The Meaning of Sustainability:
Biogeophysical Aspects

by John P. Holdren, Gretchen C. Daily, and Paul R. Ehrlich

Edited by Mohan Munasinghe and Walter Shearer -- Copyright 1995
Distributed for the United Nations University by The World Bank Washington, D.C.

This paper benefited greatly from interactions with R. Cicerone, A. Coale, T. Dietz, P. Gleick, R. Heal R. Lenski, M. McDonnell,J. Lubchenco, T. Malone, B. McCay, N. Myers, D. Pimentel, G. Rabb, D. Skole, and M. Soule (U.S. National Academy of Sciences Planning Group for a study on ecological effects of human activities); Partha Dasgupta (Cambridge University); A. Ehrlich (Department of Biological Sciences, Stanford University); W. Falcon, L. Goulder, and R. Naylor (Institute for International Studies, Stanford University); R. Howarth, A. Kinzig, S. Lele, and R. Norguard (Energy and Resources Group, University of California at Berkeley); G. Woodwell, R. Houghton, R. Ramakrishna, J. Amthor, and E. Davidson (Woods Hole Research Center); and M. Weitzman (Department of Economics, Harvard University). The responsibility for errors and infelicities, however, rests solely with the authors. Our work on this topic was supported in part by grants from the Winslow and Heinz Foundations. J. Holdren also gratefully acknowledges the hospitality of the Woods Hole Research Center during a 1992 sabbatical in which much of his part of this work was done.

A sustainable process or condition is one that can be maintained indefinitely without progressive diminution of valued qualities inside or outside the system in which the process operates or the condition prevails. (We exclude from consideration, in applying this definition, the depletion of available energy from the sun on a time scale of several billion years!) [1] Such a definition may be logically appealing, but it is hardly sufficient for addressing the meaning of sustainability in the context of practical choices about how to maintain or improve the well-being of humans on this planet. [2] What kinds of processes and conditions need to be sustained in the interest of maintaining or improving well-being? What are the sources and dimensions of the main threats to the sustainability of these? What places should be investigated and what should be measured to find out? Can sustainability be made compatible with -- or traded off against -- other desiderata relating to policy choices? (Consider, for example, sustainable development versus rapid development).

The proposition that particular human practices would prove unsustainable has cropped up in literature going all the way back to the ancient Greeks and somewhat more frequently and sweepingly in the two hundred years since the work of Malthus, above all in the period since World War II. [3] Only in the past five years, however, has sustainability become a catchword capable of capturing the attention not only of environmental scientists and activists but also of (some) mainstream economists, other social scientists, and policymakers.

This enhanced salience presumably resulted from a suite of coincident factors. For one, the world community is no longer transfixed by the Cold War. A second factor is the reluctant appreciation of the severity of the debt crisis in the developing world. A third is the substantial advancement in scientific understanding of the magnitude and consequences of ongoing global environmental transformations, including the depletion of stratospheric ozone, the buildup of greenhouse gases, and the destruction of biodiversity. Also very important has been the attention given to the notion of sustainable development in the report of the World Commission on Environment and Development (WCED 1987, also known as "the Brundtland report") and the avalanche of related studies that has followed.

Notwithstanding the extraordinary growth of the "sustainability" literature in the past few years (an unsustainable process, to be sure!), much of the analysis and discussion of this topic remains mired in terminological and conceptual ambiguities, as well as in disagreements about facts and practical implications. [4] These problems arise in part because the sustainability of the human enterprise in the broadest sense depends on technological, economic, political, and cultural factors as well as on environmental ones and in part because practitioners in the different relevant fields see different parts of the picture, typically think in terms of different time scales, and often use the same words to mean different things.

It is therefore appropriate, even though this introductory chapter and the conference of which it was originally a part are supposed to focus on the biogeophysical aspects of sustainability, to begin by locating the biogeophysical aspects within the context of the wider debate about what sustainability means and implies. We then address, in turn, some problems with defining biogeophysical sustainability in practical terms, the connection between biogeophysical sustainability and related concepts such as carrying capacity and the distinction between renewable and nonrenewable resources, the state of knowledge and debate about the character and origins of threats to biogeophysical sustainability, and some implications of the current state of knowledge and ignorance of these matters. We undertake all of this with a pronounced emphasis on the global level of analysis, leaving to the chapters that follow the task of addressing the character and measure of sustainability in particular regions and ecosystems.

Biogeophysical sustainability in context

Much of the current salience of concepts of sustainability has come from a wide-ranging international discussion about sustainable development, which has been defined variously as, for example:

These definitions have the appeal of appearing to reconcile the concerns of diverse constituencies -- above all the development and environmental communities (Lele 1991) -- but they raise at least as many questions as they answer. Is it possible to meet the needs of the present without compromising the capacity of future generations to meet their needs? How does one define needs anyway? What determines carrying capacity, and how does it vary from place to place and over time? What is the relation between economic growth and development? What constitutes fairness? Let us sketch out tentative answers to some of these broad questions -- since those answers will partly shape our understanding of the environmental issues we want to address shortly in more detail -- starting with the meaning of development. We think development ought to be understood to mean progress toward alleviating the main ills that undermine human well-being. These ills are outlined in table 1-1 in terms of perverse conditions, driving forces, and underlying human frailties. (The problems at each of these levels are themselves diversely and often tightly interconnected.) The development process is then seen to entail improving the perverse circumstances by altering the driving forces, which in turn requires overcoming, to some extent, the underlying frailties. Sustainable development then means accomplishing this in ways that do not compromise the capacity to maintain the improved conditions indefinitely.

Table 1-1: Ills That Development Must Address

Condition Meaning
Perverse conditions
1.1 billion -- 20 percent -- of the 5.5 billion people on the planet live in absolute poverty and perhaps 2 billion people do not receive a sufficiently nutritious diet to alleviate disease
Impoverishment of environment Disruption and erosion of environmental conditions and processes on which the well-being of those 5.5 billion people depend even more directly than on economic conditions and processes
Possibility of war Civil, international, global, nuclear, or conventional wars manifest in the more than 100 instances of organized armed conflict since World War II, nearly all of them in the south, with a total loss of life in the tens of millions
Oppression of human rights In forms beyond the three already listed, which deny human beings their dignity, liberty, personal security, and possibilities for shaping their own destinies
Wastage of human potential Resulting from all of the foregoing and the despair and apathy that accompany them and from the loss of cultural diversity (Ehrlich 1980)
Driving forces
Excessive population growth
Driving forces Excessive population growth Where excessive means growth that closes more options than it opens (Holdren 1973), a condition now prevailing almost everywhere
Maldistribution of consumption and investment Where the maldistribution is of three kinds: between rich and investment poor as the beneficiaries of both consumption and investment, between military and civilian forms of consumption and investment, and between the two activities themselves, that is, between too much consumption and too little investment
Misuse of technology Which occurs in forms both intentional (as in weapons of mass destruction) and inadvertent (as in the side effects of a broad spectrum of herbicides and pesticides)
Corruption and mismanagement Which are pervasive in industrial and developing countries
Powerlessness of the victims Who lack the knowledge and the resources but above all the political power change the conditions that afflict them
Underlying human frailties
Greed, selfishness, intolerance, and shortsightedness
Which collectively have been elevated by conservative political doctrine and practice (above all in the United States in 1980 92) to the status of a credo
Ignorance, stupidity, apathy, and denial The first consisting of lack of exposure to information the second of lack of capacity to absorb it, and the third and fourth of having the information but lacking the conviction or optimism or fortitude to act on it

Development by this definition should by no means be considered synonymous with economic growth, since growth by itself does not assure progress toward alleviating any of the indicated ills. (Economic growth may be a necessary condition for alleviating some of them, but it is certainly not a sufficient condition.) Note also that we have placed sustainable in front of development to mean not that the development is of a form that can be continued indefinitely but rather that the choice of processes and end states for development are compatible with maintaining the improved conditions indefinitely. Under this sort of interpretation, even the much-maligned term sustainable growth need not be an oxymoron; it can be taken simply to mean growth in forms -- and to end points -- compatible with sustainability of the improved conditions it helps bring about.

If improvements in the human condition are to be not only achieved but also sustained, all of the ills will need to be addressed; this is so because failure to address any one of them can eventually undermine the progress made on all the others.

As the human enterprise expands, interdependencies mediated through the world economy, the global environmental commons, and international political and military relations link and intensify the threats posed by each of these ills. Thus the requires meets for sustainability include not only the environmental factors to which we will shortly turn in detail but also military, political, and economic ones. The minimum requirements in each of these categories are presented in table 1-2.

Table 1-2. Requirements for Sustainable Improvements in Well-being

Area and requirement Rationale
No weapons of mass destruction
No one can be secure as long as these exist anywhere, and as long as any country insists on retaining them, others will have an incentive to acquire them
Limited capabilities of national military forces Security would be served by attaining a condition in which no nation's military forces were strong enough to threaten the existence of other states; this can be facilitated by "defense dominance," in which national forces are structured to be much stronger in defense than in offense. If stronger peacekeeping forces are needed, they should be placed under international control
Smaller political units should coalesce into or be absorbed by larger ones only by mutual consent, based on mutual advantage
Participation/empowerment Societies are not stable and hence not sustainable -- unless their citizens have an effective voice in decisions that affect their lives
The rule of law The rule of the strongest, the most devious, or the most unscrupulous is a prescription for perverse and destabilizing forms of competition
Guarantees for human rights Majority rule does not include the privilege of abusing minorities; sustainability requires respect for cultural diversity as well as biotic diversity
Reduced disparities within and between countries
The large gaps between rich and poor that characterize income distribution within and between countries today are incompatible with social stability and with cooperative approaches to achieving environmental sustainability
Internalization of environmental costs Economic markets will lead to overconsumption of environmental resources and ultimately to unsustainability if these resources are not priced or are underpriced
Assignment of property rights to future generations This approach seems essential to avoid the outcome in which high discount rates of economic actors allow actions that undermine long-term sustainability to appear economically attractive (Howarth and Norgaard 1990)
Preservation of the environmental basis of present and future well being
What this requirement consists of and the way it might be attained are the topics of the rest of this chapter

With that wider array of considerations as context, we now take a closer look at the environmental dimensions of sustainability that are the main focus of this volume.

Definitions of environmental sustainability

The environmental aspect of sustainability has been the subject of a rich literature, albeit only recently with the term sustainability appearing explicitly. [5] As with the concept of sustainable development, however, the definitions of environmental sustainability to be found in the literature recent enough to use that term are often circular or unsatisfying in other ways. Consider the following capsule definitions:

The first statement is essentially a dictionary definition of sustainability; it tells us only what we already knew sustainability to mean. The second statement introduces the interesting term "ecological debt" to describe an element of unsustainability, but the elaboration in terms of overexploiting carrying capacity and productive capacity is not much help, insofar as it merely transfers the definitional burden to over-exploitation and carrying capacity. The third statement offers an actual specification of at least one element of sustainability, but there is still buried within it a definitional problem: How is "total natural stock" to be defined and measured? Assuming this hurdle can be overcome, the further question will surely arise: What is inviolable about the current level? Can environmental scientists give a good answer? We shall return to this issue below.

Of course, all serious writers on environmental sustainability go beyond the sorts of capsule definitions cited above and elaborate what sustainability might entail and require (see, for example, boxes 1-1 and 1-2). The 1980 World Conservation Strategy of the International Union for the Conservation of Nature, the United Nations Environment Program, and the World Wildlife Fund (IUCN 1980) concludes, for example, that sustainability requires "maintenance of essential ecological processes and life-support systems; preservation of genetic diversity; and sustainable utilization of species and resources." This three-part prescription seems to consist of different facets of the same thing: preservation of genetic diversity and sustainable use are essential to maintain essential ecological processes and life support systems.

Box 1-1: Definition and Measurement of Sustainability: The Biophysical Foundations -- Keiichiro Fuwa

Environmental issues have become so popular that politicians around the world no longer need to be persuaded of their importance. Natural scientists have been using the word sustainability for a fairly long time, and recently social scientists as well as politicians have started to use it quite frequently. However, it has yet to be defined clearly.

Recommendations have been made for the definition of measurements and indicators of sustainability. Although by no means final, the following working definition of biophysical sustainability is satisfactory for the time being: Biophysical sustainability means maintaining or improving the integrity of the life support system of Earth. Sustaining the biosphere with adequate provisions for maximizing future options includes enabling current and future generations to achieve economic and social improvement within a framework of cultural diversity while maintaining (a) biological diversity and (b) the

biogeochemical integrity of the biosphere by means of conservation and proper use of air, water, and land resources. Achieving these goals requires planning and action at local, regional, and global levels and specifying short- and long-term objectives that allow for the transition to sustainability.

Biophysical refers not only to biology and physics but also to geology and chemistry. This is expressed in the definition, particularly through mention of biogeochemical integrity. Natural science has become so interdisciplinary that it is often confusing; nevertheless physics, chemistry, geology, and biology remain the most basic disciplines. Biogeophysicochemistry expresses them all in one word, albeit an exceptionally long one.

Defining terms such as sustainability and sustainable development with reference to the global environment is, to my mind, complicated by the fact that humanity has been considered special and separate from other animals and plants. This has not always been the case. The Earth is divided into three spheres: atmosphere, hydrosphere, and lithosphere. The biosphere was added later as the fourth sphere but, unlike the others, includes those parts of the atmosphere, hydrosphere, and lithosphere in which life exists. Plants and animals are, of course, part of the biosphere, but more importantly, humans are included as just one species of animal and are not treated specially. In recent years, particularly when serious environmental problems were recognized, human activity was so intense and pervasive that it came to be considered -- for example, by the Man and the Biosphere Programme as separate from the activity of other forms of life.

Biophysical sustainability must, therefore, mean the sustainability of the biosphere minus humanity. Humanity's role has to be considered separately as economic or social sustainability. Likewise, sustainable development should mean both sustainability of the biophysical medium or environment and sustainability of human development, with the latter sustaining the former.

Box 1-2. Coming to Grips with the Biogeophysical Issues in a Social Construct, or How to Talk about Sustainability without Being a Social Scientist -- Sharad Lele

"You cannot talk about sustainability without talking about people, about politics, about power and control."

Comment by a sociologist at a seminar on sustainability University of California, Berkeley, 1988

"Sustainability is maintaining the ecological basis of economic well-being, so any discussion of sustainability must incorporate economic considerations."

World Bank economist

Comments such as these threaten to create a gridlock in our discussions of the biophysical foundations of environmental sustainability. But we are clearly not (and probably nobody is) capable of conducting such an all-encompassing discussion. How then do we discuss the biophysical foundations of environmental sustainability, however defined? Social, political, and cultural issues come into play in a number of ways at two critical stages in any discussion of environmental sustainability.

Stage 1. In deciding,

  • What is to be sustained? That is, what relative ranking is to be given to, say, current resource productivity, productive potential, or genetic diversity?
  • What attributes, or combinations of attributes, of a particular system are to be maintained nondecreasing: average productivity, stability, resilience, or adaptability?
  • Over what time scale is this sustenance desired?
  • Who is to benefit? If a tradeoff is necessary between current and future consumption and well-being, or between the well-being of one community and that of another, who is to decide and how?
  • Should it be economic value of any resource flow or stock that is maintained non-decreasing, or should it be the physical quantity of that flow?

Stage 2. In understanding,

  • Why is there environmental unsustainability, however defined, in the world today?
  • How would one achieve or move toward whatever notion of an environmentally sustainable society that is decided on in stage 1?

Stage 1 requires an explication of differing individual and cultural values, preferences, as well as beliefs about and approaches to a highly uncertain and unknowable future and then the resolution of such differences through some social process. Stage 2 requires an understanding of the complex array of social, political, and cultural factors in today's world that lead to environmentally unsustainable behavior.

Once this is clearly realized, it is easier to understand where our contributions as biophysicists and ecologists can be and ought to be in informing the process of reaching some societal consensus on the issues in stage 1. At the same time, we realize that, in our work, we have often made implicit decisions about the issues raised in Stage 1. We should therefore proceed as follows:

1. Clearly state the assumptions we are making about reality in a particular case, examine whether some assumptions are commonly shared, and determine the extent to which these may be justified. For instance, perhaps most ecologists believe that whatever is to be maintained nondecreasing in an ecosystem should be measured in physical, not economic, terms. This follows from their rejection of the belief commonly held and vigorously promoted by most economists: that technological change can continuously compensate for reduction in physical resource flows, thus preventing utility from de c

2. Clearly state what value-based choices of objectives, of their ranking, of time horizons, and of users are being implicitly made in any particular case.

3. Identify a few scenarios corresponding to choices different from those that we might want to make.

Having done this, we can then proceed with our basic tasks:

4. Synthesize the current state of knowledge about the relationships between biophysical processes that affect different objectives at different temporal and spatial scales. That is, what intensity of harvesting under what technique of logging can be maintained in a tropical forest at a nondecreasing level for what time period? What would the implications of a nondecreasing resilience requirement be?

5. Identify a sparse set of indicators that best relate to each combination of objective, scale, and so forth and possibly identify threshold values for them. For instance, what would be the best indicator of stable harvests in the above-mentioned forest? What would be the indicator of resilience in the same system? What scales (spatial and temporal) may be most appropriate or sensible for measuring what attribute or type of sustainability?

6. Explore the ways in which the different scenarios interact; that is, the synergisms and contradictions among objectives, attributes, and indicators and between sustainability in general and other societal objectives. What are the tradeoffs between, say, maintaining timber productivity and maintaining biodiversity in a forest, or between average production and resilience? What are the tradeoffs between different levels of these attributes of sustainability and between the net yield or human wellbeing produced and the manner in which it is distributed within society?

If we are able to do this in a self-aware and socially sensitive manner, we will be able to overcome the paralysis of analysis and make a major contribution to the sustainability debate.

The 1991 "Strategy for Sustainable Living" by the same triad of organizations (IUCN 1991 ) says that "sustainable use means use of an organism, ecosystem, or other renewable resource at a rate within its capacity for renewal." Operating within the capacity for renewal clearly is one of the key elements of sustainability, but this formulation does not deal with either nonrenewable resources or the possible off-site, out-of-ecosystem impacts through which exploitation of one resource within its capacity for renewal might adversely affect the renewability of other resources or the sustainability of other ecosystems.

The economist Herman Daly, who has been a pioneer in thinking systematically about these matters recently offered a more helpful three-part specification of the ingredients of sustainability (Daly 1991):

The first of these conditions, by being stated in the aggregate, partly addresses the problem of off-site impacts associated with the exploitation of individual renewable resources: the regeneration rates constraints presumably reflect cross-resource or cross-ecosystem impacts occurring within the overall pattern of resource exploitation.

The second condition, the rate of use of nonrenewable resources, offers a clever solution to the question of how any use of nonrenewable resources can be contemplated within a sustainability framework. Daly offers a detailed formulation on how to ensure that this condition is met, by earmarking part of the proceeds from the exploitation of nonrenewable resources for the development of renewable alternatives.

Daly's third condition, on rates of pollution emission, does not seem as satisfying. If assimilative capacity of the environment means the capacity to assimilate the pollution without any adverse effect on human health or welfare (including through diminution of ecosystem services), the difficulty is that there are many kinds of pollution for which the assimilative capacity, so defined, is probably zero (including, for example, ionizing radiation, chlorofluorocarbons, lead, and more). It does not seem to insist on no harm from pollution as a condition of sustainability; the question is rather what level of harm is tolerable on a steady-state basis, in exchange for the benefits of the activity that produces the harm. [7]

We would also add to Daly's three-part formulation that the first condition applies to resources for which substitution at the required scale is currently and foreseeably impossible (essential resources). It is useful to distinguish those from resources for which substitutes are currently or foreseeably available (substitutable resources). Renewable substitutable resources could be sustainably exhausted on the same basis as nonrenewable substitutable resources (Daily and Ehrlich 1992).

Biogeophysical sustainability in theory and practice

The two most important questions relating to a definition of biogeophysical sustainability are "What is to be sustained?" and "For how long?" It is useful to distinguish, with respect to these questions, between what one would like the answers to be in theory and what one might have to settle for in practice (see table 1-3).

Table 1-3: Biogeophysical Sustainability in Theory and Practice

What is to be sustained? For how long?
In theory, the magnitude and quality of benefit flows that are continuously derivable from the environment Forever
In practice, the magnitude and quality of stocks of environmental resources Half-life of 500 to 1,000 years


Saying that what is to be sustained, in theory, is the magnitude and quality of benefit flows continuously derivable from the environment captures the idea that potential benefits are important, not merely the benefits that society happens to be deriving now. And, of course, saying that the time scale is forever takes the definition of sustainability seriously.

Alas, several practical problems intrude on the attractiveness of this theoretical approach. First, even the existing benefit flows from the environment -- not to mention the potentially continuously derivable benefit flows are partly unknown (indeed, partly unknowable) and also partly incommensurable. (Without commensurability, one is stuck with trying to sustain the individual, in commensurable benefit flows rather than -- more sensibly -- an aggregated total benefit flow within which tradeoffs among different types of benefits could be contemplated.)

Second, insisting that potential benefit flows remain constant over very long periods of time is problematic because environmental conditions and processes -- climate, topography, the biota -- are occurring all the time even in the absence of human interventions. The potential magnitudes of such changes over the very long term make the concept of forever essentially meaningless, at least in relation to the sustainability of conditions that humans of today care about.

Third, it is conceivable that technological improvements will permit well-being to be maintained despite diminished benefit flows from the environment. This argument is probably the one most heavily relied upon by those not convinced of the need to maintain the stream of environmental services undiminished. But attempts to substitute technology for diminishing or otherwise inadequate environmental services invariably entail monetary costs and often generate significant new environmental impacts. In some cases, these additional costs and impacts may more than offset the (presumed) benefits of the activities that necessitated augmentation of the natural environmental services in the first place; and even if it is supposed that this will not be the case, it strikes us as imprudent in the extreme to assume that suitable technology for replacing whatever environmental services are lost will become available in a timely manner and on the requisite scale.

In any case, in light of the difficulties of measuring actual and potential environmental benefit flows, and in light of the conceptual and practical problems of insisting on no degradation forever, it may be necessary in practice to settle for trying to sustain the magnitude and quality of environmental stocks. The time scale on which this ought to be ensured might be defined in practical terms by a resource or stock half-life of 500 to 1,000 years, a period much longer than current planning horizons, but much shorter than geologic time. A tentative rule for prudent practice, then, would be to constrain the degradation of monitorable environmental stocks to not more than 10 percent per century.

Note that degradation of 10 percent a century produces, strictly speaking (that is, with Q = Qo exp[-0.l0t]), a half-life of about 700 years for the resource. Degradation of 20 percent a century would mean a half-life of 350 years, leaving a quarter of the resource remaining after 700 years.

We focus on stocks in this prudent-practice approach, because that is what can most easily be measured (albeit still not all that easily). Although our approach is similar in this respect to the Costanza prescription quoted earlier, an important difference is the specification of a finite rate of degradation as opposed to insistence on maintaining the stocks at just their current level. This sidesteps slightly the argument with the economists and technologists over what is so special about the current levels; putting the argument in terms of degradation rates relies on the presumed circumstance that there is some degradation rate that is too high to be regarded as sustainable, even allowing for economic substitution and technological change.

Of course, it would not really be acceptable to run down environmental stocks at 10 percent a century indefinitely. The point is rather that a rate of 10 percent a century (which after all means about 0.1 percent a year) is slow enough to give society a reasonable chance of figuring out what this degradation is costing, which forms of degradation can be compensated for, how those forms can be stopped that cannot be compensated for or tolerated, and so on, before it is too late. At current degradation rates, by contrast, which are typically an order of magnitude or so higher (that is, in the range of 100 percent a century or more), natural services will be devastated before society even understands what is happening, let alone finds time to take evasive action on the needed scale.

Contrasting views about the sustainability of human activities

Given the above (or any other) definition of sustainability, some obvious questions present themselves:

To environmental scientists, the answer to the first question is clearly no. Current rates of degradation of essential resources are typically an order of magnitude too high (in the range of 100 percent a century or more) for them to qualify as sustainable. The margins by which sustainability is exceeded by various types and combinations of human activity are very difficult to ascertain, however. It follows from the first answer, in any case, that current practices could not possibly sustain even larger flows of goods and services, but whether best-known practices could do so requires further careful analysis.

Although environmental scientists would be in nearly unanimous agreement on the answers just given, many members of other academic disciplines and numerous policymakers would dispute not only these answers but also the relevance of the questions. It is worth looking more closely at the origins of these differences in viewpoint. They undoubtedly arise in part from ambiguities in and disagreements about the meaning of sustainability. A more important source of disagreement, however, are the differing assumptions, perceptions, and knowledge about (a) the importance of environmental conditions and processes in supporting human well-being, (b) the sensitivity of those conditions and processes to disruption, and (c) the character and amenability of society to remedy the anthropogenic impacts now threatening such disruption.

Confusion about the sensitivity of those conditions and processes to disruption is evident in the comment attributed to economist William Nordhaus that only 3 percent of gross national product (GNP) in the United States depends on the environment. In fact, the entire GNP in the US. depends, ultimately, on maintaining the biophysical requisites of sustainability. Furthermore, the importance of agriculture (the economic sector to which Nordhaus apparently was referring) is vastly underestimated by its present share of GNP.

The greatest disparities in interpretation of the relationships between the human enterprise and Earth's life support systems seem, in fact, to be those between ecologists and economists. Members of both groups tend to be highly self-selected and to differ in fundamental worldviews. Most ecologists have a passion for the natural world, where the existence of limits to growth and the consequences of exceeding those limits are apparent. Ecologists recognize that a unique combination of highly developed manual dexterity, language, and intelligence has allowed humanity to increase vastly the capacity of the planet to support Homo sapiens (Diamond 1991); nonetheless, they perceive humans as being ultimately subject to the same sorts of biophysical constraints that apply to other organisms.

Economists, in contrast, tend to receive little or no training in the physical and natural sciences (Colander and Klamer 1987). Few explore the natural world on their own, and few appreciate the extreme sensitivity of organisms -- including those upon which humanity depends for food, materials, pharmaceuticals, and free ecosystem services -- to seemingly small changes in environmental conditions. Most treat economic systems as though they were completely disconnected from the planet's basic life support systems. The narrow education and inclinations of economists in these respects are thus a major source of disagreements about sustainability.

Some of the responsibility for these continuing disagreements also rests, however, on the failure of ecologists and other environmental scientists to make a case for the importance of environmental conditions and processes and for the magnitude of anthropogenic threats to these, in terms understandable by and persuasive to others. This problem is partly a matter of too few environmental scientists having made the effort to articulate a coherent case, but also partly a matter of the great gaps in the environmental science itself. Nor has it helped that environmental scientists are often as ignorant about economic principles, and their relevance to environmental protection, as economists are about ecological principles.

Approaching consensus about biogeophysical sustainability clearly will require more research, more communication across disciplines, and more education of the public and policymakers about a multitude of issues, notably:

The causes and character of environmental damage

Understanding the amenability of the threats to remedy requires a closer look at the factors and trends that are at the root of the problem. An early approach to illuminate this issue was the "I = PAT" formula (Ehrlich and Holdren 1971,1972):

(environmental) impact = population x consumption per person (affluence) x impact per consumption (technology).

Today, a bit of further disaggregation seems useful, so as not to confuse affluence with resource use (the two being separable by means of the inverse efficiency factor, resource use per economic activity) and so as to separate measures of what technology does to the environment (stress) from measures of actual damage, which depends not only on stress but on susceptibility (itself a function of cumulative damage from previous stresses, as well as other factors). Thus,

Damage = population x economic activity per person (affluence)

x resource use per economic activity (resources)

x stress on the environment per resource use (technology)

x damage per stress (susceptibility)

Note that this expanded relation (like the previous I = PAT) is no more and no less than an identity. It is true by definition. People are free to argue about whether it is informative and useful -- and we think it is -- but to argue about whether it is right is foolishness.

Identities of this sort are instructive because they remind us that increases in population, affluence, and the ratio of environmental stress to economic activity (itself clearly a function of the composition of that activity and the technology with which it is accomplished) are multiplicative in their effect on damage, so that the impact of each factor is a matter not only of its own magnitude but also of the magnitudes of the others. [8] At the same time, such identities are deceptive, and above all deceptively simple, in that they fail to make explicit (a) the ways in which the variables on the right-hand side of the equation are not independent, (b) the ways in which institutions, beliefs, and values can influence all of the variables and the nature of the interactions among them, or (c) the ways in which the relative importance of the variables and the nature of the interactions among them vary with location and time.

With respect to the lack of independence of the variables, the magnitude and compost lion of economic activity per person, and their rates of change, are likely to depend in complicated ways on the magnitude and composition of the population and their rates of change. The nature of the technology used to generate economic activity (and thus the kind and magnitude of stresses exerted on the environment by that technology per unit of economic activity) will depend on the magnitude and composition of all economic activity (hence on population and economic activity per person) as well as on their rates of change. The damage to ecosystem services per unit of imposed environmental stress -- a form of dose-response relation -- will generally be a function both of the magnitude and composition of the stress and of their rates of change.

With respect to the role of institutions, beliefs, and values, it is clear, on reflection, that these underlie as well as modulate changes in population, economic activity per person, and the technological variables through which the combination of population and per capita activity exert stresses on ecosystems; and of course it is largely through institutions (economic, political, legal, and so on), through beliefs and values, and through changes in these that damage feeds back to population, economic activity, and technology.

The relative importance of the different causative and modulating factors and the nature and intensity of their interactions clearly vary drastically with the social and ecological contexts, hence with location as well as with time. The situation is further complicated by the wide array of mechanisms by which phenomena in one location and time -- be these phenomena demographic, economic, technological, ecological, political, cultural, or other -- propagate to and influence other locations and times.

Beyond these elaborations about the various contributing factors, it is important to be clear about what we mean by damage. Damage means reduced length or quality of life for the present generation or future generations. Damage may result from short-term alteration of environmental conditions, long-term degradation of environmental capital, and costs of attempts to avoid reductions in length and quality of life with compensating technological and social interventions.

This is of course an explicitly and self-consciously anthropocentric definition, consistent with the anthropocentric definitions of sustainable development that provide the context for this debate. The anthropocentric approach to environmental problems is not the only valid one, but (a) it is the one most likely to succeed in the policy arena and (b) the difficulties in agreeing on definitions, problems, and solutions are even greater if human well-being is not at the center of attention.

Of course, any economic activity will lead to some environmental damage except in cases where the susceptibility factor -- damage per unit of stress -- is zero. Such cases exist when environmental processes are capable of completely absorbing or buffering the imposed stress such that there is no short-term alteration of environmental conditions or long-term degradation of environmental capital of a magnitude sufficient to produce an impact on length or quality of life for any members of the present or future generations of humans. But would many types and levels of economic activity in real-world conditions actually meet this condition?

The critical issue is to specify a level of damage that is acceptable to society. An economist might argue, for example, that we should not refrain from activities that cause any damage, but only from those whose marginal costs (the sum of the internal costs plus the damages as here defined) exceed their marginal benefits. That is, if one could measure all of the costs and all of the benefits in a single currency (such as 1992 dollars), one would define the rational limit on the scale of any economic activity as the level at which the slopes of the cost and benefit curves were equal. Then maximum sustainable abuse (Daily and Ehrlich 1992) would mean the level of abuse (stress) that pushes the total marginal cost (slope of total cost curve) to just equal the total marginal benefit (slope of total benefit curve). Alas, there is no hope of quantifying and monetizing all the diverse kinds of damages associated with economic activity (even the damages occurring in the present, not to mention the problem of bringing future damages into our common currency, which requires agreeing on a discount rate).

In practice, then, cost-benefit-type approaches to determining maximum sustainable abuse are stuck with the problem of apples-and-oranges aggregation of qualitatively different damages, current and future damages, and damages and benefits. Additional daunting problems include dealing with stochasticity and establishing an appropriate margin of safety in the face of uncertainty. All these difficulties mean that tastes and preferences about the proper weighting of different categories become relevant and that the issue is political as much as technical. (A huge literature about risk perception and risk acceptance is relevant in some respects to these issues of maximum sustainable -- or maximum tolerable or maximum prudent -- abuse.)

Ignorance, knowledge, and uncertainty

As suggested earlier, the list of what is not known and what needs to be known in order to address "sustainability" with comprehensiveness and rigor is a very long one. Table 1-4 illustrates this point by presenting in abbreviated form the research agenda on ecological aspects of the issue that was developed recently as part of the Sustainable Biosphere Initiative of the Ecological Society of America (Lubehenco and others 1991). Another compact survey of research requirements related to sustainability is available in the agenda of the International Geosphere-Biosphere Programme of the International Council of Scientific Unions (ICSU 1992). The most important and demanding research sub-agenda of all may be one embedded in the environment-society elements of these lists: namely, the question of how to formulate and implement economic and social incentives for preserving the essential characteristics and functions of environmental systems.

Table 1-4 Research Needed in Ecological Science on Sustainability

Research area Need
Ecological causes and consequences Changes in climate. Changes in atmosphere, soil, and freshwater and marine chemistry.
Ecology of conservation and biodiversity Global distribution of species and change factors. Biology of rare and declining species. Effects of global and regional change on diversity.
Strategies for sustainable ecological systems Patterns and indicators of responses to stress. Guidelines and techniques for restoration. Theory for the management of ecological systems. Introduced species, pests, and pathogens. Integration of ecology with economics and other social sciences.
Source: Lubchenco and others 1991. Source: Lubchenco and others 1991.


At the same time, there is a great danger in falling into the scientist's trap of calling for more research without sufficiently emphasizing what we already know and the implications of that knowledge. We know for certain, for example, that:

This is enough to say quite a lot about what needs to be faced up to eventually (a world of zero net physical growth), what should be done now (change unsustainable practices, reduce excessive material consumption, slow down population growth), and what the penalty will be for postponing attention to population limitation (lower well-being per person).

Of course there are implications of what is not known as well as implications of what is known. The holes in society's knowledge should motivate development of strategies for minimizing the dangers associated with uncertainty. Any sensible prescription for dealing with the kinds of uncertainty we face will include adopting no-regrets strategies, buying insurance, and avoiding the biggest downside risks:

There is, of course, much more to be said about the meaning and measurement of biogeophysical sustainability and about what human societies should be doing about it. But since this chapter is intended only to set the stage for the more detailed treatments to follow, we happily leave the rest to them.


1. A billion is 1,000 million.

2. Although concerns other than the maintenance or enhancement of human well-being can be posited as principles for guiding human behavior (see, for example, Ehrenfeld 1978), we shall accept for the purposes of this chapter that the perspective focuses on the well-being of humans.

3. Some landmarks in this early sustainability literature include Marsh 1864; Vogt 1948; Osborn 1948; Brown 1954; Carson 1962; Ehrlich 1968; Cloud 1969; SCEP 1970; and Meadows and others 1972.

4. A particularly helpful review calling attention to these difficulties is that by Lele 1991.

5. In addition to references cited in note 3 above, some major works include Ehrlich and Ehrlich 1970; Institute of Ecology 1971; Ehrlich and others 1977; CEQ 1980; IUCN 1980,1991;Myers 1984;Mungalland McLaren 1990; Woodwell 1990; Turner and others 1991; Dooge and others 1992; Meadows and others 1992.

6. See, for example, Daly 1973, 1977, 1991; Daly and Cobb 1989.

7. Harm that would qualify as tolerable, in this context, could not be cumulative, else continuing additions to it would necessarily add up to unsustainable damage eventually. Thus, for example, a form and level of pollution that subtract a month from the life expectancy of the average member of the human population, or that reduce the net primary productivity of forests on the planet by 1 percent, might be deemed tolerable in exchange for very large benefits and would certainly be sustainable as long as the loss of life expectancy or reduction in productivity did not grow with time. Two of us have coined the term "maximum sustainable abuse" in the course of grappling with such ideas (Daily and Ehrlich 1992).

8. The following discussion was adapted from the unpublished report of a National Academy of Sciences study group, chaired by Holdren in 1991, on human impacts of ecosystems. See also the acknowledgments to this paper.

9. The idea of society's buying insurance is hardly unprecedented: much of the $300 billion a year that the U.S. spends on defense, for example, represents an insurance policy against contingencies considerably less likely to come about than are some of the environmental disasters one could mention.


Brown, Harrison. 1954. The Challenge of Man's Future. New York: Viking.

Carson, Rachel. 1962. Silent Spring. Boston: Houghton-MiMin.

CEQ (Council on Environmental Quality). 1980. The Global 2000 Report to the President. With the US. Department of State. Washington, D.C.: U.S. Government Printing Office.

Cloud, Preston.1969. Resources and Man. A study and recommendations by the Committee on Resources and Man of the Division of Earth Sciences, National Academy of Sciences, National Research Council, with the cooperation of the Division of Biology and Agriculture. San Francisco: W. H. Freeman.

Colander, D., and A. Klamer.1987. '`The Making of an Economist." Economic Perspectives 1, pp. 95-111.

Costanza, Robert, ed. 1991. Ecological Economics: The Science and Management of Sustainability. New York: Columbia University Press.

Daily, Gretchen C., and Paul R. Ehrlich. 1992. "Population, Sustainability, and Earth's Carrying Capacity." BioScience 42:10, pp. 761-71.

Daly, Herman E.1977. Steady-State Economics. San Francisco: W. H. Freeman.

. 1991. "Elements of Environmental Macroeconomics." In R. Costanza, ea., Ecological Economics. New York: Columbia University Press.

Daly, Herman E., ed. 1973. Toward a Steady-State Economy. San Francisco: W. H. Freeman.

Daly, Herman E., and J. B. Cobb, Jr.1989. For the Common Good: Redirecting the Economy toward Community, the Environment, and a Sustainable Future. Boston: Beacon.

Diamond, Jared.1991. The Rise and Fall of the Third Chimpanzee. London: Radius.

Dooge,J. C. I., G. T. Goodman,J. W. M. La Riviere, J. Marton-Lefevre, T. O'Riordan, F. Praderie, and M. Brennan,eds. 1992. An Agenda of Science for Environment and Development into the 21st Century. Cambridge, England: Cambridge University Press.

Ehrenfeld, David. 1978. The Arrogance of Humanism. New York: Oxford University Press.

Ehrlich, Paul R. 1968. The Population Bomb. New York: Ballantine.

.1980. "Variety Is the Key to Life." Technology Review 82 (March-April), pp. 58-68.

Ehrlich, Paul R., and Anne H. Ehrlich.1970. Population, Resources, Environment. San Francisco: W. H. Freeman.

Ehrlich, Paul R., Anne H. Ehrlich, and John P. Holdren.1977. Ecoscience. Population, Resources, Environment. San Francisco: W. H. Freeman.

Ehrlich, Paul R., and John P. Holdren.1971. "Impact of Population Growth." Science 171, pp. 1212-17.

.1972. "One-Dimensional Ecology." Bulletin of the Atomic Scientists 28:5 (May), pp.16, 18-27.

Holdren,John P.1973. "Population and the American Predicament." Daedalus (Fall), pp. 31-34.

Howarth, R. B., and R. B. Norgaard. 1990. "Intergenerational Resource Rights, Efficiency, and Social Optimality."." Land Economics 66, pp. 1-11.

Institute of Ecology.1971. Man in the Living Environment. Madison, Wisc.: Institute of Ecology.

ICSU (International Council of Scientific Unions). 1992. Global Change: Reducing Uncertainties. Stockholm: Royal Swedish Academy of Sciences, International Geosphere-Biosphere Programme.

IUCN (International Union for the Conservation of Nature). 1980. World Conservation Strategy: Living Resource Conservation. With the United Nations Environment Program and the World Wildlife Fund. Gland, Switzerland.

. 1991. Caring for the Earth: A Strategy for Sustainable Living. With the United Nations Environment Program and the World Wildlife Fund. Gland, Switzerland.

Lele, Sharachchandra M. 1991. "Sustainable Development: A Critical Review." World Development 19:6, pp. 607-21.

Lubchenco, Jane, A. Olson, L. Brubaker,, S. Carpenter, M. Holland, S. Hubbell, S. Levin, J. MacMahon, P. Matson, J. Melillo,, H. Mooney,

C. Peterson, H. Pulliam,, L. Real, P. Regal, and P. Risser.. 1991. "The Sustainable Biomass Initiative: An Ecological Research Agenda." Ecology 72, pp. 371-412.

Marsh, Gerge Perkins. 1864. Man and Nature, or Physical Geography as Modified by Human Action. Reprinted in 1965 by Cambridge, Mass.: Belknap Press of Harvard University Press.

Meadows, Donella H., Dennis L. Meadows, and Jorgen Randers. 1992. Beyond the Limits. Post Mills, Vt.: Chelsea Hills Publishing Co.

Meadows,Donella H., Dennis L.Meadows,Jorgen Randers, and W. W. Behrens III. 1972. The Limits to Growth. New York: Universe Books.

Mungall, Constance, and Digby J. McLaren, eds. 1990. Planet under Stress. New York: Oxford University Press.

Myers, Norman, ed.1984. GAIA: An Atlas of Planetary Management. New York: Doubleday.

Osborn, Fairfield.1948. Our Plundered Planet. Boston: Little, Brown & Co.

Pronk, Jan, and Mahbubul Haq.1992. Sustainable Development: From Concept to Action. The Hague Report. New York: United Nations Development Program.

SCEP (Study of Critical Environmental Problems). 1970. Man's Impact on the Global Environment. Cambridge, Mass.: M.I.T. Press.

Turner, B. L., William C. Clark, Robert W. Kates, John F. Richards, Jessica T. Mathews, and William B. Meyer, eds. 1991. The Earth as Transformed by Human Action. Cambridge, England: Cambridge University Press.

Vogt, William. 1948. Road to Survival. New York: Sloane.

Woodwell, George M., ed. 1990. The Earth in Transition: Patterns and Processes of Biotic Impoverishment. New York: Cambridge University Press.

WCED (World Commission on Environment and Development).1987. Our Common Future. Oxford, England: Oxford University Press.