Structural Human Ecology: New Essays in Risk, Energy, and Sustainability

Preface

A PERSONAL THANKS TO GENE

Paul Ehrlich

IHAVE BEEN A FAN OF GENE ROSA since he became one of a team (along with Tom Dietz and Richard York) who were among the pioneers in the social sciences taking a serious look at the environmental dimensions of the human predicament. This was especially important for humanity, since it is clear that while environmental scientists have long been aware that humanity is on a course potentially fatal to civilization, we don’t understand enough about human behavior. In short, the ball is now in the court of the social sciences, and Gene was one of the first to pick it up. His work, I must admit, has pleased me personally because he was one of the first to pay attention to the I=PAT formulation that John Holdren, Anne Ehrlich, and I developed in the 1970s, and extend it with his colleagues into the STIRPAT analyses. Gene and his colleagues were among the first social scientists to realize that the old social science model—human societies acting on an essentially constant physical-chemical environmental stage—was totally flawed as the world entered the Anthropocene. They saw that the human socio-economic-political complex adaptive system (CAS) was interacting strongly with the biosphere CAS, with potentially catastrophic results induced in the biosphere system by the human system.

Gene Rosa

Gene’s group realized that social science and environmental science could no longer remain separate if a successful evidence-based system of inquiry was to be developed to deal with the growing human population-resource-environment-equity predicament. That led the group to press their STIRPAT work, actually measuring such things as the impacts of growing populations and expanding consumption on environmental variables, and refuting a variety of claims that those impacts will be handily reduced by economic factors and technological innovation. Gene was also involved in work to improve global risk governance, finding ways to bring science to bear on risk-related decision making, especially in connection with risky technologies and how society should deal with the risks related to future natural disasters. His careful analyses of the difference between the ontology and epistemology of risk, and of cultural differences in approach to risk, have been a major factor in clarifying thought in the area.

The importance of Gene’s concerns and approaches has been underlined recently by the mounting problem of climate disruption. The science of climate change has been studied intensively by international groups of scholars, and a broad consensus reached within the scientific community that the risks could include a dissolution of society as we know it—a quite possible-infinity problem—one where the chances of occurrence are far above zero, but if it does happen the results are utterly catastrophic (near infinitely bad). That consensus has been as widely broadcast by scientists as possible, but ignorance and distorted coverage by the media have prevented the magnitude of the problem from reaching most of the public and decision makers. That failure, linked with a determined disinformation campaign financed by, among others, the oil and coal industries, has, especially in the United States, resulted in essentially no preparation for likely coming disasters.

In the face of such developments, Gene became a central figure in the MAHB (the Millennium Alliance for Humanity and the Biosphere—mahb.stanford.edu). The MAHB has been dedicated to focusing the attention of the academic community (especially the social sciences) on solving the human predicament and to attempting to unite the efforts of the extremely diverse but uncoordinated elements of civil society trying to deal with various aspects of the colossal risks we face. Gene worked tirelessly in this cause, even while struggling with an extremely serious disease. This book is a small indication of the debt I, and his other colleagues, owe him. He was a great scientist and a great friend.

I. Theoretical and Conceptual Issues

CHAPTER 1

Introduction to Structural Human Ecology

Thomas Dietz, Michigan State University Andrew K. Jorgenson, University of Utah

THE TERM HUMAN ECOLOGY has seen myriad uses. But the most straightforward approach is simply to define human ecology as the study of the inter-relations between humans and their biophysical environment. In sociology, the term human ecology is most closely associated with earlier Chicago School scholars who emphasized the importance of spatial patterning in social structure and borrowed concepts like succession from the ecology current at the time the Chicago School emerged in the early twentieth century. Later, key thinkers in the Chicago School, notably Amos Hawley, distanced themselves from the ecological/evolutionary thinking that had inspired earlier scholars (e.g. Hawley 1986) although Otis Dudley Duncan long emphasized the importance of such links (Duncan 1964).

The emphasis in the Chicago School on the biophysical environment is something of an exception in sociology. The disjunction between main-stream sociology and concerns with human interaction with the biophysical world is well documented (Catton and Dunlap 1978, 1980; Dunlap 1980; Dunlap 1997). But independent of the Chicago School tradition, a number of scholars within sociology and in the larger social and ecological science community have continued to engage with the image of the cultural evolutionary play in the human ecological theater—a metaphor inspired by G. Evelyn Hutchinson’s classic essays (Hutchinson 1965). Much of environmental sociology is concerned with what might be called human ecology and a tradition of evolutionary theory has continued (Dietz, Burns, and Buttel 1990; Lenski 2005; Burns and Dietz 1992; McLaughlin 2001). Meanwhile, work on human/environment relations have flourished in a variety of other disciplines (overviews include Moran 2006; Rosa, Diekmann, et al. 2010; Moran 2010b; Stern 1993; Richerson 1977; U.S. National Research Council 1992).

Recently, a confluence of several streams of environmental research in sociology and kindred disciplines has yielded what might be thought of as, if not a school at least a community of researchers with overlapping interests and approaches. Eugene Rosa was the first to refer to this approach as structural human ecology (Rosa 2004b; Knight 2009). While not all members of this emerging community share the same intellectual interests, three themes serve as links between many of them. The first is an interest in meta-theory—in careful thought about ontology, epistemology, normative and positive aspects of theory, and in developing coherent overarching concepts to guide research. A second theme is attention to issues of risk and uncertainty. Risk analysis became a dominant mode of environmental and technological policy analysis starting in the 1980s, and has flourished both as a practical tool and as a topic for social science investigation (Dietz, Frey, and Rosa 2002; Rosa, McCormick, and Frey 2007; Rosa, McCright, and Renn in press). The structural human ecology community is informed by research on how humans handle uncertainty and has contributed to social science analysis of risk since that literature first emerged in the 1980s. The third theme is a flourishing body of quantitative macro-comparative research on how humans place stress on the environment but also on the factors, including the environment, that influence human well-being.

It is important to note that there is no dominant paradigm or overarching theoretical program that constitutes structural human ecology, save perhaps three points frequently made by Gene Rosa: 1–Context matters; 2–Theory must be disciplined with data; 3–Progress requires careful thought about concepts and basic premises. Some scholars work on all three themes: meta-theory, risk analysis, and quantitative analysis of drivers. Some work in only one or two areas. But overall a strong network of collaboration and co-citation has produced an intellectual community that moves rapidly by sharing ideas and by respectful and supportive debate.

Eugene Rosa was a central figure and leader in this network. He contributed germinal research in all three areas and pioneered the ongoing work of building the connections between them. So when Gene was awarded the Boeing Professorship of Environmental Sociology at Washington State University in 2011, it seemed an ideal time to bring together structural human ecology scholars to reflect on the state of the science and where we might move in the future. Because of his contributions, the Symposium on Structural Human Ecology also would naturally serve as an engagement with Gene’s work at the start of his Boeing Professorship.

The Symposium was held on 24 September 2011 at Washington State University with the generous support of the Thomas S. Foley Institute for Public Policy and Public Service and the Department of Sociology, Gene’s academic homes for decades. Ten scholarly presentations and concluding remarks by Gene made for a very exciting intellectual engagement. The participants were eager to see the papers collected into a volume that would assess the state of the emerging field of Structural Human Ecology.¹ This book is the result.

The opening section contains papers by Richard York and Thomas Dietz that are reflections of Gene’s contributions to meta-theory, and in particular the issues of ontology and epistemology that he raised in his germinal paper Metatheoretical Foundations for Post-Normal Risk (Rosa 1998a) and in related work. Meta as York and Dietz refer to this body of work, is about the appropriate ways to think about ontology and epistemology in the context of risk decisions. Gene, drawing on ideas of Funtowicz and Ravitz (1993), identified a two-dimensional space defined by the degree to which our knowledge is based on phenomena that are ostensible and by the degree to which observations are repeatable. From this he argues for a HEROic approach to using science—Hierarchical Epistemology applied with a Realist Ontology.² In the areas of science most important for environmental decision making, ostensibility and/or repeatability may be limited. Unlike many sociologists of science, Gene promoted a realist ontology—there is a reality we are trying to observe rather than simply a set of social constructions. Interestingly, he also argued that most of the public engaged in risk disputes hold a realist ontology (Rosa 1998b). But because of problems of ostensibility and repeatability we must avoid hubris about our understanding of reality. In some cases it is reasonable to believe our understanding is a close match to reality, in other cases, including for many things important for structural human ecology, we have to acknowledge that our understanding involves substantial uncertainty. Thus he called for a hierarchical epistemology, open to questions for which our understanding can be asserted with great confidence and to circumstances where we must acknowledge great uncertainty and the potential for social processes to shape substantially scientific understanding.

Richard York, in his chapter, expands on the idea of HERO by noting that much of science, and nearly all of science used for policy, is about predicting the future. That means that uncertainty is inherent. He argues that there are multiple sources of uncertainty and several ways to interpret statistical analyses. In work on relatively simple systems, such as those of classical physics, uncertainty is handled by statistical procedures based on an idea of measurement error that stretches back at least to Galileo (Bennett 1998; Salsburg 2001; Bernstein 1998; Stigler 1986). This approach to uncertainty makes sense for many kinds of risk analysis deployed in engineering. In contrast, the standard model of contemporary physics that incorporates quantum mechanics allows for uncertainty all the way down. Richard offers a key insight for structural human ecology: when we are dealing with complex ecosystems, social systems, or coupled human and natural systems (Rosa and Dietz 2010; Liu et al. 2007), we have not only measurement error but also specification error—our models are almost certainly wrong.³ This has implications not only for risk analysis, the subject of Section II of this volume, but also for how we interpret the statistical results in macrocomparative analyses that are the theme of Section III. As Richard notes, realizing the several sources of error in our models and thus the limits of our ability to make predictions does not lead us to a post-modern rejection of the scientific enterprise. Rather it encourages a more nuanced understanding of what we know and of how to work with the limits of our knowledge—an enterprise where Gene’s work stands as an exemplar.

Tom Dietz’s arguments follow a parallel logic to Richard’s analysis. Logically, all decisions have to take into account both facts and values. Gene’s work has given us clarity about the problem of getting the facts right in circumstances when there is scientific uncertainty, and Richard has shown us how broadly Gene’s logic applies to science. But in many environmental and sustainability decisions, there can also be value uncertainty (Dietz 2003, 2013). When confronted with a decision about the environment, about emerging technology, about any aspect of sustainability, the implications for things we value may not be obvious to most individuals. And it may not be obvious how we should aggregate individual value considerations into a societal decision. Tom argues that Gene’s insights apply to this problem as well as to factual uncertainty. The problem of ostensibility and repeatability apply not only to our understanding of the facts but also to our ability to assess the value implications of courses of action. Many of the issues addressed by structural human ecology, and many decisions about risk, require us to develop positions about novel phenomena. That the value implications of these problems are not obvious can be thought of as another example of low ostensibility. When the phenomena are novel it is hard to draw on past experience to make value judgments and that can be thought of as a problem of low repeatability. Tom considers the implications of Gene’s work for linking scientific analysis to public deliberation, an approach Gene long advocated.

Section II of the book addresses problems of risk directly, with each paper focusing on risk associated with technologies of emerging importance as a testbed. This approach follows Gene’s long history of honing his ideas around the decades-long controversies about nuclear power. Ortwin Renn, Nadine Bratchatzek, Sylvia Hiller, and Dirk Scheer take on a major technological challenge of the twenty-first century: climate engineering. Roger Kasperson considers the broad array of technological choices we must make. Paul Stern examines a third technology that may have huge impacts on climate change: hydrofracking. Like Gene’s work on nuclear power, each of these three chapters makes a contribution to our understanding of an important set of technological risks. And like Gene’s work, they use these examples to probe deeply into how best to think about and develop reasoned societal approaches for dealing with risk.

Climate engineering, sometimes called geo-engineering, is an emerging set of methods for limiting the magnitude of climate change (U.S. National Research Council 2010, Chapter 15). One major approach is to pull greenhouse gases, especially carbon dioxide, from the atmosphere. The other is to reduce the amount of solar radiation warming the earth’s surface. There are a number of specific technologies proposed within each of these two categories. As emissions of greenhouse gases have continued to increase, interest in climate engineering as a mechanism for reducing harm from climate change has grown. From the earliest discussions of geo-engineering, the importance of social science analysis has been clear. Most major reports on the technology note the importance of effective governance mechanisms and of anticipating public acceptance or rejection of it (Royal Society 2009; U.S. National Research Council 2010, Chapter 15). While the research on public acceptance is still in its early stages, Ortwin and his collaborators provide a commanding overview of the work that has been done to assess public views of technologies that have not yet received much attention in the media. An important feature of their review is that it examines work on multiple countries—we know from the history of nuclear power and genetically modified organisms that there can be substantial differences across nations in how technologies are perceived (Rosa, Matsuda, and Kleinhesselink 2000; Kleinhesselink and Rosa 1991). As Gene often noted in his work comparing risk perceptions across cultures, context matters. Ortwin and colleagues also report results from an expert workshop intended to identify appropriate analogies in recent technological controversies to various aspects of climate engineering, a comparative approach Gene favored (e.g. Rosa and Clark 1999). Nearly all the existing literature on climate engineering calls for early public deliberation linked to ongoing scientific analysis—a theme that crosses many chapters in this volume. Given the initial social science work on climate engineering, Ortwin and his collaborators offer clear suggestions about how to implement that dialogue.

Roger Kasperson shows the utility of distinguishing three types of uncertainty: aleatory uncertainty (situations where further data and analysis could, at least in principle, reduce uncertainty); model-parameter uncertainty (where more data alone won’t help but more research to build a better understanding and better models and parameter estimates might); and deep uncertainty (where not enough is known to use data to model). It is this last category that is most troubling for decision making and often encourages delays and continuing business as usual. The parallels across types of technology is something Gene also addressed (Rosa and Clark 1999). Roger suggests two strategies for this situation: adaptive management and resilience building. Adaptive management starts with best available knowledge but makes provisions for corrections as we learn and as the situation changes. It is based on social learning (Henry 2009). Resilience building takes steps so that society is better able to cope with whatever stresses the future holds, even though we cannot predict them with much accuracy. Roger then identifies a series of institutional properties that facilitate adaptive management and resilience. However, if deploying adaptive management, building resilience, and coping with deep uncertainty were easy, we would probably have done it already. Roger notes six dilemmas, each worthy of detailed consideration, that challenge our abilities to build the institutions we need. He posits that while the challenges are substantial, there are ways to move forward, and much to be learned.

Paul Stern examines the factors that influence public acceptance or rejection of new technologies, drawing on an extensive review of the literature by a National Research Council committee on which Gene served. Paul summarizes work on nuclear power and radioactive waste management. These are topics on which Gene wrote extensively for more than three decades (Rosa and Freudenburg 1984; Rosa 1988; Rosa and Dunlap 1994; Rosa and Clark 1999; Rosa 2004a, 2007; Freudenburg and Rosa 1984; Dunlap, Kraft, and Rosa 1993), including an engagement with the recent Presidential Blue Ribbon Commission on America’s Nuclear Future (Rosa, Tuler, et al. 2010). Paul also examines the history of public concern with DNA manipulation and a variety of other public controversies over risk assessment and management and environmental assessment and decision making. In addition to these extended case studies, some insights about technological controversies can be derived inductively from social science research that is not directed at technological controversies per se. He summarizes what has been learned from work on science communication and the use of science, international policy networks and governance of common-pool resources—all extensive and robust literatures. Moving beyond this summary of what we know, Paul develops a set of design principles for governing emerging technologies. He examines the way Elinor Ostrom (Ostrom 1990, 2010a, 2010b; Dietz, Ostrom, and Stern 2003) synthesized principles for the management of common pool resources, acknowledging the differences between the local to regional institutions most extensively examined in the commons literature (Ostrom et al. 2002) and some emerging technologies. From these lessons, and drawing on the insights from the literature on technological controversies and other social science literature, he proposes seven design principles for managing emerging technologies.

The third section of the volume focuses on macro-comparative analyses of the stress humans place on the environment, human well-being, and the tradeoffs between them. It is to describe this body of research that the term Structural Human Ecology was coined. In the 1990s Gene, Thomas Dietz, and Richard York (who was Gene’s PhD student) developed a line of analysis that is given the acronym STIRPAT. Scholars since Malthus, and even earlier, had offered arguments about the effects of population, affluence, choice of technology, and other factors on the environment (Dietz and Rosa 1994). In a series of papers centered around the Presidential Commission on Population Growth and the American Future (1972)—also known as the Rockefeller Commission—Barry Commoner and colleagues debated the relative importance of drivers of environmental change with Paul Ehrlich and John Holdren (Ehrlich and Holdren 1972; Commoner 1972). From that debate emerged the IPAT equation: Impact = Population x Affluence x Technology (Dietz and Rosa 1994; Chertow 2001) which is also known as the Kaya identity (Kaya 1990a; Kaya and Yokobori 1997; Kaya 1990b). STIRPAT acknowledges that as an accounting equation IPAT cannot be used to test hypotheses. But it can be converted into a stochastic form and then becomes a vehicle for empirical research. Gene coined the term STIRPAT based both on the acronym STochastic Impacts by Regression on Population, Affluence and Technology and on the stirp (descendent) of IPAT. This formulation has helped a community of sociologists, economists, geographers, political scientists, and others engage in macro-comparative analysis of human stress on the environment. STIRPAT forms a bridge between the social and the physical and biological sciences—where IPAT originated and remains popular. But it also connects with the venerable tradition of quantitative macro-comparative work in the social sciences that has long been used to explore issues of development, economic growth, inequality and human well-being. The STIRPAT research program and Gene’s formative influence on it are described in more detail in Chapters 8 and 10.

It is hardly surprising that Gene would be an innovator in macro-comparative analysis. One of his earliest publications was a pioneering examination of the relationship between human standards of living and energy consumption using cross-national data (Mazur and Rosa 1974; see also Rosa 1997). His dissertation advisor, Allan Mazur, has continued this line of work. The first part of his chapter in this collection updates our understanding of the links between energy consumption and quality of life, finding that the general conclusion of the original paper—that energy consumption decouples from quality of life at high levels of consumption—continues to stand. He then considers what drives energy consumption, focusing in particular on the variable that has been subject to debate at least since Malthus—population. He finds that for total energy supply, while population has a modest effect compared to other factors, growth in population always leads to increased energy use while other drivers vary in their impact over time. But population growth is largely decoupled from electricity use. This distinction is especially important because, as he notes, most future growth in energy use is likely to be in the form of electricity, including the demands associated with increased use of electric vehicles.

Andrew Jorgenson begins his contribution with a comprehensive review of the STIRPAT approach, including both its logic and the literature that has applied it to the political economy of environmental change. He notes in particular the methodological progress in this line of work and especially the move towards considering the stability of the effects of population and the environment over time and across space, an approach he and collaborators initiated. Andrew then brings the concern with temporal stability and regional differences together in a new empirical analysis. He examines the effects of population, affluence (gross domestic product per capita), urbanization (often proposed as a driver of environmental stress), and international trade, looking in particular at how the effects of population and affluence may have changed over time and across regions. Population and affluence remain key drivers of stress on the environment in all analyses. However, there are important regional differences in the effects of both population and affluence over time. In the mostly developed nations of Europe, North America, and Oceania, the effects of population are stable over the time period studied (1960–2005). But in other regions the effects of population or affluence or both change over time. Again, context matters.

Sandra Marquart-Pyatt addresses another key issue for structural human ecology—the need to work across