There is a growing dissatisfaction with using standard measures of affluence, such as gross domestic product, as the sole conceptualization of human well-being. Experiments are underway with alternative metrics of well-being as ways of informing both research and policy. It is thus important to develop a theory of the production of human well-being to parallel theories of economic development and growth. The traditions of work in growth theory, sustainability theory, and household production functions provide the basis for an emerging structural human ecology of human well-being. Structural human ecology emphasizes the use of manufactured, natural, and human resources in producing well-being but is also attentive to the ways social structure shapes the production of well-being. While this approach is promising, several conceptual issues need to be addressed for it to realize its potential. In particular, we need greater clarity regarding measures of well-being and the ethical theory that underpins them and clearer thinking about the relationship between resources and capital.
Forty years ago a prescient analysis by Mazur and Rosa challenged long-standing assumptions about the relationship between natural resource use and human well-being (Mazur 2013a, 2013b; Mazur and Rosa 1974). At the time many theorists and policy makers assumed that increasing energy consumption per capita was essential for improving the standard of living; some even equated energy use per capita with development (Dietz, Burns, and Buttel 1990; Rosa 1983; Rosa, Machlis, and Keating 1988). Mazur and Rosa showed that when energy consumption was at low to moderate levels there was a strong, positive, and roughly linear relationship between per capita energy consumption and various aspects of lifestyle. But they also demonstrated that once a modest threshold was reached, increased energy consumption had no relationship to well-being. Easterlin (1974) made essentially the same argument regarding conventional economic growth and well-being—that beyond a modest threshold, increased affluence did not improve well-being.1 At about the same time, the wisdom or even the feasibility of growth was being questioned (Daly 1973; Georgescu-Roegen 1971; Meadows et al. 1972). If growth was generating environmental problems but was not substantially enhancing well-being, could and should growth continue? Of course, a counterargument was always articulated: with much of the world struggling to achieve a minimal level of consumption, further growth was justified despite its environmental costs. By the 1980s the discourse about growth was typically framed around concepts of sustainability and sustainable development.2 In most discussions of sustainable development economic growth was considered essential, but compared to conventional development arguments special attention was given to environmental impacts and resource limits. In a sense, the decoupling of well-being and environmental stress identified by Mazur and Rosa gained acceptance, but the decoupling of well-being and economic growth identified by Easterlin did not.
Despite a continued consensus on the importance of conventional economic growth and a narrow focus on growth in affluence as the goal of development policy, there have long been calls for alternative ways of thinking about well-being. As early as 1968, Robert Kennedy argued, “We cannot measure national spirit by the Dow Jones Average, nor national achievement by the Gross National Product” (Kennedy 1968). Efforts to develop social indicators emerged in the 1930s roughly in parallel with the initial development of the standard economic indicators such as gross domestic product (GDP) per capita (Bulmer 1983; Hecht 2005; Kenessey 1994; Ogburn 1933).3 Since World War II the use of economic indicators has been standardized and routinized. But until recently, alternative measures of well-being have languished, even as dissatisfaction with equating well-being with the size of an economy has persisted.
Within development research and development policy analysis, alternatives to GDP per capita are gaining some headway. For over two decades, indicators such as the Human Development Index (HDI) have been used to complement the standard national economic accounts (Anand and Sen 1994; ul Haq 2003; United Nations Development Programme 2014). The HDI combines three measures of well-being—life expectancy at birth, level of education, and gross national income per capita (in previous versions GDP per capita)—into a single measure. Recently, a number of governments have begun experimenting with measures of well-being based on happiness or life satisfaction as reported on high-quality national surveys. In France, the “Sarkozy Commission” argued that GDP per capita and related measures from conventional national economic accounting were not sufficient metrics for assessing policy and setting goals (Stiglitz, Sen, and Fitoussi 2009). The British government has issued an initial series of reports on subjective well-being as a complement to standard economic indicators (Cameron 2010; Self 2014; UK Department for Environment 2011). The Organisation for Economic Co-operation and Development and the US National Research Council have both issued reports on the measurement of subjective well-being in the service of national reporting (Organisation for Economic Co-operation and Development 2013; Stone and Mackie 2013; US National Research Council 2012). It is clear that policy discussions are beginning to explore concepts of human well-being that are broader than economic activity per capita.
This shift away from equating conventional measures of economic activity with well-being raises important challenges and opportunities for the sociology of development and environmental sociology. Here I raise two questions we must address if we are to take alternatives to GDP seriously.
—What are appropriate measures of human well-being? I will argue that selecting a measure of well-being is to a substantial degree an ethical choice. While most environmental and development sociologists bring ethical concerns to their work, little of our literature analyzes the ethical foundations that underpin issues such as selection of a measure of well-being. This is in contrast to conventional economics, where, for better or worse, arguments for economic growth are justified by the logic that increased income allows individuals to satisfy their preferences and that preference satisfaction increases utility. Thus a utilitarian ethical theory is only a few steps away from focusing on GDP per capita as an appropriate measure of well-being. The HDI is motivated largely by Sen's capabilities ethics (Nussbaum and Sen 1993; Sen 1999a, 1999b). But sociologists have not had much voice in these discussions. This is clearly an area that warrants further work. Nonetheless, I will submit that there are at least two measures of well-being—life expectancy and self-reported subjective well-being—that reflect a normative consensus and are methodologically defensible, so that empirical work can proceed even as we think more carefully about the normative framework that underpins our analysis.
—How must we reconceptualize models of growth and development to reflect an end point of human well-being? Much of development economics is about the factors that drive or retard growth. Much of the sociology of development and most macro-scale environmental sociology examine the adverse effects of conventional economic growth. But if we engage the arguments that affluence should not be equated with well-being, we will need a theory of what drives changes in well-being. A production function that links resources to well-being is a useful starting point for our thinking. But to specify such a function we first need to carefully delineate what we mean by resources, when resources are and are not capital, and how social structure shapes the use of resources to produce well-being.
Before suggesting approaches to answering these two questions, I will review the standard narrative of growth and sustainability theory to provide a context for current discussions. This leads us to an emerging approach that stands at the convergence of several streams of previous work. “Structural human ecology,” as this synthetic line of work is labeled, provides a standpoint both for developing a normative framework for well-being and for conceptualizing the production of well-being.
GROWTH AND SUSTAINABILITY THEORY
Much of classical political economy was concerned with understanding what led to increased economic activity, what we would now call increased GDP (Richerson and Boyd 1997–98). The Wealth of Nations is centrally about economic growth (Smith 1801). Malthus in his Essay on Population ( 1992), despite many limits in the analysis, was at least constructing an argument about what would lead to increased or degraded well-being. Ricardo (1817), after arguing for a labor theory of value and the distinction between use value and exchange value, devotes much of his analysis to the drivers of economic growth, including the problem of diminishing returns. Given the poverty of much of the population during the late eighteenth and early nineteenth centuries it was understandable that these theorists saw increases in economic activity as desirable.4 Nor were classical growth theorists unaware of the relationship between the size of the economy and national power. And, as has always been the case, arguments for economic growth find support from those who feel they benefit from such growth, a group that always includes owners of capital and sometimes includes labor.
As modern economics developed a consensus on a utilitarian normative theory and on the assumption of rational choice as the basis for decision making, it was inevitable that increasing affluence would be seen as essentially equivalent to human well-being. More affluence would allow more choices, and under rational choice theory these choices would inevitably enhance utility (Stigler 1950a, 1950b).5 So the move away from accepting increased GDP per capita as an adequate measure of well-being is also a rejection of rational choice theory as an adequate model for human behavior, and vice versa.6
After World War II, the theory of economic growth was put into practice. Japan, much of China, parts of Southeast Asia and North Africa, many Pacific Islands, many nations in Europe, and the western regions of the Soviet Union had been physically devastated. The war had destroyed much of the manufactured resources needed for industrial production —heavy machinery, factories and refineries, railroads and bridges, and so on. In response, economic policy after the war strongly emphasized rebuilding this infrastructure, using both bilateral aid and loans from the newly created World Bank (formally the International Bank for Reconstruction and Development and the International Development Association) as mechanisms. A major justification for this program was that economic growth fostered by rebuilding physical infrastructure would enhance well-being. But there were certainly other motivations. The West and the Soviet Union were starting the face-off that became the Cold War. Increasing industrial and agricultural production was essential to support re-arming for conflicts that many thought were inevitable. Growth in conventional economic output was also seen as a way of staving off economic policy conflicts that had strengthened during the Depression (Schnaiberg 1980). The emergence of what Eisenhower termed the “military-industrial complex” provided a mechanism through which public funds could be used to finance investment both in industrial production for defense and in the public infrastructure, such as roads and ports, that supported it (Pilisuk and Hayden 1965). Investments in the “Third World” provided new markets and new sources of raw materials. Even the vexing problem of inequality might be resolved through growth if one accepted the argument that growth ultimately led to lower levels of inequality.7 So it is not surprising that growth in GDP, and the investment in physical infrastructure to support such growth, took center stage in policy goals around the world.
From Development Theory to Sustainability Theory
During the 1960s and 1970s the environment became a more central part of the policy agenda within most nations and in the global discourse on development. By the early 1980s the emphasis on conventional economic growth was tempered with concerns about the effects of growth on the environment, about limits to growth entrained in the exhaustion of natural resources, and about the link between affluence and well-being. Arguments for sustainable development or sustainability tried to integrate the traditional emphasis on economic growth with the emerging concerns about the environment and well-being. John Robinson's (2004) insightful essay on these concepts suggests that Lester Brown (1981) was first to use the concept of sustainability, although core ideas go back at least to the writings of John Wesley Powell (Worster 2001) and George Perkins Marsh (Marsh  1965). The International Union for the Conservation of Nature under the leadership of Lee Talbot used the term sustainable development and brought the concept into policy circles (International Union for the Conservation of Nature 1980). The Brandt report engaged issues of energy, population growth, and the environment (Independent Commission on International Development Issues 1980). But the popularity of the term sustainable development dates to the famed Bruntland report, Our Common Future (World Commission on Environment and Development 1987).
The idea of “sustainable development” continued the emphasis on economic growth—increased economic activity per capita—even while paying attention to the harm done to the environment. For example, Our Common Future calls for a five- to tenfold increase in industrial activity to meet the needs of the poor. In this context sustainable development is defined as a process that “meets the needs of current generations without compromising the ability of future generation to meet their own needs” (World Commission on Environment and Development 1987:23). This standard definition is wholly anthropocentric. Changes in the biophysical environment and harm to other species may be tolerated if those actions don't put at risk the possibility that future generations can meet their needs. Of course this approach to sustainable development could justify the protection of a pristine environment and biodiversity to the extent that they are needed by humans. But under the sustainable development approach the environment is clearly a secondary consideration to meeting human needs and alleviating poverty.
Robinson (2004) also notes that the sustainable development approach led to a strong emphasis on technological fixes to environmental and resource problems. This was consistent with the legacy of conventional development. In both the capitalist West and the communist nations, development projects usually took the form of building new physical infrastructure. New technology for agriculture was also a critical element in efforts to increase food production—the “Green Revolution.” New seed types and the irrigation systems, pesticides, fertilizers, and credit systems to support them were a key part of development efforts. This shift in agriculture also was based on a shift from production for subsistence—distribution to the household and community—to production for the market.
The “technological fix” approach was not limited to hardware. The right institutions were seen as important as well. Initially there was struggle between communist and capitalist approaches to fostering growth. But by the 1980s, competitive markets with minimal government interference in the economy and no barriers to free trade (neoliberalism) became the conventional view of how to foster economic growth (Harvey 2005). The idea that the market was a perfect institution to solve the problems of developing countries reached its height in the 1980s, when “structural adjustments” were imposed on countries that required aid in response to their debt burdens (Bruce 1994; Harvey 2005; Peet 1996; Perkins 2005). From World War II onward, development policy was implemented largely through loans for the construction of physical infrastructure. Substantial funds were siphoned out of productive use via corruption, and many projects that were built never fulfilled their potential. The payment on debts from these loans was crippling for many nations. In exchange for renegotiations of these debts, governments were required to make policy changes that remodeled their countries into something closer to the neoliberal ideal of capitalism—what have come to be called structural adjustments. It was presumed by the advocates of these restructuring policies that well-functioning capitalist market institutions and free trade would lead to economic growth that would alleviate poverty (e.g., Sachs and Warner 1995). Problems of inequality were seen as secondary to problems of inefficiency and as being eventually alleviated by growth. In parallel, it was argued that the adverse effects of economic growth on the environment would eventually be ameliorated by affluence as nations moved along an environmental version of the Kuznets curve or underwent ecological modernization (Dietz, Rosa, and York 2010, 2012; Mol 1995, 2010; Selden and Song 1994; York, Rosa, and Dietz 2010). Many studies have examined the effects of these policies across a range of countries, and while results are mixed, the general pattern seems to be that structural adjustments were problematic for the environment and human well-being (Clapp and Dauvergne 2005; Glover 1995; Pfeiffer and Chapman 2010; Shandra, Shircliff, and London 2011; Shandra, Shor, and London 2008; Shandra, Shor, Maynard, and London 2008).
There were other nonhardware “fixes” intended to achieve economic growth. From the 1970s on, Garrett Hardin's hugely influential essay “The Tragedy of the Commons” erroneously convinced many that community-based management of resources was doomed to failure, despite, in many cases, the evidence of decades or even centuries of success (Dietz, Ostrom, and Stern 2003; Ostrom et al. 2002). It's not clear how much influence his arguments had on policy, since in many cases loss of community control predated his essay (Adams and McShane 1996; Anderson and Grove 1987; Peluso 1992; Peluso and Vandergeest 2001). But the idea that only markets or governments could successfully manage common-pool resources was part of the suite of beliefs that underpinned development policy in the late twentieth century. Indeed, arguments that structural adjustments would be beneficial for the environment often echoed Hardin's logic that only privately held (as opposed to community-managed) natural resources would be managed effectively.
From Growth to Sustainability Theory
The need for economic growth remains a taken-for-granted assumption in most discussions of development, and the tendency to equate GDP per capita with well-being persists. But even in the 1970s, before sustainable development was part of the lexicon, a number of arguments for “limits to growth” had already attracted broad attention. These included Paul Ehrlich's (1968) influential The Population Bomb, the highly visible Limits to Growth study (Meadows et al. 1972), and the Carter administration's Global 2000 report, which was largely suppressed by the Reagan administration (Barney 1980; Speth 2008). While these analyses were sometimes very simplistic and have often been labeled neo-Malthusian, they did initiate serious questions about whether continuous increases in economic activity, with its attendant stress on the environment as both a source of raw materials and a sink for waste, could expand indefinitely. While not garnering as much public attention, Kenneth Boulding (1966) and Herman Daly (1973) were laying the groundwork for arguments that economic growth as conventionally defined—an increase in GDP—was not an adequate goal. Daly, in particular, was an early advocate for metrics of well-being that could complement or substitute for GDP (Daly and Cobb 1989; Hecht 2005). The debate between those who questioned the desirability and feasibility of continued growth and those who advocated for growth has led to a renewed examination of what is necessary for producing goods and services. This rethinking of the idea of a production function lays the groundwork for addressing questions of how we produce well-being.
The core of the debate is a conflict about whether elements of the production function can be substituted for one another. Both conventional economic growth theory and sustainability theory examine the resources that are deployed to produce a desired outcome, in particular increased economic activity, or GDP.8 Most analyses focus on three key inputs: manufactured resources, natural resources, and human resources. Other factors may matter as well. Indeed neoclassical economic growth theory argues that much of economic growth is the result of changes in technology; thus the emphasis on technological fixes in many development policies (Solow 1957, 1994).
If the goal of sustainability is to enhance well-being at present while ensuring that future generations have sufficient wherewithal to provide for their well-being, can we deplete resources in one category but increase resources in another and still be providing an adequate portfolio to future generations (Neumayer 2010)? Advocates of weak sustainability suggest that there is substantial substitutability. So, for example, it might be a wise policy to dam a wild river for electrical production or convert high-biodiversity rainforest to commercial monocrop agriculture if that depletion of natural resources led to investments in infrastructure (manufactured resources) and education (human resources). Future generations would inherit a depleted portfolio of natural resources but could still achieve high well-being because of the increased stock of manufactured and human resources.
In contrast, advocates of strong sustainability argue that future generations have a right to an undiminished stock of each resource. One underpinning for this logic is ethical. Natural resources make a unique contribution to human well-being, and investments in human resources and manufactured resources cannot compensate for degradation of ecosystems. The other argument is that the technical possibility for substitution is limited. For example, greenhouse gas emissions may lead to a level of disruption of the global climate system that cannot be compensated for by increased manufactured infrastructure or education.
The debate about strong versus weak sustainability, substitutability, and the drivers of growth continues and is often quite technical. But it has the merit of focusing attention on the production function—how resources are deployed to achieve a desired end. The production function is a very useful way of framing a move from economic growth as a presumed societal goal to a concern with well-being more broadly defined. An emerging literature, usually referred to as “structural human ecology” to differentiate it from earlier versions of human ecology, is building on the idea of production functions to examine how well-being is produced (Dietz and Jorgenson 2013). In the next section of this paper I summarize the basics of this literature before turning to the two major research questions that the move toward a focus on well-being involves.
FROM HOUSEHOLDS TO STRUCTURAL HUMAN ECOLOGY
Household Production Functions
While classical growth theory applied the idea of a production function to a nation or to a firm, starting in the 1960s the “new home economics” theory suggested that the production function idea could be applied to the household (Becker 1965, 1960; Berk and Berk 1983; Lancaster 1966). The initial applications of this approach attempted to explain nearly all behavior in terms of self-interested utility maximization (Becker 1976). These efforts were certainly provocative. They inspired increased attention to issues such as time allocation within the household and women's opportunity costs associated with child rearing in a gendered division of household labor. Of course, from the start sociologists were skeptical of some of the unrealistic simplifying assumptions required to deploy these models (Blake 1968).
Despite the overemphasis on rational choice assumptions in early versions of household production theory, the idea of thinking about how well-being is produced within a household has led to a nuanced literature examining how households deploy the resources available to them in order to generate livelihoods (Bebbington 1999; de Sherbinin et al. 2008; Serageldin and Steer 1994). This approach takes into account resource constraints faced by households and also households' agency in crafting strategies to generate well-being in the face of those constraints. It allows for consideration of inequality within the household and across households. It incorporates power differentials and the power embedded in institutions and cultures. Much of the literature explicitly acknowledges that production of material goods and services and income is only one appropriate outcome for household activity; Sen's idea of capabilities is central to this line of research (Sen 1999b, 1982). And the household production approach supports the logic undergirding the Millennium Ecosystem Assessment, which took as its central theme the dependence of human well-being on ecosystems (Alcalmo et al. 2003; Carpenter et al. 2009; Reid et al. 2005).
Using the household production perspective, we can understand the interplay between ecosystems and human well-being in specific local contexts. For example, one of the largest threats to the giant panda in China's Wolong Panda Reserve is the extensive use of firewood by the households who live in the reserve (Chen et al. 2012). The households would prefer to use electricity, but even with government subsidies it is relatively expensive. The opportunities for earning cash locally are very limited. However, members of some households are able to find work in distant cities and send money back to Wolong. This substantially decreases the use of firewood. A key element in this livelihood strategy is using “weak ties”—indirect social connections—to find work in the distant cities. Some households are able to use social network connections to facilitate the use of human resources to generate cash, which in turn displaces the problematic use of natural resources. The household production framework also was able to show the effects of the 2008 earthquake (whose epicenter was adjacent to the reserve) on household resource use and resilience (Yang, Dietz, Kramer, et al. 2013; Yang, Dietz, Liu, et al. 2013).
As the Wolong example demonstrates, analysis guided by the logic of a production function, broadly defined, has been very fruitful not only for looking at how households generate livelihoods and well-being but also for understanding what drives household use of natural resources and the resulting stress placed on the environment. Such work underpins much of our understanding of land use change around the globe (Entwistle and Stern 2005; Moran 2010; Moran and Ostrom 2005; Rudel 2009; Rudel et al. 2009, 2005; Rudel and Horowitz 1993). There is also a literature examining the drivers of household-level use of energy in the developed world (Dietz, Stern, and Weber 2013; Stern 2014). And household-level analysis is an important way of framing problems of vulnerability to environmental change (McLaughlin 2012; McLaughlin and Dietz 2008; Turner 2010; Turner et al. 2003).
However, there are some important limits to household-level analysis that come from the lack of investment in environmental social science research. Most household studies are local to regional in scope. They draw their data from households within one or a few communities, and usually a study is conducted in only one nation. While this has yielded a rich literature, the lack of variation in institutions, policies, and culture means that such studies cannot disentangle the effects of these key influences on both human well-being and stress on the environment. To address this issue we have to engage a more macro-comparative perspective (Bebbington 1999).
Toward a Structural Human Ecology
Eugene Rosa often argued that a key finding of sociology is that “context matters” (Dietz 2013b; York 2013). Despite the many contributions of work at the household and local level, we have not learned as much about the effects of institutions, culture, and other structural factors as we would like. The problem is not one of conceptualization. The ideal research design is well known (Dietz et al. 2010). We would collect data at the individual level nested successively within households, communities, local geopolitical units (cities, counties, provinces, etc.), and ecosystems, with all those units nested in turn within nations. With such data, multilevel modeling could be used to account for contextual effects where features of larger units, such as institutions, culture, ecosystem structure and function, and climate, can influence causal relationships at the micro level. In addition, social and biophysical connections that are local as well as those that operate at a distance could be incorporated into models (Liu et al. 2013). Data on the same units over time—panel data—would allow powerful assessment of causation. While the social sciences have the conceptual and methodological framework in place to do this work, funding has never been available for anything approaching the kind of data collection effort required. The World Fertility Survey (Cleland, Scott, and Whitelegge 1987) and the International Forestry Resources and Institutions project (www.ifriresearch.net) (Wollenberg et al. 2007) are excellent examples of what can be achieved with data that captures variation across local contexts, even when we lack ideal data sets.
In the absence of data for contextual and network analysis that match the conceptual ideal, how can progress be made? Certainly work at the household level using local and regional samples will continue to be important for our understanding of well-being. But in parallel, macro-comparative work, building on a rich tradition in environmental sociology and the sociology of development, can contribute a great deal to our understanding of the production of well-being and the factors that influence it.
World systems theory and dependency theory have inspired a rich empirical literature showing the effects of global power relations on economic growth, on inequality, on well-being, and on the environment (Babones and Chase-Dunn 2012; Bornschier, Chase-Dunn, and Rubinson 1978; Burns, Kentor, and Jorgenson 2003; Burns et al. 1994; Chase-Dunn 1989; Clausen and Longo 2012; Jorgenson 2003, 2004, 2006, 2009; Jorgenson and Burns 2004; Lough 1999; Roberts and Grimes 2002; Rosa and Dietz 2012; Shandra et al. 2004; Shandra, Shor, and London 2009; Snyder and Kick 1979; York, Rosa, and Dietz 2003). The most recent development in that literature is a series of analyses of what is being called the Ecological Intensity of Well-Being (EIWB) (Dietz, Rosa, and York 2009). The EIWB literature asks what structural factors influence the production of human well-being and anthropogenic stress on the environment. Initial studies used a production function of the sort articulated in discussions of strong versus weak sustainability (Dietz, Rosa, and York 2009; Vemuri and Costanza 2006). Human well-being was modeled as a function of national affluence, human resources, and natural resources. Following the approach of Mazur and Rosa (1974), more recent applications have compared measures of anthropogenic stress on the environment (usually the ecological footprint or greenhouse gas emissions) to a measure of human well-being (usually life expectancy or subjective well-being) (Dietz and Jorgenson 2014; Dietz et al. 2012; Jorgenson 2014; Jorgenson, Alekseyko, and Giedraitis 2014; Jorgenson and Dietz 2014; Knight and Rosa 2011; Steinberger et al. 2012; Steinberger and Roberts 2010). By making comparisons across macrounits, typically nation-states, these analyses have begun to address structural influences: the role of institutions and the distribution of power, culture, and other factors hard to examine with research confined to the local level in one or a few places. Together, the macro approach and the micro approach complement each other and will allow for robust and nuanced understandings of the production of well-being, human stress on the environment, and the trade-offs between the two. This structural human ecology provides a vehicle for sociological contributions to debates about conventional economic growth and sustainability (Dietz and Jorgenson 2013). As we develop alternative approaches to conventional growth theory, many new research questions arise. I will suggest two challenges that are especially worthy of our attention.
CHALLENGES FOR STRUCTURAL HUMAN ECOLOGY
What Is Well-Being?
Selecting measures of well-being to use in theoretical and empirical analysis is inherently a normative or ethical decision. One advantage of a research program on the production of well-being is that it will encourage or indeed require researchers to be explicit about the ethical theory that underpins their choice of well-being measures. While sociologists have often been critical of the use of income or wealth as a normative standard, at least economists can invoke utilitarianism and a substantial body of normative theory as the basis for that focus (Dietz 1994; US National Research Council 1999b). I believe that nearly all sociologists, and certainly nearly all environmental sociologists and sociologists of development, have strong ethical concerns that underpin their work. But there is relatively little explicit discussion of ethical theory in our literature.
Perhaps the most prevalent normative sociological theory traces its roots to Dewey and Habermas and their argument for a deliberative foundation for ethics (Dewey 1923; Dietz 1994; Dietz and Stern 1998; Habermas 1993, 1996). This approach, often called analytic deliberation, has been incorporated into environmental social science initially around discussions of impact and risk assessment, then in larger discussions of environmental politics, and most recently in discussions of how best to integrate science and democracy (Dietz 2013a; Rosa, McCright, and Renn 2013; US National Research Council 2008). It has also been influential in the “capabilities” approach of Nussbaum and Sen, and thus in discussions of development (Nussbaum 2006; Sen 1999b). As noted, the HDI is largely driven by the capabilities approach and thus is only one step removed from deliberative ethics logic.
The capabilities approach may also be helpful in addressing a major problem for current approaches to well-being—they are wholly anthropocentric (Donaldson and Kymlicka forthcoming). Nussbaum (2006) has offered a persuasive argument that the capabilities approach can be extended to give ethical consideration to the well-being of nonhuman species. Clearly, a robust normative approach must be attentive to both the role and the well-being of nonhuman species in society (Dietz and York 2014; York and Mancus 2013). Careful consideration of how to take account of nonhumans in the ethical arguments underpinning our work is an important step in that direction.
How might a research program on the production of human well-being be organized around a deliberative ethics? An implication of deliberative ethics is that a broad range of views, including those of the public (interested and affected parties in Dewey's  definition), must be engaged in drawing any ethical conclusions. Analytic deliberation has become a well-accepted approach when science is being deployed to inform decision making and assessments (US National Research Council 1996, 2008). But for more general research programs that are not tied to specific decisions or assessments, one solution has been to look toward public statements that can plausibly be taken to reflect a broad consensus. For example, Our Common Journey, a response by the scientific community to Our Common Future, uses major UN resolutions such as the Millennium Development Goals as guidance for what measures of well-being represent a broad consensus (Kates et al. 2001; US National Research Council 1999a).
This suggests a first criterion for any measure of well-being: there must be a broad normative consensus that the measure captures a feature of the world that is desirable. I would propose two further criteria. The measure must be driven by the condition of the bulk of the population and thus resistant to overweighting the situation of the most affluent and powerful: that is, the measure must be attentive to equity. Also, the methods used to assess the measure at levels from nation-states down to individuals should be well developed, robust, and feasible to deploy in a wide variety of circumstances.
I submit that there are at least two measures that meet these criteria. Generally, indicators of well-being are divided into objective measures and subjective measures. Among the objective measures, life expectancy at birth is the most commonly used. It has several advantages. There is a broad consensus within and across societies that lower mortality rates and higher life expectancies are desirable.9 For example, life expectancy is one of three elements of the HDI.10 Life expectancy is heavily influenced by infant and child mortality, which in turn are strongly correlated with maternal mortality, so life expectancy captures another set of measures that are often emphasized in such global statements as the Millennium Development Goals. It also captures access to sanitation, clean water, adequate and nutritious food, and freedom from indoor and outdoor air pollution and from other exposure to toxics. In addition, life expectancy is attentive to inequality. A society can achieve a high average level of affluence as measured by GDP per capita even with great inequality because a small number of super-rich can outweigh masses in poverty. But a few super-rich who can achieve exceptionally long life expectancies will not have any substantial influence on average life expectancy if most members of the society are impoverished and suffering from ill health and other risks. A few additional life years among the privileged cannot compensate for early deaths among the impoverished. So life expectancy has a built-in weighting against extreme inequality (one can emphasize this weighting by examining the life expectancy of disadvantaged groups when data are available). Finally, in most societies life expectancy is one of the easiest indicators of human well-being to measure with reasonable accuracy.11
Subjective measures of well-being ask representative samples of respondents to self-report how happy they are or how satisfied they are with their life overall or with aspects of their life (e.g., work, relationships, health). The methodology for conducting such surveys has evolved for several decades and is now quite sophisticated (Diener 2013; Diener, Inglehart, and Tay 2013; Diener et al. 2009; Kahneman and Krueger 2006; Kahneman et al. 1999; Krueger 2009; Organisation for Economic Co-operation and Development 2013; US National Research Council 2013). As a result, subjective well-being measures are being discussed as possible parallel metrics to the traditional national economic accounts and in some cases are even being implemented on a trial basis (Self 2014; Stiglitz et al. 2009; UK Department for Environment 2011). While there is less consensus for subjective well-being as a desirable social goal than there is for life expectancy, I believe there is enough agreement that this approach also is justified as a focus for further investigation. As noted, the methodology is well developed and can be applied wherever it is feasible to conduct a high-quality survey. And measures of subjective well-being are equitable in that a small subset of the population can have limited effect on the central tendency.
What Is Capital?
Readers familiar with the literature on strong versus weak sustainability or on household livelihoods will note that I have been using the term resources (i.e., manufactured resources, human resources, natural resources) while the literature usually refers to “capitals” (i.e. manufactured capital, human capital, and natural capital). I have shifted to the resource language because the term capital has multiple meanings that can be conflated too easily when everything is seen as a form of capital. I prefer to reserve the term capital for resources that are used with the expectation of a positive return on investment—a profit. Or as Bourdieu (2008:280) puts it, capital is the “potential capacity to produce profits and to reproduce itself in identical or expanded form.” The growth theory that underpins much of sustainability theory is centrally concerned with increasing economic activity per capita or with increasing output per worker. So it is understandable that development, growth, and sustainability theory see as capital what I call resources. But to term material infrastructure, education, and ecosystem goods and services “capital” presumes an intention to deploy them to generate exchange values. It also assumes a set of social relations, norms, and institutions that allow these resources to be deployed to generate a profit via market exchange. Of course, in most of the world today resources are fungible—can be exchanged for money—and can be invested to seek profit.12 Thus most resources in the contemporary world are capital.
However, I believe that the analytical distinction between manufactured resources (e.g., machinery, software, infrastructure), human resources (e.g., labor supported by education, skills), and natural resources (e.g., ecosystem goods and services) on the one hand and capital on the other is useful for conceptualizing past, present, and future human ecologies. All societies throughout the history of our species have deployed manufactured, human, and natural resources to generate well-being. But for most of human history the culture and institutions in place meant that profit as a result of exchange was not the typical goal in using those resources. It is undoubtedly true that in food-foraging societies better tools, better gathering techniques, and ecosystems more productive of food and fiber were valued. But it appears that the kind of exchange relationships that typify markets and that make various kinds of resources fungible were of modest importance for the majority of human history (Fukuyama 2011; Gowdy 1997).13 Further, even in contemporary societies, there is substantial contestation about what resources should and should not be exchanged to yield profit. Some forms of human resources are subject to fungible market exchange and thus are capital. But we place sharp limits on that exchange through laws, norms, and institutions to prevent, for example, slavery, child labor, sexual trafficking, and other abusive practices. We extend these limits to nonhuman resources as well. The first animal protection law was enacted in Ireland in 1635 to protect farm animals—horses and sheep in particular (Kalof 2007). It was the progenitor of numerous laws and norms that limit the exploitation of nonhuman animals as resources, capital, and labor (Dietz and York 2014). Protected areas such as national parks produce use value and perhaps some exchange value in the form of visitor expenditures. But while protected areas provide some services to humans, we have explicitly decided not to allow these ecosystems and their services to be fungible via exchange values in the market.
For all these reasons I have found that my thinking is much clearer when I differentiate between various forms of resources and the use of those resources as a form of capital. This distinction is also intended to respect and eventually engage the substantial work outside neoclassical economics on the nature of capital (e.g., Foster 1999, 2014; Foster, Clark, and York 2011; Harvey 2011; Mészáros and Foster 2010; Piketty 2014). Drawing a distinction between resources and capital avoids the presumption that the only value of a resource is the exchange value that comes from market transactions. The distinction also makes it easier to see that what is and is not used as capital is a result of agency and political struggle. Thus it allows visions of a future in which we can change what we consider to be appropriate to use as capital and what we do not.
The problems of equating resources and capital are exacerbated when we consider influences on well-being beyond the classical manufactured, human, and natural resources. Even the earliest analyses of the production of livelihoods by households discussed the idea of social capital, and political and cultural capital have also been invoked as possible components of a production function (Serageldin and Steer 1994). It is certainly true that factors beyond manufactured, human, and natural resources influence the production of well-being. But while the ideas of social, political, and cultural capital may be capturing some of those influences, I would suggest that conceptualizing them as capital is likely to be confusing.
Social capital is usually defined in terms of social network connections and trust. There is no doubt that networks and trust are important in generating human well-being. They can be key factors in gaining access to situations in which other resources can be either acquired or deployed (Chen et al. 2012; Coleman 2000; Dasgupta 1999;). But while network connections and trust certainly need to be considered in developing a better understanding of well-being, they are of a much different character than manufactured, human, or natural resources, especially in a capitalist market society. Typically network connections and trust are not privately “owned” and therefore cannot be exchanged directly for money or directly purchased with money. Granted, in most contemporary societies money can be used to carry out activities that can enhance an individual's social resources. As Bourdieu (2008:282) notes, manufactured capital (roughly his economic capital) is “immediately and directly convertible to money,” but his cultural capital (which is roughly human capital) and social capital are convertible “on certain conditions” and “in certain circumstances” respectively. In the contemporary world, manufactured, human, and natural resources are generally fungible, but social and other factors that influence the production of well-being are usually not. So households or individuals may be able to use trust and network connections as resources to enhance well-being. But to think of trust and network connections as capital—a resource invested to generate a profit in a market exchange—seems inappropriate.14
When we move from the individual or household to geopolitical units such as the nation, treating network connections or trust as resources seems even more problematic. First, it is critical to differentiate between social resources within a nation and the social resources of a nation in the world system. Certainly levels of trust and the network connections of individuals and organizations within a nation have a substantial influence on the ability of the nation to generate well-being for its citizens. For example, corruption, while certainly based on trust and network connections of a sort, is antithetical to the kind of social resources that might enhance the ability of a nation to create well-being for its citizens, and there is evidence that high levels of corruption facilitate environmental degradation (Mitra 2000; Welsch 2004). But these within-nation characteristics are different in kind from the relationships between nations.
A major insight of the sociology of development is the importance of a nation's position in the world system, which has been operationalized as a set of network relations at least since Snyder and Kick's important work (Chase-Dunn 1989; Chase-Dunn and Grimes 1995; Snyder and Kick 1979; Wallerstein 1974). Indeed, world polity theory suggests that these connections are dominant features in shaping policy and institutions (Buttel 2000; Frank 1997; Frank, Hironaka, and Schofer 2000; Shandra 2007; Shandra et al. 2009). International relations scholars also have argued that participation in the scientific epistemic community is critical for policy adoption (Cross 2013; Haas 1993). But here again we must be cautious about labeling such factors as resources, let alone as capital. Dependency and world systems theory and the substantial literature supporting them tell us that strong network ties can be disadvantageous for some nations because they represent relations of dominance and subordination. Under those conditions, it seems odd to refer to such strong network connections as a resource or a form of capital for the subordinate nation.
I would also suggest that power, and especially the ability to deploy violence, is too nuanced and important a concept to be treated analytically as simply political resources or political capital (Henry 2011; Mann 1986; Shwom 2011). Certainly, violence has been a major factor in the evolution of the world systems since its inception and has had profound implications for human well-being and the environment (Lough 1999; Mann 1986; Stonich and Vandergeest 2001). Wealth can be, and often is, deployed to finance violent actions. Militarization of nations seems to be detrimental to the environment and to human well-being (Clark, Jorgenson, and Kentor 2010; Jorgenson and Clark 2009; Kick, Nasser, Davis, and Bean 1990). But, as with network connections and trust, it may be more fruitful to conceptualize power and the mobilization of violence as factors that shape the ability to successfully deploy resources rather than as resources in themselves. Power can influence the return that comes from using resources by shaping culture and institutions to favor those with power over those without. In evolutionary terms, power reshapes the adaptive landscape (Dietz and Burns 1992; McLaughlin 2012; McLaughlin and Dietz 2008). Thus power seems too complex and dynamic a factor in societal evolution to conceptualize simply as a resource or as capital.
My goal in this essay was to outline the intellectual foundations of, and challenges for, a structural human ecology of well-being. This line of work is promising in that it builds on several strong research traditions: the strong-versus-weak-sustainability debate, household production, and especially the macro-comparative tradition in both environmental sociology and the sociology of development. It has the potential to advance our theoretical understanding even as it provides results that are useful for shaping political debate and policy. The concerns I raise—encouraging a more nuanced discussion of how we might conceptualize well-being, drawing a distinction between resources and forms of capital, and avoiding theorizing all influences on well-being as resources or capital—are not fundamental obstacles to progress. Indeed, engaging with them should hone our theoretical and methodological thinking.
Of course, this short description of an emerging field, and especially the treatment of its intellectual forebears, is inevitably superficial. I have certainly smoothed over important subtleties and distinctions, and my characterization of some research traditions may be unsatisfying to those working in those traditions. But I hope that this essay, despite its limits, will encourage more scholars from diverse traditions to engage with the emerging literature that is attempting to understand how structural forces, including institutions and policies, impinge on our ability to improve human well-being.
I thank J. Foster, R. Gunderson, A. Jorgenson, J. Shandra, R. York, and J. Zhao for very helpful suggestions on these issues; R. Kelly for able assistance; and especially the anonymous reviewer for Sociology of Development for insightful comments that have greatly improved the paper. This research was supported in part by Michigan AgBio Research.
I will use the term affluence in the conventional sense to mean the amount of economic activity in a nation and in particular economic activity per capita.
GDP is roughly the value of all goods and services purchased in an economy in a given year. GDP can be thought of as the size of an economy or the amount of economic activity taking place in a nation. The differences between GDP, gross national product, and national net national income are important for some purposes but need not concern us here. Hecht (2005) provides a useful overview of the standard measures of the national economic accounts and several important alternatives to them.
Even from the start of utility theory in 1738, Daniel Bernoulli, by specifying that marginal utility is a logarithmic function of income, acknowledged that there were diminishing returns to utility from increased consumption. But the basic argument linking growth to utility still holds.
The conflicts and complementarities between economic growth theory and sustainability theory are well articulated in papers and commentaries in Ecological Economics 22(3).
Arguments that quality of life at the end of life may decline are important for policy and decision making, but extending the life of older members of a society by a few years has little impact on overall life expectancy.
I prefer not to use the HDI itself as a measure of well-being because its other two components are both measures of resources that might be used to achieve well-being. Education is a human resource that almost certainly enhances well-being, but it is better thought of as an input rather than an end in itself. And of course the search for alternative measures of human well-being is motivated in large part by dissatisfaction with treating GDP per capita as well-being.
Adjustments to national economic accounts are another possible objective measure of well-being (Hecht 2005). In some cases the justification for a measure invokes normative arguments (Bagstad, Berik, and Gaddis 2014; Kubiszewski et al. 2013; New Economics Foundation 2009; Talberth, Cobb, and Slattery 2007). But the theoretical justifications for most of these measures are responses to the challenges of the strong-versus-weak-sustainability debate and growth theory in general. Many of them are trying to adjust traditional national accounts, such as GDP, to take account of degradation of the environment when used as a sink for pollution and of depletion of natural resources (Arrow et al. 2012 and commentaries; Dasgupta 2009; Hamilton and Hepburn 2014 and related papers in special issue; World Bank 2006). Surely these metrics need further consideration by the sociological community, but so far they have not been much discussed or deployed in empirical work in sociology.
As Ingham (2004:3) notes: “Money is one of our essential social technologies.” Marx, Weber, and Simmel made important contributions to the theory of money. Sociologists, and in particular sociologists of development and environmental sociologists, might benefit from a more nuanced examination of how culture and social institutions shape the relationships among money, resources, capital, and power.
Of course, some business models for Facebook, LinkedIn, and other social networking services would have the service “own” the information about network connections and attempt to profit from that knowledge. And it is certainly true that lobbyists, media producers, and other “deal makers” sell their network connections and the trust of others. So there may be situations where network connections and trust are acting as capital. But I would argue that it is best to be cautious and treat these as special cases rather than to label all trust and network connections as capital.