. Hollander 1986 Desire—induced, innate, insatiable? Journal of Macromarketing 6(2):4-24. Richards, T. 1991 The Commodity Culture of Victorian Britain: Advertising and Spectacle, 1851-1914. Stanford, Calif.: Stanford University Press. Sherry, J., ed. 1995 Contemporary Marketing and Consumer Behavior: An Anthropological Sourcebook. Thousand Oaks, Calif.: Sage Publications. 

EXAMINING THE DRIVING FORCES 115 Sklair, L. 1994 The culture-ideology of consumerism in urban China. Research in Consumer Be- havior, Vol. 7. Greenwich, Conn.: JAI Press. Tiersten, L. 1993 Redefining consumer culture: Recent literature on consumption and the bour- geoisie in Western Europe. Radical History Review 57:116-159. Tobin, J., ed. 1992 Re-Made in Japan: Everyday Life and Consumer Taste in a Changing Society. New Haven, Conn.: Yale University Press. Tomlinson, J. 1991 Cultural Imperialism: A Critical Introduction. Baltimore, Md.: Johns Hopkins Uni- versity Press. Wagnleitner, R. 1994 Coca-Colonization and the Cold War. Chapel Hill: University of North Carolina Press. Ware, W., and M. Dupagne 1994 Effects of U.S. television programs on foreign audiences: A meta-analysis. Jour- nalism Quarterly 71(4):947-959. Wilk, R. 1995 The local and the global in the political economy of beauty: From Miss Belize to Miss World. Review of International Political Economy 2(1):117-134. 

116 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION CULTURAL AND SOCIAL EVOLUTIONARY DETERMINANTS OF CONSUMPTION Willett Kempton and Christopher Payne In the workshop paper summarized here (see Kempton and Payne, forthcoming, for the complete version), we draw data from a wide range of human societies to ask: What can cross-cultural comparisons teach us about the relationship between consumption and quality of life? We argue that the dependence of quality of life on consumption is not mono- tonic and is both weaker and more complex than is often assumed. We begin by addressing two myths that underlie much thinking about consumption. The first myth is that quality of life generally increases with higher consumption levels—that is, more consumption of goods and services increases the quality of life. This relationship is believed to hold across societies and across social strata within any society. Parts of it are parodied in the tee-shirt slogan “He who dies with the most toys wins.” The second myth is that society evolves and changes to improve the lot of individuals. If our society previously had one form of government, kin- ship system, economy, or whatever and another form replaces it, the societal change improves the quality of life of members of the changed society. We call these the “most toys” myth and the “social evolution for individual benefit” myth. They are addressed at several points here. We begin by considering the types of societies within which biologi- cally modern humans evolved. These societies are small, organized around family relationships, and subsist by hunting and gathering. Both mobility and their social organization limit consumption. Mobile societ- ies shift residences, whether on a predictable yearly cycle based upon seasonal cycles of wild crops and game or moving more opportunistically to follow herds, water, or areas not yet exhausted of plant resources. Individuals in these societies limit consumption simply via the limit on inventories—you can’t possess more than you can carry. Socially, hunting and gathering societies are organized around family relationships and are egalitarian. We also briefly examine a subsequent form of subsistence, swidden agriculture. This pattern relies on cutting and burning forest, farming the plot for one or a few years, and abandoning it for decades to lie fallow and regrow. As societies moved from hunting and gathering to swidden agriculture, and then to fixed agriculture, changes in social and political organization accompanied these production and settlement changes. Among other things, these changes increase status differentiation. With larger populations in settlements and social differentiation comes the need 

EXAMINING THE DRIVING FORCES 117 for display of status by means of prestige goods. Subsequently in social evolution, material consumption is driven partially by status competi- tion. Remarkably, the consumption literature rarely distinguishes con- sumption for social-status display from sustenance, enjoyment, or other (sometimes overlapping) motivations for consumption. Social status con- sumption is a zero-sum game, which drives competing individuals or groups toward higher consumption—ending not with “need satisfaction” but only with exhaustion of an individual’s resources. Acting alone, each individual competing for status seeks to make the best of his or her position. But satisfaction of these individual preferences itself alters the situation that faces others seeking to satisfy similar wants. A round of transactions to act out personal wants of this kind therefore leaves each individual with a worse bargain than was reckoned on when the transaction was undertaken, because the sum of such acts does not correspondingly improve the position of all individuals taken together. There is an “adding-up” problem (Hirsch, 1976). LEVELS OF CONSUMPTION Next we address relative levels of consumption across societies. We compare consumption of the main two throughputs of environmental interest, mass and energy, and further divide mass throughputs into re- cycled and nonrecycled categories. Energy use has been thoroughly stud- ied in a number of indigenous societies. Total mass throughputs of indig- enous societies have not been studied explicitly, but we can make estimates from existing ethnographic data. The bulk of the mass used by indigenous peoples is biodegradable and recycled by biological processes. Wood, hide, reed or bamboo, foodstuffs, and such, when discarded, de- grade and feed biological cycles. These societies also create a one-way (nonrecycled) flow of materials for stone tools, such as chert, flint, and obsidian. Ceramic vessels can survive 10,000 years before reintegration into the soil, so we consider them nonrecycled as well. Table 4-6 shows four types of societies, with estimates for energy, nonrecycled material, and recycled materials. Note that hunter–gatherers function at two orders of magnitude less energy and one order of magni- tude less materials than the United States. Swiddeners (based on Beckerman, 1976) use about the same level of materials as does the United States but differ from the United States in that over 99 percent of the materials are recycled. The supposedly modern concept of “sus- tainability” has been achieved in most hunting–gathering and swidden agricultural societies, as evidenced by the fact that many of these societies can be shown to have run their systems of materials and energy through- put in the same locations for millennia. These societies modify their 

118 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION TABLE 4-6 A Rough Quantitative Comparison of Energy and Materials Use Across Diverse Types of Societies Nonrecycled Energy Materials Recycled Materials (kW/capita) (kg/capita/day) (kg/capita/day) Hunting and gathering 0.11 0.035 3.6 Swidden horticulture 0.25 0.15 50–100 Agriculture in a 1-3 0.5 4–50 developing country U.S.A. 11 56 2.7 NOTE: The full paper explains how the quantities were calculated or estimated. environments initially—especially swiddeners—but then continue in the same location for very long periods without continuing environmental degradation. Their long-term durability within the ecosystem is not matched by durability in contact with the global political economic sys- tem—upon this contact they are quickly absorbed into the nonsustainable world economy. HOW CAN ONE COMPARE QUALITY OF LIFE? Low-consumption societies are not very relevant if the life they live is “nasty, brutish and short” [Hobbes, 1968(1651):186]. Comparing quality of life across societies is fraught with problems, but anthropologists have developed some measures. To summarize briefly, our paper suggests potential candidates such as nutrition, health, life-span, work time vs. leisure time, in- vs. out-migration (“revealed preference” for a given soci- ety), and relative perceived quality of life by ethnographers. Of course, this is not a complete list of all the measures we would like to have. What the above measures have to recommend them is that they are in available data—ethnographic, archaeological, human biological, or paleoarchaeological records. When objective measures are applied to compare the quality of life across widely divergent societies, the results are surprising. We concentrate here on work time and health; other mea- sures are covered in Kempton and Payne (forthcoming). RESULTS OF COMPARISON Regarding health measures, studies of skeletal remains show that health declined—not improved as might be expected—after transitions from hunting-gathering to early agriculture, then from early agriculture 

EXAMINING THE DRIVING FORCES 119 to archaic states. Decreases in health occurred due to the greatly reduced range of plant species eaten, social stratification resulting in separation of decision makers from the bulk of the population, and high population densities leading to infectious disease (Diamond, 1987). Health did not improve markedly until the nineteenth and twentieth centuries, as a re- sult of public health measures in the cities and the advent of modern antibiotics. Life-span is longer in industrialized societies than in any of the indigenous societies we discuss. One component of the “most toys” myth is that the devices we con- sume reduce our work time—i.e., that life is easier today than in earlier historical periods or in technologically primitive societies. Regarding earlier historical periods, Juliet Schor has challenged the myth of less work in the modern era. She takes the comparison back to medieval time (Schor, 1991), finding a large increase in work time during the industrial revolution and a decline back to medieval levels during the twentieth century. We feel that the work-time comparison gets more interesting when extended to indigenous peoples. Several sources demonstrate that sustenance requires less work in primitive societies than in our own. In one study of indigenous people, Johnson (1978), compares middle- class France with the Machiguenga, a horticultural group in the Peruvian rain forest, another society autonomous from the global economy. The middle-class French had 10 hours of free time per day, and the Amazo- nian people had 14 hours of free time per day; free time in each case included about 8 hours of sleep. Qualitatively, Johnson made parallel observations about his own time sense while living there: “[In] their communities . . . I sense a definite decrease in time pressure . . . when I return home [to the United States] I am conscious of the pressure and sense of hurry building up to its former level” (1978:53). Other studies of time required for subsistence are reported in Sahlins (1972). Hunting– gathering societies often require only 3-4 hours of work to provide an ample and varied diet. In sum, hunter-gatherers and swidden agricultur- alists work less and have more leisure than citizens of industrialized soci- eties. Other studies of contemporary indigenous peoples being drawn into the market economy similarly demonstrate the forces that lead to incorporation into market-based, higher-consumption lifestyles. The at- traction at time of entry occurs despite eventual degradation in the incor- porated peoples’ quality of life (e.g., Barlett and Brown 1985; Bodley, 1990). To briefly consider urban-industrial society, many authors argue that many of the historical developments of the past century have been incon- sistent with a higher quality of life for individuals. Historians like Wiebe (1967), Hughes (1989), and Marcus and Segal (1989) have traced the rise of the technological society during the twentieth century and identified its 

120 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION defining focus to be the growth of large-scale organizational systems. Organizational systems can be defined as operational structures that pro- vide their members with efficient means of achieving given ends [com- pare the administrative theories of Simon (1957) and the economic theo- ries of Galbraith (1967)]. Social critics such as Mumford (1934, 1967, 1970) and Ellul (1964) have argued that these organizational goals result in isolated, dehuman- ized individuals, while benefiting the organization itself. Furthermore, organizational theorists have argued that large-scale organizations pre- vent individuals from developing fully on a psychological level. Draw- ing heavily on the work of psychologists such as Jung and Marcuse, Denhardt (1981) has argued that there is a fundamental tension between the individual and collective psyches. This presents a problem of inte- grating the individual and collective psyches into a self-actualized whole. In this view, the development of organizational systems has led to a col- lective psyche that values certain aspects of the human psyche (rational- ity, instrumentality, etc.) at the expense of others (emotion, expression, etc). The repression of these emotive values hampers individual develop- ment. Because of the structure of organizations, therefore, social and individual development is inhibited. The perspectives of these authors suggest that organizations have become autonomous actors in our society, furthering their own aims rather than human welfare. When people believe that the efficient pro- duction of goods is the means for improving their quality of life, this pursuit makes sense. As people come to recognize the destructive charac- teristics of the material lifestyle and of the systems that support it, they see organizations as, in many respects, fulfilling organizational aims to the detriment of human individuals. The aspects of life that define us as human—expressive, creative, unique—are those aspects that are in con- flict with the needs of organizational structures for efficient operation. It is, finally, organizational operations that are supported by the myth of consumption. CONCLUSION Obviously, we do not advocate abandoning fixed communities, agri- culture, and modern technology, a change impossible at current world population levels. Rather, we wish to list the following observations from the data outlined here and explored more in the full paper: (1) Most early societies had consumption levels several orders of magnitude smaller than industrial societies today. However, some indig- enous societies had very high levels of per capita materials consumption, 

EXAMINING THE DRIVING FORCES 121 similar in magnitude to the United States today but with virtually all in materials promptly recycled by the biosphere. (2) By objective indicators other than life-span, the quality of life in some ultra-low-consumption societies seems rather high—higher than the societies they next evolved into, and by many indicators higher than ours today. (3) Major social transitions can occur if they provide benefits to deci- sion-making elites and greater “fitness” at the societal level (e.g., military advantage or rapid growth and spread of the sociopolitical system). (4) Increasing the quality of life of the broad masses of individuals is not a criterion by which organizations survive, nor has it been a force determining the direction of social evolution. ACKNOWLEDGMENTS We are grateful to Jill Neitzel for major conceptual, literature, and reference suggestions. Steven Beckerman and Thomas Rocek provided important data. Abigail Jahiel, Faith Mitchell, and various participants at the National Research Council workshop provided helpful comments on the argument and logic of drafts of this paper. None of these commenta- tors and contributors are responsible for its content. REFERENCES Barlett, P.F., and P.J. Brown 1985/ Agricultural development and the quality of life. Agriculture and Human Values. 1994 Pp 175-182 reprinted in A. Podolefsky and P. J. Brown, Applying Anthropology: An Introductory Reader. Mountain View, Calif.: Mayfield Publishing. Beckerman, S.J. 1976 The Cultural Energetics of the Barí of Northern Columbia. Unpublished Ph.D. dissertation, University of New Mexico, Department of Cultural Anthropology. Bodley, J.H. 1990 Victims of Progress. Mountain View, Calif.: Mayfield Publishing. Denhardt, R.B. 1981 In the Shadow of Organization. Lawrence. Kans.: The Regents Press of Kansas. Diamond, J. 1987 The worst mistake in the history of the human race. Discover (May) 64-66. Ellul, J. 1964/ The Technological Society. Translated by J. Wilkinson. New York: Alfred 1954 A. Knopf and Random House. Galbraith, J.K. 1967 The New Industrial State. New York: Signet. Hirsch, F. 1976 Social Limits to Growth. Cambridge, Mass.: Harvard University Press. 

122 ENVIRONMENTALLY SIGNIFICANT CONSUMPTION Hobbes, T. 1968/ Leviathan. Penguin Classics, 1985 ed. New York: Penguin Books. 1651 Hughes, T.P. 1989 American Genesis: A Century of Invention and Technological Enthusiasm 1870-1970. New York: Penguin Books. Johnson, A. 1978 In search of the affluent society. Human Nature (Sept):50-59. Kempton, W., and C. Payne in Cultural and social evolutionary determinants of consumption. In T. Dietz, ed., press Environmental Impacts of Consumption. Marcus, A.I., and H.P. Segal 1989 Technology in America: A Brief History. San Diego, Calif.: Harcourt Brace Jovanovich. Mumford, L. 1934 Technics and Civilization. San Diego, Calif.: Harvest/Harcourt Brace and Company. 1967 Technics and Human Development. The Myth of the Machine, Vol. 1, lst ed. New York: Harcourt Brace Jovanovich. 1970 The Pentagon of Power. The Myth of the Machine, Vol. 2, lst ed. New York: Harcourt Brace Jovanovich. Sahlins, M. 1972 Stone Age Economics. Chicago: Aldine-Atherton. Schor, J.B. 1991 The Overworked American: The Unexpected Decline of Leisure. New York: Basic Books. Simon, H.A. 1957 Administrative Behavior: A Study of Decision-Making Process in Administrative Orga- nization, 2nd ed. New York: Free Press. Wiebe, R.H. 1967 The Search for Order: 1877-1920. New York: Hill and Wang. BIBLIOGRAPHY British Petroleum Company 1994 Statistical Review of World Energy. London: British Petroleum Company. Cohen, M.N., and G.J. Armelagos 1984 Paleopathology and the Origins of Agriculture. New York: Academic Press. Crèvecoeur, J.H. St. John 1782/ Letter from an American farmer. In A. Kolodny, Among the Indians: The uses 1993 of captivity. New York Times Book Review. (Jan): 1. The Harwood Group 1995 Yearning for Balance: Views of Americans on Consumption, Materialism, and the Envi- ronment. Takoma Park, Md.: Merck Family Fund. Kempton, W., J.S. Boster, and J.A. Hartley 1995 Environmental Values in American Culture. Cambridge, Mass.: M.I.T Press. Kolodny, A. 1993 Among the Indians: The uses of captivity. New York Times Book Review (Jan): 1. Lee, R.B. 1969 !Kung bushman subsistence: An input-output analysis. In A.P. Vayda, ed., Envi- ronment & Cultural Behavior. Austin, Tex.: University of Texas Press. 

EXAMINING THE DRIVING FORCES 123 1979 The !Kung San. Cambridge, England: Cambridge University Press. Lightfoot, R.R. 1994 The Duckfoot site. Archaeology of the House and Household, Vol. 2; Cortez, Colo.: Crow Canyon Archaeological Center. Rappaport, R.A. 1971 The flow of energy in an agricultural society. Scientific American 225(3):117-122. U.S. Bureau of the Census 1992 Statistical Abstract of the United States: 1992, 112th ed. Washington, D.C.: U.S. Department of Commerce. Wernick, I.K., and J.H. Ausubel 1995 National materials flow and the environment. Annual Review of Energy and Envi- ronment 20:463-492.