Social Science Research K.W. Knight, E.A. Rosa/Social Science Research 40 (2011) 931–949 the well-being of the world’s poor.’’ Their analyses indicate that for various measures

Social Science Research 40 (2011) 931–949

Contents lists available at ScienceDirect

Social Science Research

journal homepage: www.elsevier .com/locate /ssresearch

The environmental efficiency of well-being: A cross-national analysis

Kyle W. Knight ⇑, Eugene A. RosaDepartment of Sociology, PO Box 644020, Washington State University, Pullman, WA 99164-4020, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 8 May 2010Available online 21 November 2010

Keywords:SustainabilitySubjective well-beingConsumptionEnvironmental sociologyEcological footprintLife satisfaction

0049-089X/$ - see front matter � 2010 Elsevier Incdoi:10.1016/j.ssresearch.2010.11.002

⇑ Corresponding author. Fax: +1 509 335 6419.E-mail addresses: [email protected] (K.W.

Recent research has conceptualized sustainability as the environmental efficiency of well-being (EWEB). This concept takes into account the benefits societies are able to producefrom their demands on the environment. Research along these lines indicates that coun-tries vary widely in the efficiency with which they transform the Earth’s resources intowell-being. Here, we take up this finding as a puzzle to be explained. We construct anew measure of EWEB using the ecological footprint per capita (a measure of environmen-tal consumption) and average life satisfaction (a measure of subjective well-being). Wedraw hypotheses from political economy, modernization, and sustainable consumptiontheories in the environmental social sciences. Using full information maximum likelihoodestimation, we test the effects of climate, political, economic, and social factors on EWEBwith a sample of 105 countries. Key findings include a negative quadratic effect of eco-nomic development on EWEB, a negative effect of income inequality, and a positive effectof social capital.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

Definitions of sustainability typically refer, implicitly or explicitly, to striking a balance between societal demands on theenvironment and social well-being (of current and future generations). For instance, the influential 2005 Millennium Ecosys-tem Assessment was titled Ecosystems and Human Well-Being: Synthesis (Reid et al., 2005). The goal of sustainability is tominimize environmental impacts and maximize human well-being (Dietz et al., 2009; Prescott-Allen, 2001). Chamberset al. (2000, p. 3), argue that, ‘‘To make sustainability happen, we need to balance the basic conflict between the two com-peting goals of ensuring a quality of life and living within the limits of nature.’’ Thus, sustainability is the sustaining of hu-man well-being into the indefinite future by minimizing humanity’s impact on the natural environment upon which thatwell-being ultimately depends. The goal, as stated in the Brundtland Report, is to ‘‘[meet] the needs of the present withoutcompromising the ability of future generations to meet their own needs’’ (WCED, 1987, p. 8). According to that influentialreport this is to be achieved by focusing on economic growth, social equity, and environmental protection (WCED, 1987).Much of the debate on sustainability and the means to attain it (i.e., sustainable development) centers on the primacy,and desirability, of each of these three dimensions of sustainability. In particular, a focused debate has ensued over whethereconomic development or well-being should be the goal. We address this issue in detail below.

Since the landmark Rio Earth Summit and publication of Agenda 21 in 1992, sustainable consumption—how to define it,measure it, and attain it—has become a major topic and policy issue. Here we focus on consumption as the key aspect ofsustainability because economics is dominated by the assumption of a strong link between consumption and utility, orwell-being, and it is where the issue of well-being has been most researched (see Jackson, 2005; Seyfang, 2009).

. All rights reserved.

Knight), [email protected] (E.A. Rosa).

http://dx.doi.org/10.1016/j.ssresearch.2010.11.002

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.ssresearch.2010.11.002

http://www.sciencedirect.com/science/journal/0049089X

http://www.elsevier.com/locate/ssresearch

932 K.W. Knight, E.A. Rosa / Social Science Research 40 (2011) 931–949

There are two main competing views on what sustainable consumption means. The first sees it, not as a reduction in con-sumption, but as ‘consuming differently.’ The focus is on ‘green consumerism,’ market-mechanisms, and production efficien-cies (Mol and Spaargaren, 2004; Vincent and Panayotou, 1997). This understanding of sustainable consumption relies on theassumptions of neoclassical economics that equates consumption with utility or well-being1 and thus does not advocate forlower levels of consumption. According to this view, if sustainability requires reducing consumption levels, then we face a bleakchoice between maintaining our own well-being at the expense of future generations or reducing our present well-being (Markset al., 2007). The route out of this dilemma is to make our consumption more benign or more efficient.

The second approach is critical of the link between consumption and well-being and defines sustainable consumption as‘consuming less’ in absolute terms. Hence, it is critical of both the environmental costs of high levels of consumption as well asthe purported well-being benefits of consumption. This perspective has been marginalized relative to the former in main-stream policy discussions (Seyfang, 2009; Jackson and Michaelis, 2003). In contrast to the continue to consume, but more effi-ciently, ‘‘greening,’’ view of sustainable consumption, this second view suggests that, because consumption levels and well-being are not tightly coupled, reaching sustainability by reducing consumption will not require reductions in well-being.

A significant body of research has demonstrated that consumption is a major cause of global environmental degradation(Redclift, 1996; Rothman, 1998; Worldwatch Institute, 2004; York et al., 2003). Consumption, therefore, has been attractingincreased recognition as an environmental problem in both academic and policy circles (Cohen, 2001; Fuchs and Lorek,2005; Seyfang, 2009). However, most of the empirical research has focused on consumption as a driver of environmentalimpacts without regard to the social well-being that consumption may generate. There is a virtual absence of social scienceresearch on the social benefits of environmental consumption (for an exception, see Dietz et al., 2009). A number of studieshave explored the relationship between natural capital or environmental quality and human well-being (Engelbrecht, 2009;Vemuri and Costanza, 2006); however, these studies have not focused on the role of environmental consumption per se.Some researchers from the world-systems perspective have recently analyzed the direct and indirect influence of environ-mental consumption on various measures of objective well-being. Specifically, Jorgenson and Burns (2004) have demon-strated a negative indirect effect of environmental consumption on infant mortality while Rice (2008) has found anegative relationship between environmental consumption and maternal mortality rates in less-developed countries. Objec-tive quality of life indicators, such as these, provide the parameters or inputs to well-being and, are, therefore, indirectestimates of well-being (Veenhoven, 1995). Examining the subjective dimension of well-being is a more direct measureof well-being and the natural complement to its objective inputs.2 Given that conceptualizations of sustainability oftenencompass both environmental impacts and human well-being, empirical research in this area is essential for a more completeunderstanding of the dimensions of sustainability.

2. Previous research on consumption and well-being

As noted above, while most research on the relationship between consumption and the environment has neglected theconsumption and well-being nexus, several studies have begun to explore the topic. The question posed by these researchersis straightforward: what is the nature of the relationship (if any) between levels of consumption and social well-being?

Conventionally, especially in economics and policy, measures of affluence (i.e., GDP per capita) have typically been usedas valid measures of well-being. Based on the axioms of revealed preferences (Samuelson, 1938) and utility maximization,economics equates higher levels of GDP with higher levels of well-being (Royo, 2007). This assumes rather than investigatesthe equivalence of consumption with well-being. There is good reason to question the use of GDP as a complete proxy forwelfare or well-being.3

First, it leaves out natural capital which sustainable well-being depends upon but includes expenditures on positional(i.e., goods that provide satisfaction as a function of comparison) and defensive goods (i.e., goods that protect rather thanincrease satisfaction) as well as negative externalities none of which contributes to well-being (e.g., the costs of cleaningup pollution add to GDP rather than detract) (Daly and Cobb, 1994; Hirsch, 1977; Jackson, 2009). As Ayres and Martinas(2005, p. 128) note, GDP ‘‘is not really a measure of welfare. . . [it] is a measure of economic activity, nothing more.’’

Second, the relationship between affluence and both objective and subjective well-being is curvilinear, with higher levelscontributing less and less to well-being (i.e., diminishing marginal returns) (Dietz et al., 2001; Goldstein, 1985; Inglehart andKlingemann, 2000; Layard, 2005; Preston, 1975). Fig. 1 illustrates the curvilinear relationship between GDP per capita andsubjective well-being. Furthermore, recent cross-national research has questioned the equating of economic growth withwell-being in less-developed countries. Brady et al. (2007, p. 2) criticize the ‘‘growth consensus: the conventional wisdomthat economic growth should be the primary priority for less-developed countries because it most effectively improves

1 Not all economists are in agreement that utility and well-being are the same thing (see, for example, Becker and Rayo, 2008).2 Objective measures of well-being are social indicators that reflect the extent to which human needs are met (e.g., life expectancy, infant mortality, etc.).

These measures do not require input from the people they describe. On the other hand, subjective measures of well-being are self-reported levels of well-being(e.g., happiness, life satisfaction, etc.) as perceived by the individual his or herself.

3 This is a transformational point of growing recognition, even among economists. Earlier recognition can be found in Nussbaum and Sen (1993) and veryrecent refinements are in the influential report commissioned by French President Nicolas Sarkozy in 2008, directed by Nobel Laureates Joseph Stiglitz andAmartya Sen and distinguished French economist Jean-Paul Fitoussi, Report by the Commission on the Measurement of Economic Performance and Social Progress(Stiglitz et al., 2010).

Fig. 1. Scatterplot of average life satisfaction, 2006–2009, and gross domestic product (GDP) per capita, 2005 (N = 105).

K.W. Knight, E.A. Rosa / Social Science Research 40 (2011) 931–949 933

the well-being of the world’s poor.’’ Their analyses indicate that for various measures of objective well-being, economicgrowth has no robust effects, has become less effective over time, and that factors other than growth, such as fertility rates,urbanization, and education, are more influential in predicting well-being in less-developed countries.4

2.1. The economics of subjective well-being

Economist Richard Easterlin (1974, 1995) pioneered5 the field of the economics of happiness or subjective well-being andarrived at these paradoxical findings: (1) within a society, rich people tend to be much happier than poor people; (2) rich soci-eties tend to be no happier (or not by much) than poor societies, and (3) as countries get richer they do not get happier. Fur-thermore, some economists, building on the Easterlin findings, have argued for a satiation point—a point where increases inincome increase well-being no further (Frey and Stutzer, 2002; Layard, 2005). Easterlin explains his findings, referred to eversince as ‘‘the Easterlin paradox,’’ with two social factors: social comparison and hedonic adaptation. He argued that once basicneeds are met, relative income is more important than absolute income in determining subjective well-being. Social comparisonrefers to the idea that within countries higher income is associated with higher subjective well-being because individuals com-pare their own consumption to the prevailing norm; those consuming at higher levels feel better off than those consuming atlower levels. But, as an already wealthy nation becomes richer over time, consumption norms undergo an upward shift so thatwhile consumption may be increasing, national average well-being does not change. Easterlin refers to this phenomenon as he-donic adaptation. Recently Easterlin and colleagues (Easterlin and Sawangfa, 2009; Easterlin and Angelescu, 2009) found thateconomic growth over time does not lead to higher subjective well-being in either developed or developing countries.

Further investigation of the ‘‘Easterlin paradox’’ comes from research demonstrating that economic growth in industrial-ized countries since WWII has not resulted in substantial increases in subjective well-being (Diener and Oishi, 2000). Recentfindings by Diener et al. (2010, p. 61) show ‘‘that more than money is needed for quality of life, and the social psychologicalforms of prosperity correlate only moderately with economic development.’’ Furthermore, Helliwell (2003) finds that socialfactors other than affluence are more important in predicting national-level life satisfaction such as low corruption, highlevels of mutual trust, and effective social and political institutions. In addition, Inglehart (2009, p. 351) finds that nationalincome per capita ‘‘follows a curve of diminishing returns’’ in relation to subjective well-being.

The Easterlin paradox has attracted its share of critics. The literature on this topic is too vast and growing too rapidly toadequately review here. Nevertheless, it is important to note that several recent articles that have called into question theempirical validity of the Easterlin paradox (e.g., Deaton, 2008; Hagerty and Veenhoven, 2003; Stevenson and Wolfers, 2008).Graham et al. (2009) argue that the national income-subjective well-being relationship is highly complex and that the effects

4 Thicker description is provided for such cross-national findings by detailed case studies showing that considerably high quality of life can be achieved inpoor countries or even regions of countries. The point has been amply demonstrated in the Indian state of Kerala (population 32 million) which, althoughhaving one of the lowest GDPs in India, has high literacy rates and life expectancies for both sexes (even higher than African–Americans) and effective healthcare. These high objective well-being measures can be traced to policies and programs that promote literacy, that expand health care, and are attentive toinequalities (for a descriptive summary see Sen (1993)).

5 Another pioneering economist on this topic was Scitovsky (1976).


of income are sensitive to a number of factors including survey question-framing, the type of well-being analyzed, samplecomposition, selected time frame, previous levels of economic development, and rates of economic growth.

Despite the limitations in the data, Stevenson and Wolfers (2008), by examining many datasets, provide the most thor-ough test of the Easterlin hypothesis to date. Their principal test is of the hypotheses that poorer nations are just as happy asrich ones and that there is an income upper bound or satiation point. They find support for neither. Instead they establish apositive link between the log of GDP per capita and subjective well-being and no evidence of a satiation point where wealth-ier countries have no further increases in well-being.

In sum, the Easterlin paradox is neither definitively confirmed nor disconfirmed by this collection of recent studies.According to Diener et al. (2009b, p. xvi), the debate regarding income and subjective well-being is ‘‘best regarded as unset-tled.’’ Resolution of the paradox will require the collection of consistent panel data for many countries, spanning many years.Such data is likely to remain unavailable for the foreseeable future. In the meantime, it appears that the link between incomeand well-being is clearly a concave, monotonic function, as illustrated in Fig. 1. It apparently takes large increases in incomeat upper levels to produce modest increases in well-being. The question the relationship poses in the environmental contextwe are investigating here is about the link between consequential environmental threats and well-being.

2.2. Environmental consumption and well-being

A second body of research with sociological roots avoids the conventional economic assumption of equivalence by inves-tigating the effects of environmental consumption directly on objective and subjective measures of well-being. Early re-search by Mazur and Rosa (1974) documented the decoupling in the US of energy consumption (a critical naturalresource) and various objective measures of quality of life. Rosa (1997) later found that carbon emissions and objectivewell-being had become decoupled between 1970 and 1985 in a sample of advanced industrial countries. Recent researchhas found that environmental consumption, measured as the ecological footprint (EF)6, has no direct effect, net of income,on objective well-being measured as life expectancy (Dietz et al., 2009). Furthermore, Knight and Rosa (2009) assessed the di-rect effect of environmental consumption, measured as EF on subjective well-being measured as average life satisfaction. Thefindings show that: (1) increased consumption does not lead to higher levels of, and may actually decrease, well-being in high-consumption countries, (2) high consumption levels are not a necessary condition for high levels of well-being, and (3) verylarge increases in consumption since 1961 have not led to substantial improvements in life satisfaction in the US.

These findings suggest that improvements in well-being may be attainable without escalating the consumption of envi-ronmental resources. In short, environmental consumption and well-being are apparently not linearly related. Furthermore,countries with high levels of well-being vary widely in their levels of consumption. For instance, Costa Rica has achieved alevel of subjective well-being rivaling that of the US with an EF per capita four times smaller (Abdallah et al., 2009; Knightand Rosa, 2009). Further evidence comes from Marks et al.’s (2006) computation and analysis of the Happy Planet Index,which measured the ratio of well-being to environmental consumption and found wide variation among countries.

Research on the relationship between environmental consumption and well-being led to a key refinement. It refined theidea of sustainability as the maximization of human well-being and minimization of environmental impact by conceptual-izing sustainability as environmental efficiency of well-being (EWEB) (Dietz et al., 2009). The EWEB concept takes into ac-count the benefits (alternatively measured as subjective or objective well-being) societies are able to produce from theirdemands on the environment. EWEB is an important concept and indicator of sustainability because measures of environ-mental sustainability, like the EF do not measure sustainability per se, but only the ecological and distributional aspects of it(Chambers et al., 2000). By taking into account both well-being and level of environmental consumption, EWEB is more fullyconsonant with the prevailing definitions of sustainability. It indicates the degree to which a society is achieving a desirablelevel of well-being without compromising the well-being of future generations by depleting natural capital. As noted above,research along these lines indicates that countries vary widely in the efficiency with which they transform the Earth’s re-sources into well-being (Dietz et al., 2009; Knight and Rosa, 2009; Marks et al., 2006). This is an especially significant finding,but it begs the question: What social factors influence a country’s environmental efficiency of well-being?

We address this question with a quantitative cross-national analysis in several steps. First, we graphically depict the rela-tionship between environmental consumption and a measure of subjective well-being. Second, we devise and compute anew indicator of the Environmental Efficiency of Well-being (EWEB) using a measure of subjective well-being and a measureof consumption-based environmental pressure, the EF. Third, we examine several factors in the light of prior theorizationand findings that may influence EWEB. Fourth, we estimate the effects of social, political, economic, demographic, and cli-mate variables on this new measure of EWEB using Arbuckle’s (1997) full information maximum likelihood estimation (MLE)on a large sample of countries.

3. Measuring EWEB

There have been several attempts to operationalize EWEB in the last decade. First, Dietz et al. (2001) calculated a ratioincluding infant mortality (transformed into infant survival) as the numerator and a combined index including EF per capita

6 See Section 3.2 below for a definition and description of the ecological footprint.


(measuring consumption-based environmental impact) and the Environmental Systems Indicator (a subset of the Environ-mental Sustainability Index measuring environmental quality) as the denominator (Samuel-Johnson, 2001). Second, a similarmeasure was developed by Prescott-Allen (2001), the Wellbeing/Stress Index, which includes a composite index of objectivewell-being measures in the numerator and a composite index of ecosystem health (environmental quality) in the denomi-nator. Third, the Happy Planet Index (HPI) (Marks et al., 2006; Abdallah et al., 2009) is the ratio of average life satisfactionmultiplied by life expectancy over the EF per capita.7 A related EWEB measure is Ng’s (2008) Environmentally ResponsibleHappy Nation Index, which is an attempt to revise conceptually and mathematically the HPI. Fourth, Dietz et al. (2009) esti-mated the efficiency of well-being using a stochastic frontier production model with ecological footprint, affluence, and educa-tion as predictors of life expectancy.

These approaches have key limitations for several reasons. First, as recognized by Marks et al. (2006) and Abdallah et al.(2009) in calculating the HPI, ratios are often dominated by either the numerator or denominator depending on which hasgreater dispersion (in this case the denominator). This means that EWEB indicators measured as ratios are dominated by EFscores and thus can lead to misleading results. The HPI corrects for this by adding a constant to the EF per capita in order tomatch its coefficient of variation with the numerator; however, this artificial deflation of dispersion in the EF may result inless straightforward interpretation. The adjustment of the EF in the HPI transforms it from an indicator of absolute footprintscores to an indicator of the position of each country in the overall distribution of footprint scores. Thus, the HPI measuresthe degree to which a country has a higher level of well-being relative to countries with similar levels of environmental con-sumption. It is important to note that this adjustment has major consequences for the resulting indicator. In order to illus-trate this, we calculated an EWEB score by dividing average life satisfaction by the EF per capita and then calculated anotherscore in which the coefficients of variation were matched by adding a constant to the footprint.8,9 The two resulting scoreswere only correlated at .39 (p < .01, N = 105). Thus, this transformation is not trivial and furthermore, the HPI is not a straight-forward ratio. However, as we note below, this is not necessarily a downside of the HPI.

Second, the use of a ratio variable as a dependent variable in regression analyses is statistically questionable, often result-ing in spurious correlations (Kronmal, 1993). The major concern here is the inclusion of an independent variable that is thesame as or highly correlated with either the numerator or denominator of the ratio in the dependent variable. This can resultin a positive or negative effect, respectively, simply because an independent variable is common (or nearly so) to a compo-nent of the dependent variable. Furthermore, interpretation of regression coefficients can be difficult because the indepen-dent variables may differentially affect the numerator and denominator of the ratio in the dependent variable. This meansthat regressing a ratio measure of EWEB on several variables would make it difficult to discern the actual effects of each var-iable and may result in misleading inferences. One solution to this problem would be to treat the EF as an independent var-iable in a model predicting well-being; however, this is not feasible because level of economic development is a keyindependent variable of interest and adding the EF as another independent variable would result in excessive multicolline-arity and thus unstable parameter estimates.

Third, while Dietz et al. (2001) use a measure of objective well-being, the HPI uses a combined measure of objective andsubjective well-being and the Wellbeing/Stress Index (Prescott-Allen, 2001) uses a composite index of objective well-being.Furthermore, Dietz et al. (2001) use a combined measure of environmental impact including both resource consumption andenvironmental quality while the HPI uses the EF alone and the Wellbeing/Stress Index uses a composite index of environ-mental quality. The choice of environmental quality or environmental consumption as the denominator of the ratio can af-fect the results because these are two different types of environmental degradation. As Jorgenson et al. (2009) have noted,countries with higher consumption often face lower levels of degradation within their borders due to the process of ‘ecolog-ically unequal exchange’ by which more-developed countries externalize the environmental costs of their high levels of con-sumption to less-developed countries10; for example, countries with higher resource consumption actually experience lowerlevels of water pollution and deforestation. Thus, combining such measures in the denominator or numerator results in muchless straightforward interpretation. In our view, a simpler approach which selects one type of well-being measure and one typeof environmental impact measure is more useful for understanding EWEB at this early stage of research. Here, we focus on envi-ronmental consumption measures, particularly the EF.

The approach of Dietz et al. (2009), while useful and innovative, has not been adapted for research on the determinants ofcross-national variation in EWEB. Dietz and colleagues estimate countries’ efficiency in enhancing human well-beingthrough the use of economic, natural and human resources, but do not investigate the factors which affect this efficiency.Furthermore, the efficiency measure produced in their approach takes into account economic and human capital inputsin addition to environmental consumption which makes it less amenable to the approach taken here which is concernedwith the factors which may affect EWEB alone.

7 Average life satisfaction is a measure of subjective well-being which we describe at length below in the data and methods section. The product of lifeexpectancy and average life satisfaction is a measure of well-being known as ‘happy life years’ developed by Veenhoven (1996, 2005).

8 For our sample, this constant was calculated to be 5.89.9 Both life satisfaction and ecological footprint are described in detail below.

10 In the ecology literature this is referred to as the ‘‘Netherlands Fallacy’’ (York et al., 2003). The term refers to the pattern in some densely populated andeconomically prosperous nations, such as The Netherlands, which have a relatively clean environment, despite high consumption of resources, because theyimport a large share of resources they consume and export their wastes, thereby externalizing their environmental impacts.


Ng (2008) offers a critical correction to the HPI. He calculates an alternative Environmentally Responsible Happy NationIndex while criticizing several features of the HPI. Among other changes, Ng advocates for the use of CO2 per capita as a mea-sure of the external costs of global environmental disruption contributed by a country, which is different from what Markset al. (2006) claim to be measuring with EF per capita (i.e., the resources consumed by a population). He does so because heviews the EF as not ‘‘internationally acceptable’’ for comparing nations’ environmental impacts. Ng’s final measure is com-puted as adjusted happy life years minus carbon dioxide per capita as a proportion of average world emissions per capitameasured in terms of the proportion of the average happy life years of the current generation. Ng provides no comparisonof the final scores he calculates with the values of the HPI. Thus, it is not clear how much of a difference or improvement hisrevisions make. However, as Ng (2008, p. 444) declares, this ‘‘estimation is meant to be no more than an illustration. . .’’ Fur-thermore, Ng (2008, p. 443) relies on numerous assumptions and estimations of variables in calculating his index andacknowledges that ‘‘An adequate estimate of these variables is beyond the scope of this paper and probably the capabilityof any individual researcher.’’ In view of these limitations we see little grounds suggesting that Ng’s index is an advance onprevious EWEB indicators.

To avoid the main limitations of the approaches above we devised a measure of EWEB that takes into account thesubjective well-being and EFs of countries without the limitations inherent in a ratio measure. Next, we describe thecomponents of this measure and our calculation of it.

3.1. Life satisfaction

We chose to focus on subjective well-being for several reasons. First, most of the previous research toward developing anEWEB indicator has focused on objective indicators of well-being such as life expectancy and infant mortality or combina-tions of objective and subjective well-being indicators. Hence, we already have some understanding of the relationship be-tween the quality of living conditions and threats to the environment. By looking at subjective well-being, we move thatresearch into uncharted territory where there is considerable potential for advancing our understanding of this topic. Second,a focus on subjective, as opposed to objective, well-being is more congruent with the sustainable consumption literature(Jackson, 2005; Manno, 2002; Seyfang, 2009). Findings from this study are, therefore, aligned with that literature and holdpromise for contributing to it. Third, there is a vertiginously growing literature on subjective well-being in the social sci-ences. (For a summary of this literature, see Diener et al., 2009a; Dutt and Radcliff, 2009; Eid and Larsen, 2008; Stevensonand Wolfers, 2008.) Analyzing EWEB in terms of subjective well-being takes this research in a new direction and providesmuch-needed linkages between environmental social science perspectives and this important and burgeoning literature.Veenhoven (1995, 1996), a pioneer in the social science of subjective well-being, has pointed out that objective well-beingindicators are measures of inputs to well-being or quality of life, while subjective measures reflect actual experience of one’squality of life. He further argued that subjective measures of well-being provide a more comprehensive assessment of overallquality of life (Veenhoven, 2002). We can also recognize that available objective measures of well-being are typically limitedto assessing the degree to which basic human needs are met, such as infant mortality, while subjective measures of life sat-isfaction represent a wider range of life experiences. Helliwell and Putnam go even further, arguing that subjective well-being should be ‘‘the ultimate dependent variable in social science’’ (2004, p. 1435). We are in agreement with Stiglitzet al. (2010) that both objective and subjective measures of well-being provide important information about quality of life.However, addressing both types of well-being as a component of EWEB is beyond the scope of this paper. We suggest thatfuture research operationalize, compute and analyze EWEB scores based on various objective measures of well-being andcompare the results to those based on subjective well-being.

We use a measure of national average life satisfaction as our indicator of subjective well-being. Sometimes happiness andwell-being are used interchangeably in the literature. But, there is an important distinction between ‘happiness’ and ‘life sat-isfaction’ and they are typically measured using different types of survey instruments. As Helliwell and Putnam (2004, p.1435) point out, ‘‘Generally speaking, self-ratings of ‘happiness’ turn out to reflect relatively short-term, situation-dependentexpressions of mood, whereas self-ratings of ‘life satisfaction’ appear to measure longer-term, more stable evaluations. . .’’Since we are more concerned with people’s sense of the state of their life experiences rather than fluctuating moods (evenif averaged), life satisfaction more likely than happiness taps this dimension. Measures of life satisfaction have been found tobe valid, reliable indicators within countries and, for the most part, across countries and cultures (Abdallah et al., 2008; Die-ner and Suh, 2000; Helliwell and Putnam, 2004). Indeed, compelling recent research has demonstrated that among US statessubjective well-being correlates highly with measures based on objective well-being, thus indicating that ‘‘subjective well-being data contain genuine information about the quality of human lives’’ (Oswald and Wu, 2010, p. 579). Furthermore,Helliwell (2007) has demonstrated that both suicide rates and life satisfaction data are similarly well explained by the samestatistical model. This suggests that reservations about the use of subjective indicators because of misgivings about theirvalidity compared to objective measures or suggestions that the survey questions of subjective measures may be interpreteddifferently across cultures are misguided. In our own data, the correlation between our measure of life satisfaction and lifeexpectancy, a common measure of objective well-being, is .83 (n = 105; p < .01). 11 For these reasons we focus our presentanalysis of EWEB on subjective life satisfaction.

11 Life expectancy data are from World Bank (2007).


Two survey questions in the Gallup World Poll have been repeatedly used to measure life satisfaction in a large number ofcountries. The first, known as the Cantril self-anchoring ladder, reads as follows:

Please imagine a ladder with steps numbered from 0 at the bottom to 10 at the top. Suppose we say that the top of the ladder(10) represents the best possible life for you and the bottom of the ladder (0) represents the worst possible life for you. If thetop step is 10 and the bottom step is 0, on which step of the ladder do you feel you personally stand at the present time?

The second survey question asks respondents to answer the following question on a scale of 0–10:

All things considered, how satisfied are you with your life these days?

Recently, Bjornskov (2008) raised concerns about the comparability of data based on the Cantril ladder with data basedon the life satisfaction question. However, Helliwell et al. (2009) have allayed these concerns by determining that an averageof the two indicators provides the best measure of life satisfaction. We follow their advice and calculate the mean of thesetwo indicators using country-level mean responses to these questions, as provided by the World Database of Happiness(Veenhoven, 2009).

3.2. Ecological footprint

For our indicator of environmental consumption, we use the per capita ecological footprints (EF) of nations. The EF is ameasure of consumption-based pressure on the environment developed by Wackernagel and Rees (1996). It is constructedfrom five basic forms of human consumption: food, housing, transportation, consumer goods, and services; these categoriescomprise a more direct measure of consumption than GDP which also includes gross domestic investment, net exports, andgovernment expenditures (BEA, 2009). Further, these categories of consumption are converted into a common metric per-mitting harmonized comparisons. The footprint is defined by Rees as ‘‘the area of land and water ecosystems required ona continuous basis to produce the resources that the population consumes, and to assimilate (some of) the wastes thatthe population produces, wherever on Earth the relevant land/water may be located’’ (2006, p. 145). The common metricis, therefore, not based upon market prices like GDP, but the amount of land area at average levels of productivity toreplenish the consumption levels at a specific point in time.

We have chosen the EF as the most suitable indicator of environmental threat for assessing EWEB because it is the mostcomprehensive indicator of the resource demands nations make on the environment. Furthermore, it does not suffer fromthe same limitations as its competitors, as explained above. Using an indicator of local environmental quality is less desirablebecause it does not take into account the resources consumed and ecosystems affected that do not originate or are not lo-cated within that country.

The (EF) is widely considered to be a reliable indicator of country-level threats to the environment and to sustainability.Indeed, it has become one of the most widely used measure of anthropogenic environmental pressure in the environmentalsocial sciences (e.g., Dietz et al., 2007; Hayden and Shandra, 2009; Jorgenson, 2003; Jorgenson and Burns, 2007; Jorgensonet al., 2009; Özler and Obach, 2009; Rosa et al., 2004; York et al., 2003). However, the EF is not without its critics. Perhaps thesingle most repeated criticism of the footprint is that it does not measure actual impacts to the environment (Dietz et al.,2010). However, this criticism is misplaced since the EF is not intended to be an impact measure, but a hypothetical measureof threats to the environment. We have been careful here to emphasize the latter interpretation. Indeed, Rees (2006),co-developer of the EF, acknowledges that the EF as a hypothetical concept likely under-estimates the actual anthropogenicimpact on the environment.

Another criticism, from the developers of the EF, include its omission of key environmental threats such as water con-sumption, toxic and hazardous wastes, pollution in general, and greenhouse gases other than carbon dioxide (Wackernagelet al., 2002). One of the earliest and most durable critiques questions the appropriateness of combining various types of re-source consumption and land in one single indicator and the large contribution of energy consumption to footprint measure-ments (Van den Bergh and Verbruggen, 1999). Still others question the lack of consideration given to the environmentalspace needed for the preservation of non-human species (Özler and Obach, 2009). More recently, Fiala (2008) has criticizedthe assumption of zero greenhouse gas emissions and national borders as arbitrary, the inadequacy of the footprint inaccounting for intensive versus extensive production growth via technological change, and the lack of significant correlationbetween the footprint and measures of land degradation. Despite all these limitations, Jorgenson and Clark (in press) impor-tantly note that the EF remains a useful indicator for hypothesis testing for a number of reasons: It is comprehensive, yetconservative; the calculations are based on rigorous methodology; it is widely accepted and used throughout the social sci-ence disciplines; it has high face validity given its strong correlation with other environmental impact indicators such as car-bon dioxide emissions; and it is available for a wide array of nations.12

Because the EF is a comprehensive, consumption-based measure, all the resources consumed by a country, includingimports, are attributed to its own footprint while its exports are attributed to the footprints of the importing countries(i.e., consumption = production + imports � exports). This is particularly important because, due to international trade, the

12 For a more in-depth review of criticisms of the ecological footprint and responses to them, see Rees (2006) and for recent developments in footprintmethods see the special section of Ecological Economics, vol. 68, issue 7, 2009.

Fig. 2. Scatterplot of average life satisfaction, 2006–2009, and ecological footprint per capita, 2005, with vertical line indicating an ecological footprint of2.1 ha, the global biocapacity per capita (N = 105).


environmental consequences of consumption often occur far from the site of consumption as noted above (Jorgenson et al.,2009; York et al., 2003).

As already mentioned, one major analytical advantage of the EF is that, like money for economic goods, it combines theconsumption of many different types of resources, the use of land for living space, and the production of wastes (currentlylimited to CO2 emissions due to data limitations) into a single common metric: land. Its specific measure is global hectares. Aglobal hectare is a unit (2.41 acres) of biologically productive land or water adjusted to the world average productivity (Rees,2006). The EF includes over 200 resource categories as well as the embodied energy in more than 600 categories of products(Kitzes et al., 2007). A considerable advantage of this measure is that it captures tradeoffs between the consumption of dif-ferent resources (e.g., biomass fuels vs. fossil fuels) (York et al., 2009). For the present analysis, we use the most recent EFdata available, which is for 2005 (Global Footprint Network, 2008a).

3.3. Calculating EWEB

Fig. 2 illustrates the relationship between the EF and average life satisfaction. Note the horizontal spread of observationsat high levels of life satisfaction (i.e., around seven). This indicates, in line with much of the previous research reviewed here,that among countries with high life satisfaction there is wide variation in environmental consumption. Furthermore, notethe observations with life satisfaction around seven near the vertical line at 2.1 global hectares of EF, which is the global bio-capacity per capita available in 2005 (Global Footprint Network, 2008b). This indicates that relatively high levels of well-being are achievable with sustainable levels of consumption. Thus, it is clear that countries vary widely in how efficient theyare at transforming environmental consumption into human well-being.

In order to calculate our indicator of EWEB we borrow a methodological innovation from the macro-comparative politicaleconomy literature on urbanization. Timberlake and Kentor (1983) determined that by regressing urbanization on economicdevelopment (GDP per capita) and taking the residuals, they could measure overurbanization without the drawbacks ofusing a ratio measure (e.g., urbanization/GDP per capita). The residuals represent the level of urbanization relative to thelevel of GDP per capita. Countries with large positive residuals are overurbanized relative to their level of economic devel-opment, while those with large negative residuals are underurbanized. This measure of urbanization has been used innumerous cross-national studies (e.g., Bradshaw, 1985; Walton and Ragin, 1990; Shandra et al., 2003).

We follow the same procedure to calculate our EWEB indicator. We regress average life satisfaction on EF per capita andcompute the unstandardized residuals.13 Countries with large positive residuals have high well-being relative to their level of

13 We use ecological footprint per capita rather than the log of the footprint because we think that in computing the level of life satisfaction relative to thefootprint, the linear scale is more appropriate than the logarithmic. We calculated the EWEB score using both methods and found important differences. Forinstance, whereas the United States (US) has an EWEB score of �1.31 using our preferred model specification, the alternative model using the log of thefootprint produces an EWEB score of �.057. The latter is misleading because despite its high level of life satisfaction (it is the tenth highest), the US also has thelargest footprint in our sample—by far. An EWEB score close to zero gives the impression that the US is about average. On the other hand, our preferred EWEBscore indicates that because other countries are able to achieve as high or higher levels of life satisfaction as the US at much lower levels of consumption. Thismeans that the environmental efficiency of well-being of the US is quite low, not average.

Table 1Unstandardized regression coefficients of ecological footprint per capitaon average life satisfaction.

Ecological footprint, p.c. 0.490 (.042)***

Constant 4.262 (.141)R2 0.571

*** Notes: p 6 .01; N = 105.


environmental consumption while those with large negative residuals have low well-being relative to their level of environ-mental consumption. Thus, the residuals provide a useful way of measuring EWEB absent the limitations of ratio measures.The equation used to estimate the residuals is presented in Table 1.

Table 2 lists each country and their EWEB scores by World Bank (2009) income classification in 2005 rank ordered withineach category. Data necessary for calculating EWEB were available for 105 countries.

We determined that our measure of EWEB is actually quite comparable to the HPI; the two variables are correlated at .89(p < .001). Furthermore, when we use happy life years (the product of life satisfaction and life expectancy) as our measure ofwell-being as is the case with the HPI, we find that the HPI and our measure of EWEB, calculated as a residual, correlate at .94(p < .001). Therefore, the matching of coefficients of variation in the calculation of the HPI makes the EWEB score similar toour EWEB score based on the residuals produced by regression well-being on EF per capita. However, our new measure doeshave two advantages over the HPI: First, we do not combine objective and subjective measures of well-being. This is impor-tant because in order to have analytical clarity, it is important that we determine the correlates of EWEB scores based oneach measure of well-being separately before they are combined. Second, our EWEB indicator is not measured as a ratio,which, as discussed above, reduces the risk of spurious findings.

Table 2Environmental efficiency of well-being (EWEB) scores by world bank income classification in 2005.

HIGH INCOME Turkey �0.39 Cameroon �0.86Netherlands 1.25 Hungary �0.45 LOW INCOMESaudi Arabia 1.06 Russia �0.51 Laos 0.87Finland 1.01 Uruguay �0.75 India 0.82Sweden 0.9 Latvia �0.8 Pakistan 0.69Austria 0.78 Lebanon �0.89 Viet Nam 0.54Belgium 0.59 Botswana �1.14 Uzbekistan 0.5Germany 0.51 Estonia �1.86 Bangladesh 0.44Ireland 0.45 LOWER MIDDLE Nepal 0.34Singapore 0.45 Brazil 1.57 Tajikistan 0.19Israel 0.39 Colombia 1.57 Kyrgyzstan 0.01France 0.36 Dominican Rep. 1.41 Zambia �0.08UK 0.33 Honduras 1.06 Nigeria �0.12Spain 0.33 El Salvador 0.98 Ghana �0.16Italy 0.27 Jordan 0.8 Cambodia �0.24Norway 0.16 Thailand 0.76 Rwanda �0.35Slovenia 0.02 Nicaragua 0.75 Senegal �0.37Canada �0.01 Bolivia 0.62 Mauritania �0.39Japan �0.11 Morocco 0.59 Afghanistan �0.4South Korea �0.17 Indonesia 0.55 Mozambique �0.44New Zealand �0.42 Egypt 0.51 Chad �0.53Australia �0.47 Ecuador 0.47 Uganda �0.54Greece �0.47 Algeria 0.47 Madagascar �0.77Portugal �0.8 Peru 0.43 Mongolia �0.84United States �1.31 Philippines 0.38 Kenya �0.86UPPER MIDDLE Moldova 0.35 Ethiopia �0.87Costa Rica 2.36 Paraguay 0.21 Sierra Leone �1.15Panama 1.34 Kazakhstan �0.05 Mali �1.15Mexico 1.33 Sri Lanka �0.05 Niger �1.19Argentina 1.26 Iran �0.16 Burundi �1.21Malaysia 0.92 Armenia �0.22 Benin �1.44Chile 0.42 Azerbaijan �0.33 Tanzania �1.51South Africa 0.18 Angola �0.41 Burkina Faso �1.52Poland �0.02 Ukraine �0.41 Zimbabwe �1.66Romania �0.04 Belarus �0.51 Togo �1.74Lithuania �0.07 Iraq �0.58Czech Republic �0.1 Georgia �0.7


Now that we have devised an indicator of EWEB, we shift our attention to the factors which may affect this variable. In thenext section we review empirical and theoretical literatures which explicitly or implicitly offer guidance into what indepen-dent variables may be of importance here.

4. Factors influencing EWEB

4.1. Modernization theories

Modernization theories, including ecological modernization theory from sociology and economic modernization theoryfrom economics have emphasized the positive role of economic growth and development in achieving sustainability. Ecolog-ical modernization theory (EMT), in particular, argues that when it comes to sustainable consumption, the focus should notbe on ‘how much,’ and derisively labels those who would focus on ecological limits to consumption as Malthusian (Mol andSpaargaren, 2004). EMT characterizes approaches that advocate for ‘‘less or limits to consumption’’ as taking up a ‘‘demod-ernization’’ position, implicitly suggesting that reducing consumption is tantamount to reversing historical gains in well-being (Mol and Spaargaren, 2004, p. 263). The modernization path to sustainability relies heavily on the assumptions of neo-classical economics and therefore eschews limits to absolute consumption and assumes a strong link between consumptionand well-being.

In addition, economic growth in this view is seen not as an environmental problem, but as a solution. Researchers in thistradition have also suggested the existence of an ‘environmental Kuznets curve’ which purportedly demonstrates that envi-ronmental impacts increase with economic development up to a point after which impacts decline with further economicgrowth (Grossman and Kreuger, 1995). However, many studies have failed to corroborate the existence of the Kuznets, in-verted-U relationship; with most indicating that environmental impacts actually continue to rise with economic growth (Yorket al., 2003, 2009). Nevertheless, it is appropriate to examine from this perspective the prediction of a linear, positive relation-ship between economic development and EWEB, since this is the relationship consistent with modernization theories.

Furthermore, EMT emphasizes the importance of democracy in making states responsive to extra-institutional environ-mental politics, suggesting that more democratic countries will have lower environmental impacts (Mol and Sonnenfeld,2000; Murphy, 2000). This may be especially important in predicting EWEB because previous research has determined thatdemocracy is positively related to life satisfaction (Layard, 2005; Inglehart and Klingemann, 2000). Thus, the theoreticalexpectation is that democracy will be positively related to EWEB.

4.2. Political economy

Political economy perspectives are critical of the link between sustainability and economic growth. The treadmill of pro-duction theory, in particular, argues that economic growth is the main driver of environmental degradation (Gould et al.,2004; Schnaiberg, 1980). Proponents of this view argue that in advanced industrial countries, economic growth is drivenby investments in resource-intensive and highly polluting production technologies which displace labor resulting in theneed for more economic growth to absorb the displaced labor.

The treadmill of production theory describes high income countries as encountering ‘‘decreased social efficiency of nat-ural resource utilization’’ (Gould et al., 2004, p. 297). That is, because of the capitalist political economy, economic growth inthese countries results in each consumer dollar generating more environmental damage and less welfare for individuals thanit did at lower levels of development (Schnaiberg, 1991). This suggests that EWEB may be highest at intermediate levels ofdevelopment; thus, the relationship between economic development and EWEB may follow an inverted-U shape—similar inshape to the environmental Kuznets curve (Grossman and Kreuger, 1995) but with a drastically different interpretation of theconsequences of economic growth for sustainability. 14 Partial evidence for this view comes from Rice (2008) who, in exam-ining bivariate correlations, found fundamental differences in the correlation of environmental consumption with well-beingbetween low and high-income nations. Among low income countries several indicators of objective well-being improve sub-stantially with increased environmental consumption. In sharp contrast, among high income countries increased environmentalconsumption has deleterious effects on several objective well-being indicators. Overall, Rice (2008) argues that this suggeststhat the marginal utility of environmental consumption is substantially higher among developing countries compared to thosecountries at high levels of economic development. Knight and Rosa (2009) demonstrated a similar relationship between envi-ronmental consumption and subjective well-being as illustrated here in Fig. 2.

In addition, income inequality may be an important factor affecting EWEB. In the treadmill of production theory and otherpolitical economy theories, inequality is an important component of the social systems’ effects on the environment (Gouldet al., 2004). Inequality may affect EWEB by disproportionately distributing consumption and whatever well-being benefitsit entails, so that, among marginalized groups and the poor, consumption is restricted while the privileged enjoy high levelsof consumption. Among the former, higher consumption may actually result in higher well-being, but among the latter, the

14 As previously noted, the environmental Kuznets curve predicts that further economic development will lead to environmental sustainability. In contrast, ifeconomic development and EWEB have a similar inverted U-shaped relationship, then further economic development may impede, rather than contribute to,sustainability.


effect is likely to be small or non-existent (Diener and Biswas-Diener, 2002). Higher levels of inequality may also spur status-based competitive consumption which would increase environmental impacts without increasing overall societal well-being(Jackson, 2009). Furthermore, Jorgenson (2004) found that among high income core countries, greater inequality increasesenvironmental consumption.15 In view of Knight and Rosa’s (2009) and Rice’s (2008) findings discussed above, this suggeststhat inequality may be negatively associated with EWEB.

4.3. Sustainable consumption, community, and social capital

The relationship between environmental consumption and subjective well-being has been most explicitly theorized inthe sustainable consumption literature. For instance, Manno (2002) developed the concept of ‘consumption efficiency’ whichis the ratio of personal and social well-being to material consumption. He argues that the goal of sustainable developmentshould be to increase consumption efficiency but that doing so is difficult within a society dominated by ‘‘commoditiza-tion.’’16 Focusing, not on commodities, but on nurturing the ability of social relationships and communities to meet needswould not only result in lower levels of consumption but also higher well-being. This idea of consumption efficiency is congru-ent with Jackson’s (2005) concept of the ‘double-dividend,’ which refers to the possibility that we might be able to achieve high-er well-being while simultaneously reducing environmental consumption. In this view, more effective needs-satisfaction, andthus higher well-being, may be achieved through the dematerialization of satisfiers (i.e., various ways of satisfying needs) bymeeting needs through social relationships. For example, Seyfang (2009) describes local provisioning (e.g., farmers’ markets,cooperatives), mutual aid programs, sharing, and local currencies as examples of community building and social capital whichpromote sustainable consumption. Furthermore, research on subjective well-being has found that one of the strongest predic-tors of national-level life satisfaction is high levels of social capital (Helliwell, 2003; Helliwell and Putnam, 2004; Layard, 2005;Veenhoven, 2008). Countries with high levels of social capital may be able to meet some needs through social relationships andcommunity thereby reducing material consumption and boosting life satisfaction by reinforcing social participation. Thus, thisliterature would lead us to expect a positive relationship between social capital and EWEB.

4.4. Climate

Van de Vliert (2009) identified climate as an important factor influencing life satisfaction, arguing that populations inharsher climates experience lower levels of subjective well-being. Van de Vliert (2009) argues that while colder climate isnegatively correlated with happiness, this relationship is mediated by affluence so that among rich countries colder climateis associated with high levels of happiness (e.g., the Scandinavian countries) while among poor countries colder climate isassociated with low levels of happiness. Rehdanz and Maddison (2005) similarly find that colder winters are associated withlower average happiness. In addition, research in human ecology has found that countries in colder climates exhibit higherlevels of environmental consumption (York et al., 2003). Therefore, we include climate as a control in our model because itmay have a significant effect on EWEB.

4.5. Region

A consistent finding in the subjective well-being literature is that two regions of the world stand out. Latin Americancountries tend to have higher life satisfaction than countries at similar levels of development, while Former Soviet Union(FSU) countries tend to have lower life satisfaction (Engelbrecht, 2009; Helliwell et al., 2009; Inglehart and Klingemann,2000; Inglehart et al., 2008). The Latin American effect may be due to a particular cultural or social context while the ex-So-viet effect may be due to the lasting social effects of the breakdown of the Soviet Union (Helliwell et al., 2009; Inglehart andKlingemann, 2000). Because we produce our EWEB score by regressing life satisfaction on EF, it is likely that these unusuallyhigh or low values show up in the residuals that make up our dependent variable. Therefore, we include two dummy vari-ables in our model to control for these regional effects.

5. Variables and data sources

5.1. Dependent variable

The environmental efficiency of well-being (EWEB), as described above, is measured as the unstandardized residual pro-duced by regressing the average of two measures of average life satisfaction on EF per capita. Life satisfaction data are from

15 Jorgenson (2003), working from an ecostructural world-systems approach, found that, in a sample of countries in all three zones of the world system (i.e.,core, semiperiphery, and periphery), income inequality is negatively associated with environmental consumption measured as the ecological footprint.However, further research (Jorgenson, 2004) indicated that the effect is positive in the core and negative in the other zones. According to this research, world-system position influences level of inequality and ecological footprint as well as the relationship between the two.

16 Karl Marx, of course, introduced the concept of commodification, namely the transformation of all goods and services into commercial commodities;Manno (2002) adopts the conceptually similar term ‘‘commoditization’’ in order to emphasize that this process is active rather than passive.


the Gallup World Poll and were made available in Veenhoven’s (2009) World Database of Happiness. These data cover theyears 2006–2009. The EF data are for 2005 and are from Global Footprint Network (2008a).17

5.2. Independent variables

Economic development is measured by the GDP per capita in 2005 in US dollars. We take the natural log to correct forskewness. We also include the quadratic of GDP per capita in the model. A linear relationship between economic develop-ment and EWEB would indicate support for modernization theories. A negative quadratic relationship would support thepolitical economy perspective of Schnaiberg (1980) and colleagues (e.g., Gould et al., 2004). The variable in the quadraticwas centered before squaring, by subtracting the mean, in order to reduce collinearity with the linear term. These dataare from the World Bank (2007).

Our measure of democracy is for 2005 and is from the 2008 version of the Polity IV dataset (Marshall et al., 2008). We usethe ‘‘Polity Score’’ which conceptualizes types of regime authority as a spectrum ranging from fully institutionalized autoc-racies to fully institutionalized democracies and is measured on a scale from �10 to +10, where �10 is a hereditary monar-chy and +10 is a consolidated democracy.

Inequality is measured as the GINI coefficient for net income in 2000. We use this year of data because it has the widestcoverage of countries.18 The GINI coefficient is a measure of income inequality which ranges from 0 to 100 where 0 is the mostequal and 100 is the most unequal. These data are from Solt’s (2009) Standardized World Income Inequality Database (SWIID).Cross-national GINI coefficient data typically present a tradeoff between coverage and comparability (i.e., income definitionsand reference units); the more countries covered the less comparable the measure for each country. Solt’s SWIID database max-imizes comparability for the largest possible number of countries and years. It does so by standardizing and augmenting theUnited Nation University’s World Income Inequality Database through the use of data from the Luxembourg Income Studyand a custom algorithm. While this dataset is quite new, it is already being employed in research published in top social sciencejournals such as American Political Science Review (Baldwin and Huber, 2010).

The idea of social capital comprises reciprocity, membership in voluntary organizations, and trust (Putnam, 1995). Basedupon data availability, we use the level of trust as our rough indicator of social capital. Previous research has used trust as aproxy for social capital because it has relatively wide coverage among countries (e.g., Engelbrecht, 2009; Helliwell, 2007).Trust is measured with the survey question:

Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing withpeople?

This question was asked in wave 4 of the World Values Survey between 1999 and 2001. Trust is measured as the percent-age of people answering ‘most people can be trusted.’ These data are from World Values Survey Association (2009).

We measure climate with a variable developed and calculated by Van de Vliert (2009). His variable, ‘‘Winters,’’ is the sumof absolute deviations from 22 �C (72 �F) for the average lowest and highest temperatures in the coldest month. This mea-sures the harshness of a country’s winters and is an improvement over average temperature or latitude as a measure ofclimate.

As noted above, we include two dummy variables to control for possible regional effects of Latin American and FormerSoviet Union countries. These variables are coded 1 if a country is in that region and 0 if not. Descriptive statistics for allvariables are presented in Table 3.

6. Model estimation

Because of a large amount of missing data on a key variable, trust, we decided to estimate our multivariate regressionmodel with Arbuckle’s (1997) full information maximum likelihood (MLE) routine with AMOS software. Dealing with miss-ing data by using listwise deletion would have resulted in sample of less than 50 countries. By allowing us to include allcountries for which we have data on the dependent variable, MLE estimation makes it possible to include 105 countriesin our analysis. Furthermore, this estimation procedure provides more consistent and less biased results than listwise orpairwise deletion in OLS regression (Arbuckle, 1997; Jorgenson, 2004). This method has been used by a number of macro-comparative researchers, including Hayden and Shandra (2009), Jorgenson (2004), and Kentor (2001). In order to assessthe reliability of our results, we also computed OLS regression models using listwise deletion (see Appendix). We estimatetwo MLE models: Model 1 includes all independent variables and Model 2 includes only the variables that were significant in

17 While longitudinal ecological footprint data are available for a large number of countries between 1960 and 2005, life satisfaction data are much sparser.The Gallup World Poll, which provides the widest coverage of life satisfaction data (more than 130 countries) only began in 2006. Abdallah et al. (2009) recentlyreleased a rather lengthy longitudinal data set (1961–2005) of the Happy Planet Index for 25 OECD countries, based heavily on estimates of life satisfaction, anda shorter three time-point (1990–2005) dataset including components for 36 countries, both of which may prove useful for future research. However, wesuggest that a cross-sectional analysis that includes a large number of geographically diverse countries at varying levels of development is more suitable andappropriate for assessing the effects on EWEB at this early stage of research.

18 Levels of income inequality have been found to be relatively stable over time, thus the choice of year for these data are not a concern in relation to the othervariables in our analysis (Bergesen and Bata, 2002; Jorgenson, 2003).

Table 4Maximum likelihood estimates for environmental efficiency of well-being(EWEB).

Model 1 Model 2

lnGDP per capita .236** .233**

.118 .116(.058) (.052)

(lnGDP per capita)2 �.204** �.209**

�.066 �.068(.030) (.028)

Winters �.207* �.216**

�.008 �.008(.004) (.003)

Trust .331*** .310***

.018 .017(.006) (.006)

Democracy �.038�.005(.014)

Inequality �.221* �.237*

�.022 �.023(.014) (.014)

Latin America .564*** .555***

1.276 1.255(.243) (.239)

Former Soviet Union �.030�.072(.237)

R2 .500 .485

Notes: N = 105; standardized MLE coefficients are in bold, MLE coefficients arein italics, standard errors are in parentheses.

* p 6 .10 (two-tailed tests).** p 6 .05 (two-tailed tests).

*** p 6 .01 (two-tailed tests).

Table 3Descriptive statistics.

N Mean Std. deviation

EWEB 105 0.00 0.82lnGDP per capita 103 7.65 1.65(lnGDP per capita)2 103 2.69 2.52Winters 105 36.68 22.11Democracy 103 4.91 5.82Inequality 90 38.82 8.47Trust 55 28.38 14.84Latin America 105 0.15Former Soviet Union 105 0.13


the full model. Standardized MLE coefficients, unstandardized MLE coefficients, and standard errors are reported in Table 4.The R-square for the models are moderately strong, with 50% of the variation in EWEB explained in the full model (Model 1)and 48.5% in the reduced model (Model 2). Variance inflation factors (based on the listwise analysis) indicate no seriousproblems with multicollinearity for the models reported in Table 4.

7. Results and discussion

The regression results show that economic development has a negative quadratic relationship with EWEB. The coefficientof the linear term of GDP per capita is positive and significant while the quadratic is negative and significant. This suggeststhat the environmental efficiency with which well-being is produced increases with affluence at low to moderate levels ofeconomic development but declines at high levels. This supports the treadmill of production theory which suggests that thesocial efficiency of natural resource utilization declines at high levels of economic development. This is counter to modern-ization theories (both ecological modernization theory and economic modernization theory). These theories, by equatingconsumption with well-being and arguing that sustainability can only be achieved through higher economic growth, predict


a linear, monotonically increasing relationship between economic development and EWEB. That prediction is called intoquestion by these findings.

The coefficient for democracy is not significant. Hence, the type of regime authority is apparently unrelated to EWEB. Thisfinding is counter to EMT and other modernization theories which suggest that sustainability is intimately linked with liberaldemocracy. Alternative measures of democracy and political freedom were tested in models not reported here. These indi-cators included indexes of political rights and civil liberties, a score based on these two indexes, and a dummy variable indi-cating whether a country is an electoral democracy (all data are from Freedom House, 2010). The results from the Marshallet al. (2008) Polity Score were sustained; democracy had no significant effect on EWEB. This non-finding is consistent withother research results. First, recent research indicates that the effect of democracy on life satisfaction is positive and signif-icant only for countries with higher levels of economic development and also, we would expect, larger EFs (Helliwell andHuang, 2008). Second, the relationship between democracy and EF, contrary to prediction, has been found to be eithernon-significant or positive (York et al., 2003; Marquart-Pyatt, 2010).

Inequality has a negative, significant effect on EWEB. The prediction from political economy that countries with highinequality produce well-being less efficiently than more egalitarian countries is supported by our analysis. We find supportfor the link between economic equality and sustainability measured as EWEB. As argued by Jackson (2009) and Wilkinsonand Pickett (2009) inequality may increase status-based consumption (a constant push to keep up with others) which doesnot add to overall well-being while also resulting in lower levels of overall social health and well-being.

Trust, our single measure of social capital, has a positive, significant relationship to EWEB. This supports recent findingsshowing that social capital is an important factor in explaining cross-national variation in life satisfaction (Helliwell, 2007).Furthermore, it reaffirms the nature of humans as social creatures whose well-being is tied to the relationships with theirfellow humans. The finding also supports claims in the sustainable consumption literature that there may be a ‘double-div-idend’ in reducing consumption, whereby increased community and social capital increase well-being while reducing con-sumption levels (Jackson, 2005; Manno, 2002).

Climate, measured as the harshness of winters, has a significant, negative effect on EWEB. This likely occurs becausecolder climates reduce well-being while increasing resource consumption demands. For example, heating demands placean additional burden on energy supplies.

We found that there is a significant regional effect for Latin American countries. Net of other factors, Latin American coun-tries tend to have higher EWEB. This is consistent with previous research that has showed that Latin American countrieshave higher life satisfaction compared to countries at similar levels of development and may be due to particular featuresof Latin American societies and culture (Helliwell et al., 2009). The Former Soviet Union dummy variable had no significanteffect on EWEB.

In order to determine if there are other important regional effects, and as a sensitivity test, we estimated a MLE modelincluding dummy variables for major regions of the globe: Asia, Africa, Latin America, and the Former Soviet Union countries,with the affluent nations—Europe (including the US, Canada, Australia, and New Zealand) as the comparison category (seeModel 7 in the Appendix). The results differed substantially with income inequality and the effect of the linear term ofGDP per capita dropping below statistical significance while all other variables were substantively the same. The Africa dum-my variable was the only regional variable besides Latin America to come up significant (and negative), but model fit was notsubstantially improved with the addition of the regional variables. In contrast to the results presented in Table 4, an insig-nificant coefficient for GDP per capita and negative coefficient for the squared term suggest that economic development maynot affect EWEB until high levels of development at which further growth decreases EWEB. However, we must interpretthese results with caution because the variance inflation factors (for the unreported listwise analysis) indicate there maybe problems of multicollinearity in this model (VIFs near and above 5). 19 Thus, we have more confidence in the results basedon the more parsimonious model.

As noted above, we also computed OLS regression models employing listwise deletion as a check on the reliability of theMLE results reported in Table 4 (Model 1). We specified two models: (1) a model which excludes the variable with the mostmissing data (i.e., Trust); and (2) a model including all variables. All of these models are reported in the Appendix. The firstmodel (N = 90) has one major discrepancy in that the quadratic of GDP per capita is not significant; otherwise, the findingsare not different from the results in Table 4 (see Model 3). However, further analysis indicated that GDP per capita is onlysignificant in this model when log-transformed, suggesting that the relationship is not linear but rather curvilinear, perhapsdemonstrating diminishing marginal returns (see Models 4 and 5). The listwise model including trust (N = 50) is substan-tially different (see Model 6). Only two variables are significant in this model: the Latin American dummy variable and Trust.We suggest that this is an artifact due to the drastic decrease in sample size. However, it is telling that even with fewer thanhalf the cases as in the full MLE model, the variable trust is statistically significant. With regard to the linear rather than in-verted U-shaped relationship between GDP per capita and EWEB found in the first listwise model, we suggest that becauseMLE estimates are based on a larger sample and are more consistent and efficient than those produced with listwise deletion(Arbuckle, 1997), more weight should be given to these results. However, we do acknowledge that the OLS results do ques-tion the robustness of this finding of a negative quadratic relationship and therefore force us to interpret the findings in the

19 There is no universally agreed upon single VIF standard to judge multicollinearity. VIF cutoff points range from 2.5 (Allison 1999) to 10 (see O’Brien, 2007for a summary). We have, following Greene (2000) chosen a VIF of 5 as a reasonable compromise between those extremes.


MLE models (Table 4) more cautiously. However, all models indicated that the relationship was non-linear and thus in oppo-sition to modernization theories. Nevertheless, further investigation is needed on the relationship between economic devel-opment and EWEB.

8. Conclusion

A variety of sources have conceptualized sustainability as the level of human well-being relative to environmentalthreats and impacts. Researchers have found that measures of this concept indicate that countries vary widely in the envi-ronmental efficiency with which they produce well-being (Dietz et al., 2009; Knight and Rosa, 2009; Marks et al., 2006).Here, we took up this finding as a puzzle to be explained. We devised and computed a new measure of environmentalefficiency of well-being (EWEB) based on the residuals produced by regressing average life satisfaction on EF per capita.Then, we performed a multivariate analysis of the effects of various factors, identified by previous research and theoriza-tion, on EWEB.

Ruut Veenhoven, the most prominent sociologist studying subjective well-being, has advanced the idea that average lifesatisfaction in a society depends on the extent to which that society is ‘‘livable’’ (1995). In this conceptualization, the better asociety meets the needs of its population and advances a high objective quality of life, the happier that population will be.We observe that research on EWEB furthers our understanding of these issues by conceptualizing social well-being in rela-tion to environmental sustainability. EWEB underscores the importance of meeting the needs of society without compromis-ing future generations by depleting natural capital and services. A focus on consumption, as Princen (2005) argues, shifts thequestion of achieving sustainability from production efficiency to consumption ‘sufficiency’. Sufficiency ‘‘is the sense that, asone does more and more of an activity, there can be enough and there can be too much’’ (Princen, 2005, p. 6). In these terms,our analysis identifies key factors that could contribute to the ability of societies to become more sufficient in terms ofachieving livable societies with high levels of well-being but without high levels of consumption. Research on EWEB seeksto contribute to bridging the current divide between the subjective well-being literature and the environmental social sci-ence literature on sustainability and consumption.

Political economy theories question the utility of economic development in achieving sustainability while modernizationtheorists argue that it is the primary means. Our test of these theories indicates support for the former. Indeed, the moststriking finding is that economic development has an inverted U-shaped relationship to EWEB—consistent with the predic-tions of political economy but diametrically opposite to those of modernization theories. The results show that efficiencyincreases at low to moderate levels of consumption, but reaches a turning point and then declines. Additional analyses withOLS models using listwise deletion and an MLE model including additional regional dummies produced discrepant results.However, in both of these models, the relationship was non-linear, suggesting in the former that the relationship is curvi-linear with increases at high levels of economic development increasing EWEB less and less while the latter model suggestedno relationship until high levels of development at which point further increases negatively impacted EWEB. Taken together,the results suggest that the path to sustainability is not necessarily, as economic and ecological modernization theories sug-gest, paved by economic growth. The policy implications of this are clear: If governments focused more on improving well-being directly, rather than indirectly though increasing incomes and consumption, sustainability and high levels of well-being might become more realistic goals. Furthermore, policies designed to alleviate economic inequality may help to movecountries toward sustainability because more egalitarian countries are able to achieve higher levels of well-being with lowerlevels of environmental consumption.

In addition, our findings suggest that countries with higher social capital (measured as trust) transform environmentalconsumption into well-being more efficiently. This supports sustainable consumption scholars who argue that communityand social capital are important factors in achieving the ‘double-dividend’ of lower consumption and higher well-being(Jackson, 2005). As Manno (2002) has argued, meeting needs through social relationships and community ties rather thancommodities decreases material consumption and increases well-being. Thus, policies aimed at increasing social participa-tion and community cohesion may be a key factor in achieving sustainability.

In sum, our findings offer an encouraging alternative to the sustainability strategy dependent on unbounded economicgrowth. Our analyses suggest a more important role for decreasing inequality and increasing social capital in achieving sus-tainability. However, this analysis is only a starting point for further research on EWEB. The present analysis has not ex-hausted the list of potential correlates of EWEB. Future investigation should be undertaken of other factors influencingcross-national variation in environmental efficiency of well-being (EWEB). We have focused on factors internal to nations,such as economic development, social capital, and domestic inequality that have been identified by major theoretical per-spectives in the environmental social sciences, but scholars may wish to identify the external and relational factors involvedas is common from the world-systems perspective of sociology. We recognize the limitations inherent in using cross-sec-tional data, but suggest that given the lack of previous research on EWEB and the lack of longitudinal data with sufficientcoverage, our chosen analysis is appropriate. However, longitudinal analyses should be encouraged when appropriate databecome available. Finally, as is apparent from our discussion of operationalizing EWEB, alternative indicators of well-beingor environmental impacts may be used for measuring EWEB. We suggest that future research should refine and extend thisconcept, explore alternative indicators, and seek to refine its measurement. A promising line of research would be to


compute and compare EWEB indicators based on various measures of subjective well-being, objective well-being, and per-haps a combination of the two.

Further understanding of the link between human use of the environment and well-being can provide empirical guidanceon the path to sustainability by identifying possible social structural changes that can preserve, and maybe enhance, humanwell-being while reducing environmental threats and impacts.

Acknowledgments

We wish to thank Maurie Cohen, Thomas Dietz, John Ehrenfeld, William Rees, Juliet Schor, Rachel Shwom, Andrew Szasz,and Peter Victor for comments on a previous version of this manuscript.

Appendix A

See Table A1.

Table A1Sensitivity analyses of alternative specifications and techniques.

OLS OLS OLS OLS MLEModel 3 Model 4 Model 5 Model 6 Model 7

lnGDP per capita .134** .121** �.065 .119/.059(.061) (.060) (.096) (.060)

(lnGDP per capita)2 �.043 .005 �.153/.051*

(.034) (.046) (.030)

GDP per capita .000(.000)

Winters �.009** �.009** �.007* �.005 �.253/�.009**

(.004) (.004) (.004) (.006) (.004)

Trust .018 .295/.015**

(.007)** (.006)

Democracy �.020 �.021 �.015 .016 �.070/�.010(.015) (.015) (.014) (.021) (.014)

Inequality �.032** �.030** �.036** �.018 �.106/�.011(.013) (.013) (.013) .016 (.014)

Latin America 1.225*** 1.330*** 1.464*** 1.14*** .394/.892***

(.251) (.239) (.232) (.382) (.345)

Former Soviet Union �.127 �.033 �.060 �.398 �.124/�.151(.247) (.237) (.247) (.311) (.290)

Africa �.334/�.613*

Asia �.078/�.151(.257)

Constant .684 .569 1.550 .883(.801) (.799) (.572) (.007)

R2 .414 .402 .382 .398 .533

N 90 90 90 50 105

Notes: In Models 1–4, OLS coefficients are presented with standard errors in parentheses. In Model 5, standardized MLE coefficients are in bold, MLEcoefficients are in italics, standard errors are in parentheses.

* p 6 .10 (two-tailed tests).** p 6 .05 (two-tailed tests).

*** p 6 .01 (two-tailed tests).

References

Abdallah, Saamah, Thompson, Sam, Marks, Nic, 2008. Estimating worldwide life satisfaction. Ecological Economics 65, 35–47.Abdallah, Saamah, Thompson, Sam, Michaelson, Juliet, Marks, Nic, Steuer, Nicola, 2009. The (un)Happy Planet Index 2.0. Why good lives don’t have to cost

the Earth. New Economics Foundation, London.Allison, Paul, 1999. Multiple Regression: A Primer. Pine Forge Press, Thousand Oaks, CA.Arbuckle, J.L., 1997. AMOS User’s Guide. Small Waters Corporation, Chicago, IL.Ayres, Robert U., Martinas, Katalin, 2005. On the Reappraisal of Microeconomics: Economic Growth and Change in a Material World. Edward Elgar,

Cheltenham, UK.


Baldwin, Kate, Huber, John, 2010. Economic versus Cultural Differences: Forms of Ethnic Diversity and Public Goods Provision. American Political ScienceReview, doi:10.1017/S0003055410000419.

Bureau of Economic Analysis (BEA), 2009. National Economic Accounts. Washington, DC. Bureau of Commerce. <http://www.bea.gov/newsreleases/national/gdp/gdpnewsrelease.htm> (accessed October 2009).

Becker, Gary S., Rayo, Luis, 2008. Comments and Discussion. Brookings Papers on Economic Activity. Spring, pp. 88–95.Bergesen, Albert, Bata, Michelle, 2002. Global and national inequality: are they connected? Journal of World-Systems Research 8, 130–144.Bjornskov, Christian, 2008. How comparable are the gallup world poll life satisfaction data? Journal of Happiness Studies. doi:10.1007/s10902-008-9121-6.Bradshaw, York, 1985. Overurbanization and underdevelopment in sub-saharan Africa: a cross-national study. Studies in Comparative International

Development 20, 74–101.Brady, David, Kaya, Yunus, Beckfield, Jason, 2007. Reassessing the effect of economic growth on well-being in less-developed countries, 1980–2003. Studies

in Comparative International Development 42 (1), 1–35.Chambers, Nicky, Simmons, Craig, Wackernagel, Mathis, 2000. Sharing Nature’s Interest: Ecological Footprints as and Indicator of Sustainability. Earthscan,

Sterling, VA.Cohen, Maurie J., 2001. The emergent policy discourse on sustainable consumption. In: Cohen, Maurie J., Murphy, Joseph (Eds.), Exploring Sustainable

Consumption: Environmental Policy and the Social Sciences. Pergamon, Amsterdam, pp. 21–37.Daly, Herman E., Cobb, John B., 1994. For the Common Good. Beacon Press, Boston, MA.Deaton, Angus, 2008. Income, health, and well-being around the world: evidence from the gallup world poll. Journal of Economic Perspectives 22 (2), 53–72.Diener, Ed, Biswas-Diener, Robert, 2002. Will money increase subjective well-being? Social Indicators Research 57, 119–169.Diener, Ed, Oishi, Shigehiro, 2000. Money and happiness: income and subjective well-being across nations. In: Diener, Ed, Suh, Eunkook M. (Eds.), Culture

and Subjective Well-being. The MIT Press, Cambridge, MA, pp. 185–218.Diener, Ed, Suh, Eunkook M., 2000. Measuring subjective well-being to compare the quality of life of cultures. In: Diener, Ed, Suh, Eunkook M. (Eds.), Culture

and Subjective Well-being. The MIT Press, Cambridge, MA, pp. 3–12.Diener, Ed, Helliwell, John F., Kahneman, Daniel (Eds.), 2009a. International Differences in Well-Being. Oxford University Press, New York, NY.Diener, Ed, Helliwell, John F., Kahneman, Daniel, 2009b. Introduction. In: Diener, Ed, Helliwell, John F., Kahneman, Daniel (Eds.), International Differences in

Well-Being. Oxford University Press, New York, NY, pp. ix–xviii.Diener, Ed, Ng, Weiting, Harter, James, Raksha, Arora, 2010. Wealth and happiness across the world: material prosperity predicts life evaluation, whereas

psychosocial prosperity predicts positive feeling. Journal of Personality and Social Psychology 99 (1), 52–61.Dietz, Thomas, York, Richard, Rosa, Eugene A., 2001. Ecological democracy and sustainable development. In: Paper Presented at the 2001 Open Meeting of

the Human Dimensions of Global Environmental Change Research Community, Rio de Janeiro, Brazil, 8 October 2001.Dietz, Thomas, Rosa, Eugene A., York, Richard, 2007. Driving the human ecological footprint. Frontiers in Ecology and the Environment 5 (1), 13–18.Dietz, Thomas, Rosa, Eugene A., York, Richard, 2009. Environmentally efficient well-being: rethinking sustainability as the relationship between human

well-being and environmental impacts. Human Ecology Review 16 (1), 114–123.Dietz, Thomas, Rosa, Eugene A., York, Richard, 2010. Human driving forces of global change: dominant perspectives. In: Rosa, Eugene A., Diekmann, Andreas,

Dietz, Thomas, Jaeger, Carlo C. (Eds.), Human Footprints on the Global Environment: Threats to Sustainability. The MIT Press, Cambridge, MA, pp. 83–134.

Dutt, Amitava Krishna, Radcliff, Benjamin (Eds.), 2009. Happiness, Economics, and Politics: Towards a Multi-disciplinary Approach. Edward Elgar,Cheltenham, UK.

Easterlin, Richard, 1974. Does economic growth improve the human lot? Some empirical evidence. In: David, P.A., Reder, M.W. (Eds.), Nations andHouseholds in Economic Growth. Academic Press, New York, pp. 89–125.

Easterlin, Richard, 1995. Will raising the incomes of all increase the happiness of all? Journal of Economic Behavior and Organization 27, 35–47.Easterlin, Richard A., Angelescu, Laura, 2009. Happiness and Growth the World Over: Time Series Evidence on the Happiness–Income Paradox, Discussion

Paper No. 4060. Bonn, Germany: Institute for the Study of Labor, IZA.Easterlin, Richard, Sawangfa, Onnicha, 2009. Happiness and economic growth: does the cross section predict time trends? Evidence from developing

countries. In: Diener, Ed, Helliwell, John F., Kahneman, Daniel (Eds.), International Differences in Well-Being. Oxford University Press, New York, NY, pp.166–216.

Eid, Michael, Larsen, Randy (Eds.), 2008. The Science of Subjective Well-Being: A Tribute to Ed Diener. Guilford Publications, New York.Engelbrecht, Hans-Jurgen, 2009. Natural capital, subjective well-being, and the new welfare economics of sustainability: some evidence from cross-country

regressions. Ecological Economics 69, 380–388.Fiala, Nathan, 2008. Measuring sustainability: why the ecological footprint is bad economics and bad environmental science. Ecological Economics 67, 519–

525.Freedom House, 2010. Freedom in the World Comparative and Historical Data. <http://www.freedomhouse.org/template.cfm?page=439> (accessed

01.08.10).Frey, Bruno., Stutzer, Alois., 2002. Happiness and Economics. Princeton University Press, Princeton and Oxford.Fuchs, Doris A., Lorek, Sylvia, 2005. Sustainable consumption governance: A history of promises and failures. Journal of Consumer Policy 28, 261–288.Global Footprint Network, 2008b. The Ecological Footprint Atlas 2008. Global Footprint Network, Oakland, CA.Global Footprint Network, 2008a. National Footprints and Biocapacity Accounts, 2008 Edition. <http://www.footprintnetwork.org>.Goldstein, Joshua., 1985. Basic human needs: the plateau curve. World Development 13 (5), 595–609.Gould, Kenneth A., Pellow, David N., Schnaiberg, Allan, 2004. Interrogating the treadmill of production. Organization and Environment 17 (3), 296–316.Graham, Carol, Chattopadhyay, Soumya, Picon, Mario, 2009. The Easterlin and other paradoxes: why both sides of the debate may be correct. In: Diener, Ed,

Helliwell, John F., Kahneman, Daniel (Eds.), International Differences in Well-Being. Oxford University Press, New York, NY, pp. 247–290.Greene, William H., 2000. Econometric Analysis, Fourth ed. Prentice-Hall, Upper Saddle, NJ.Grossman, G.M., Kreuger, A.B., 1995. Economic growth and the environment. Quarterly Journal of Economics 112, 353–378.Hagerty, Michael R., Veenhoven, Ruut, 2003. Wealth and happiness revisited: growing national income does go with greater happiness. Social Indicators

Research 64, 1–27.Hayden, Anders, Shandra, John M., 2009. Hours of work and the ecological footprint: an exploratory analysis. Local Environment 14 (6), 574–600.Helliwell, John F., 2003. How’s life? Combining individual and national variables to explain subjective well-being. Economic Modeling 20 (2), 331–360.Helliwell, John F., 2007. Well-being and social capital: does suicide pose a puzzle? Social Indicators Research 81, 455–496.Helliwell, John F., Huang, Haifang, 2008. How’s your government? International evidence linking good government and well-being. British Journal of

Political Science 38, 595–619.Helliwell, John F., Putnam, Robert D., 2004. The social context of well-being. Philosophical Transactions of the Royal Society of London 359, 1435–1446.Helliwell, John F., Barrington-Lee, Christopher P., Harris, Anthony, Huang, Haifang, 2009. International evidence on the social context of well-being. In:

Diener, Ed, Helliwell, John F., Kahneman, Daniel (Eds.), International Differences in Well-Being. Oxford University Press, New York, NY, pp. 291–327.Hirsch, Fred, 1977. Social Limits to Growth. Harvard University Press, Cambridge, MA.Inglehart, Ronald, 2009. Faith and freedom: traditional and modern ways to happiness. In: Diener, Ed, Helliwell, John F., Kahneman, Daniel (Eds.),

International Differences in Well-Being. Oxford University Press, New York, NY, pp. 351–397.Inglehart, Ronald, Klingemann, Hans-Dieter, 2000. Genes, culture, democracy, and happiness. In: Diener, Ed, Suh, Eunkook M. (Eds.), Culture and Subjective

Well-being. The MIT Press, Cambridge, MA, pp. 165–183.Inglehart, R., Foa, R., Peterson, C., Welzel, C., 2008. Development, freedom, and rising happiness. Perspectives on Psychological Science 3 (4), 264–285.Jackson, Tim, 2005. Live better by consuming less? Is there a ‘double dividend’ in sustainable consumption? Industrial Ecology 9 (1–2), 19–36.

http://dx.doi.org/10.1017/S0003055410000419

http://www.bea.gov/newsreleases/national/gdp/gdpnewsrelease.htm

http://www.bea.gov/newsreleases/national/gdp/gdpnewsrelease.htm

http://dx.doi.org/10.1007/s10902-008-9121-6

http://www.freedomhouse.org/template.cfm?page=439

http://www.footprintnetwork.org


Jackson, Tim, 2009. Prosperity Without Growth: Economics for a Finite Planet. Earthscan, London.Jackson, Tim, Michaelis, Laurie, 2003. Policies for Sustainable Consumption. Report to the UK Sustainable Development Commission.Jorgenson, Andrew K., 2003. Consumption and environmental degradation: a cross-national analysis of the ecological footprint. Social Problems 50 (3), 374–

394.Jorgenson, Andrew K., 2004. Uneven processes and environmental degradation in the world-economy. Human Ecology Review 11 (2), 103–117.Jorgenson, Andrew K., Burns, Thomas J., 2004. Globalization, the environment, and infant mortality: a cross-national study. Humboldt Journal of Social

Relations 28 (1), 7–52.Jorgenson, Andrew K., Burns, Thomas J., 2007. The political-economic causes of change in the ecological footprints of nations, 1991–2001: a quantitative

investigation. Social Science Research 36, 834–853.Jorgenson, Andrew K., Clark, Brett, 2011. Societies consuming nature: a panel study of the ecological footprints of nations, 1960–2003. Social Science

Research 40 (1), 226–244.Jorgenson, Andrew, Austin, Kelly, Dick, Christopher, 2009. Ecologically unequal exchange and the resource consumption/environmental degradation

paradox: a panel study of less-developed countries, 1970–2000. International Journal of Comparative Sociology 50 (3–4), 263–284.Kentor, Jeffrey, 2001. The long-term effects of globalization on income inequality, population growth, and economic development. Social Problems 48 (4),

435–455.Kitzes, Justin, Peller, Audrey, Goldfinger, Steve, Wackernagel, Mathis, 2007. Current methods for calculating national ecological footprint accounts. Science

for Environment & Sustainable Society 4 (1), 1–9.Knight, Kyle W., Rosa, Eugene A., 2009. The environmental costs of life satisfaction: a cross-national empirical test. In: Presented at the SCORAI Workshop on

Individual Consumption and Systemic Societal Transformation. Clark University, Worcester, Massachusetts, October 15–17, 2009.Kronmal, Richard A., 1993. Spurious correlation and the fallacy of the ratio standard revisited. Journal of the Royal Statistical Society 156 (3), 379–392.Layard, Richard, 2005. Happiness: Lessons from a New Science. Penguin Press, New York.Manno, Jack, 2002. Commoditization: consumption efficiency and an economy of care and connection. In: Princen, Thomas, Maniates, Michael, Conca, Ken

(Eds.), Confronting Consumption. The MIT Press, Cambridge, MA, pp. 23–42.Marks, Nic, Abdallah, Saamah, Simms, Andrew, Thompson, Sam, 2006. The (un)Happy Planet Index: An Index of Human Well-being and Ecological Impact.

New Economics Foundation, London.Marks, Nic, Thompson, Sam, Eckersly, Richard, Jackson, Tim, Kasser, Tim. 2007. Well-being and sustainable development: relationships, challenges, and

policy implications. In: Presentation at the Defra/Sustainable Development Commission Well-Being Conference, May 22, 2007.Marquart-Pyatt, Sandra T., 2010. Environmental sustainability: a closer look at factors influencing national ecological footprints. International Journal of

Sociology 40 (2), 65–84.Marshall, Monty G., Jaggers, Keith, Gurr, Ted Robert, 2008. Polity IV Project: Political Regime Characteristics and Transitions, 1800–2008. Center for Systemic

Peace, Severn, MD. <http://www.systemicpeace.org/polity/polity4.htm> (accessed January 2009).Mazur, Allan, Rosa, Eugene, 1974. Energy and life-style. Science 186, 607–610.Mol, Arthur P.J., Sonnenfeld, David A. (Eds.), 2000. Ecological Modernization around the World: Perspectives and Critical Debates. Frank Cass, London.Mol, Arthur P.J., Spaargaren, Gert, 2004. Ecological modernization and consumption: a reply. Society and Natural Resources 17, 261–265.Murphy, Joseph, 2000. Ecological modernisation. Geoforum 31 (1), 1–8.Ng, Yew-Kwang, 2008. Environmentally responsible happy nation index: towards an internationally acceptable national success indicator. Social Indicators

Research 85, 425–446.Nussbaum, Martha C., Sen, Amartya (Eds.), 1993. The Quality of Life. Oxford University Press, Oxford.O’Brien, Robert M., 2007. A caution regarding rules of thumb for variance inflation factors. Quality and Quantity 41, 673–690.Oswald, Andrew J., Wu, Stephen, 2010. Objective confirmation of subjective measures of human well-being: evidence from the USA. Science 327, 576–579.Özler, S. Ilgü, Obach, Brian K., 2009. Capitalism, state economic policy and ecological footprint: an international comparative analysis. Global Environmental

Politics 9 (1), 79–108.Prescott-Allen, Robert, 2001. The Well-Being of Nations: A Country-by-Country Index of Quality of Life and the Environment. Island Press, Washington, DC.Preston, Samuel H., 1975. The changing relation between mortality and economic development. Population Studies 29 (2), 231–248.Princen, Thomas, 2005. The Logic of Sufficiency. MIT Press, Cambridge, MA.Putnam, Robert, 1995. Bowling alone: America’s declining social capital. Journal of Democracy 6 (1), 65–79.Redclift, Michael, 1996. Wasted: Counting the Costs of Global Consumption. Earthscan, London.Rees, William E., 2006. Ecological footprints and bio-capacity: essential elements in sustainability assessment. In: Dewulf, J., Van Langenhove, Herman

(Eds.), Renewables-based Technology: Sustainability Assessment. John Wiley and Sons, Chichester, UK, pp. 143–158.Rehdanz, Katrin, Maddison, David, 2005. Climate and happiness. Ecological Economics 52, 111–125.Reid, W.V., Mooney, H.A., Cropper, A., Capistrano, D., Carpenter, S.R., Chopra, K., Dasgupta, P., Dietz, T., Duraiappah, A.K., Hassan, R., Kasperson, R., Leemans,

R., May, R.M., McMichael, T. (A.J.), Pingali, P., Samper, C., Sholes, R., Watson, R.T., Zakri, A.H., Shidong, Z., Ash, N.J., Bennett, E., Kumar, P., Lee, M.J.,Raudsepp-Hearne, C., Simons, H., Thonell, J., Zurek, M.B., 2005. Ecosystems and Human Well-Being: Synthesis. Island Press, Washington, DC.

Rice, James, 2008. Material consumption and social well-being within the periphery of the world economy: an ecological analysis of maternal mortality.Social Science Research 37, 1292–1309.

Rosa, Eugene A., 1997. Cross national trends in fossil fuel consumption, societal well-being and carbon releases. In: Stern, P.C., Dietz, T., Ruttan, V.W., Socolow,R.H., Sweeney, J.L. (Eds.), Environmentally Significant Consumption: Research Directions. National Academy Press, Washington, DC, pp. 100–109.

Rosa, Eugene A., York, Richard, Dietz, Thomas, 2004. Tracking the anthropogenic drivers of ecological impacts. Ambio 33 (8), 509–512.Rothman, Dale, 1998. Environmental Kuznets curves—real progress or passing the buck? A case for consumption-based approaches. Ecological Economics

25, 177–194.Royo, Monica Guillen, 2007. Well-being and consumption: towards a theoretical approach based on human needs satisfaction. In: Bruni, Luigino, Porta, Pier

Luigi (Eds.), Handbook on the Economics of Happiness. Edward Elgar, Cheltenham, UK, pp. 151–169.Samuel-Johnson, Kim, 2001. Environmental Sustainability Index. World Economic Forum, Davos, Switzerland. <http://sedac.ciesin.columbia.edu/es/esi/

archive.html> (accessed April 2009).Samuelson, Paul, 1938. A note on the pure theory of consumers’ behaviour. Economica 5, 61–71.Schnaiberg, Allan, 1980. The Environment: From Surplus to Scarcity. Oxford University Press, New York.Schnaiberg, Allan. 1991. The political economy of consumption: ecological policy limits. In: Paper Presented at the 1991 Annual Meeting of the American

Association for the Advancement of Science, February 18, 1991.Scitovsky, Tibor, 1976. The Joyless Economy. Oxford University Press, New York.Sen, Amartya, 1993. The Economics of Life and Death. Scientific American, May.Seyfang, Gill, 2009. The New Economics of Sustainable Consumption: Seeds of Change. Palgrave Macmillan, New York.Shandra, John M., London, Bruce, Williamson, John B., 2003. Environmental degradation, environmental sustainability, and overurbanization in the

developing world: a quantitative, cross-national analysis. Sociological Perspectives 46 (3), 309–329.Solt, Frederick, 2009. Standardizing the world income inequality database. Social Science Quarterly 90(2), 231–242. SWIID Version 2.0, July 2009.Stevenson, Betsey, Wolfers, Justin, 2008. Economic Growth and Happiness: Reassessing the Easterlin Paradox. Brookings Papers on Economic Activity,

Spring, pp. 1–87.Stiglitz, Joseph, Sen, Amartya, Fitoussi, Jean-Paul, 2010. Report by the Commission on the Measurement of Economic Performance and Social Progress.

<www.stiglitz-sen-fitoussi.fr>.

http://www.systemicpeace.org/polity/polity4.htm

http://sedac.ciesin.columbia.edu/es/esi/archive.html

http://sedac.ciesin.columbia.edu/es/esi/archive.html

http://www.stiglitz-sen-fitoussi.fr


Timberlake, Michael, Kentor, Jeffrey, 1983. Economic dependence, overurbanization, and economic growth: a study of less-developed countries. TheSociological Quarterly 24 (4), 489–507.

Van de Vliert, Evert, 2009. Climate, Affluence, and Culture. Cambridge University Press, NY.Van den Bergh, Jeroen C.J.M., Verbruggen, Harmen, 1999. Spatial sustainability, trade and indicators: an evaluation of the ‘ecological footprint’. Ecological

Economics 29, 61–72.Veenhoven, Ruut, 1995. The cross-national pattern of happiness. Social Indicators Research 34, 33–68.Veenhoven, Ruut, 1996. Happy life expectancy: a comprehensive measure of quality-of-life in nations. Social Indicators Research 39, 1–58.Veenhoven, Ruut, 2002. Why social policy needs subjective indicators. Social Indicators Research 58, 33–45.Veenhoven, Ruut, 2005. Apparent quality-of-life in nations: how long and happy people live. Social Indicators Research 71, 61–68.Veenhoven, Ruut, 2008. Sociological theories of subjective well-being. In: Eid, Michael, Larsen, Randy (Eds.), The Science of Subjective Well-Being: A Tribute

to Ed Diener. Guilford Publications, New York, pp. 44–61.Veenhoven, Ruut, 2009. World Database of Happiness, Erasmus University Rotterdam. <http://worlddatabaseofhappiness.eur.nl>. (retrieved September

2009).Vemuri, Amanda W., Costanza, Robert, 2006. The role of human, social, built, and natural capital in explaining life satisfaction at the country level: toward a

National Well-Being Index (NWI). Ecological Economics 58, 119–133.Vincent, Jeffrey R., Panayotou, Theodore, 1997. Consumption: challenge to sustainable development. . . or distraction. Science 276, 53–57.Wackernagel, Mathis, Rees, William, 1996. Our Ecological Footprint: Reducing Human Impact on Earth. New Society, Gabriola Island, British Columbia.Wackernagel, Mathis, Schulz, Niels B., Deumling, Diana, Linares, Alejandro Callejas, Jenkins, Martin, Kapos, Valerie, Monfreda, Chad, Loh, Jonathan, Myers,

Norman, Norgaard, Richard, Randers, Jørgen, 2002. Tracking the ecological overshoot of the human economy. Proceedings of the National Academy ofSciences 99 (14), 9266–9271.

Walton, John, Ragin, Charles, 1990. Global and national sources of political protest: third world responses to the debt crisis. American Sociological Review55 (6), 876–890.

Wilkinson, Richard, Pickett, Kate, 2009. The Spirit Level: Why Greater Equality Makes Societies Stronger. Bloomsbury Press, New York.World Bank, 2007. World Development Indicators (CD). World Bank, Washington, DC.World Bank, 2009. Historical Income Classifications. World Bank, Washington, DC. <http://www.worldbank.org> (retrieved August 2009).World Commission on Environment and Development (WCED), 1987. Our Common Future. Oxford University Press, New York.World Values Survey Association, 2009. World Values Survey 1981–2008 Official Aggregate v.20090901. Aggregate File Producer: ASEP/JDS, Madrid

<worldvaluessurvey.org>.Worldwatch Institute, 2004. State of the World 2004. W.W. Norton & Company, New York.York, Richard, Rosa, Eugene A., Dietz, Thomas, 2003. Footprints on the earth: the environmental consequences of modernity. American Sociological Review

68, 279–300.York, Richard, Rosa, Eugene A., Dietz, Thomas, 2009. A tale of contrasting trends: three measures of the ecological footprint in China, India, Japan, and the

United States, 1961–2003. Journal of World Systems Research 15 (2), 134–146.

http://worlddatabaseofhappiness.eur.nl

http://www.worldbank.org

http://worldvaluessurvey.org

Documents

Social Science Research K.W. Knight, E.A. Rosa/Social Science Research 40 (2011) 931–949 the well-being of the world’s poor.’’ Their analyses indicate that for various measures