Resilience, Sustainability, and Environmentalism
S. Lélé / Pacific Institute 2000
Ecological Economics Unit
Institute for Social and Economic Change
Bangalore 560 072
An occasional paper of the:
Pacific Institute for Studies in Development, Environment, and Security
654 13th St.
Oakland, CA 94612 USA
Editor Editor’s ’s Note:
An earlier version of this paper was published in the journal 3(2) E ENVIRONMENT & DEVELOPMENT ECONOMICS 249- 254 (1998),
as a part of a policy forum on resilience and sustainability
Resilience is turning out to be a resilient concept. First proposed way back in the 1970s in the context of ecosystem dynamics, it was then dissected and elaborated, spawning terms such as “malleability,” “elasticity,” “hysterisis,” “inertia,” “resistance,” “amplitude,” as ecologists struggled to make it into something measurable, usable, and distinct from its notoriously slippery predecessor concept, “stability.” But in the post- Brundtland era, the focus appeared to have shifted to the umbrella concepts of “sustainability” (Brown et al., 1987; Lubchenco et al., 1991, Levin, 1993), “ecosystem health” (Schaeffer et al., 1988; Costanza et al., 1992) or “ecosystem integrity” (Regier, 1993; Angermeier and Karr, 1994). More recently, however, the concept of resilience has regained prominence: a group of ecologists and economists (Levin et al., 1998) has in fact made a strong pitch for applying it to the integrated analysis of ecological and socioeconomic systems and called upon policymakers to “build resilience” into social and environmental systems. Why has resilience been resuscitated? How can it be operationalised? How does it relate to sustainability? And can either resilience or sustainability be considered sufficient for environmental soundness?
Resilience, “the ability to recover from [presumably severe shocks or stress]” as the dictionary puts it, is a flag to draw attention to the need to incorporate non-linearities in models of socio-ecological systems. Superficially, it might seem surprising that it should take so much effort to get us to recognize the importance of non-linearities. After all, we experience nonlinearities and irreversibilities on a daily basis: plants die of too much watering as well as too little, children falter with too little guidance but rebel at too much. Non-linearities are well recognized in traditional philosophies and knowledge systems, be they those of Indian sages, Amazonian hunters, dialectical Marxists, or even 19th century natural historians. Nevertheless, those branches of natural science (agriculture, fisheries, forestry, even medicine) and of social science (essentially economics) that most directly inform public policy today appear to be oblivious to such phenomena, sticking to their linear mechanistic mindsets. That such approaches continue to hold sway in face of everyday experience, traditional philosophy, and also recent developments in non-linear systems theory attests to their hegemony over the policy-making process (Holling et al., 1995; Norgaard, 1987). Any attempt to break these mindsets and hegemonies is therefore to be vigorously applauded.
Once we accept the basic thrust of the argument that the concept of resilience should be built into our policy process, we can proceed in the traditional manner of science, viz., developing a clear taxonomy and hypotheses. How then should resilience be defined? Here, I found Levin et al.’s treatment inadequate and had to take recourse to the work of Holling (1973), who states that resilience is the size of the stability domain around stable time-invariant equilibria (point attractors) or stable oscillations (periodic attractors). From this definition, a number of points follow. First, resilience should not be measured in terms of the distance of the current state of the system from the edge of the stability domain, which is a constantly changing parameter. This in turn implies that reductions in the magnitude of the excursion away from equilibrium should not be viewed as increases in resilience. Second, reductions in the perturbing force or in the deviation per unit perturbing force (the latter being the inverse of “robustness” or “inertia” (Westman, 1986) should not be confused with increases in resilience. They are complementary ways of ensuring the same end result (viz., system stays within the stability domain) as increasing resilience (increasing the size of the domain). Third, resilience must refer to a situation where the perturbation applied to the system is significant (to cause it to move far away from equilibrium) but temporary (a shock or a period of stress: Westman, 1986), because a continuously applied perturbation will eventually drive any system out of its stability domain, regardless of the size of the domain.
In the absence of sufficient resilience, i.e., if the perturbation drives the system out of a stability domain, the system may either collapse (when there is no other stable equilibrium) or may enter another stability domain, characterised by a different system structure, such as a reversed Gulf Stream. Intuitively, this seems to be the situation in which one should use the term adaptability: the ability of a system to keep some “ultimate” desired state variables (say net food production) at the desired level in the face of domain shifts in “underlying” ecosystems (say the agro-climatic system). Again, one can talk about the time required to adapt, the extent of recovery to the old level of the desired variable, etc. as different aspects of adaptability. But it is clearly useful to distinguish adaptability from resilience. For instance, an agricultural system may be resilient to (able to recover from) occasional severe droughts, but it may not be able to adapt to a shift to a significantly drier climatic regime.
Using the above framework, one can begin to evaluate various hypotheses about the resilience of socio-ecological systems advanced by Levin et al. and a few others. First, if the perturbation is anthropogenic (such as net CO2 emissions) and clearly of significant magnitude, then, regardless of the exact nature of the system and one’s distance from equilibrium or the edge of the stability domain, reducing the perturbing force (stopping the burning of fossil fuels) is clearly one (and probably the best) means of reducing the chances of a disastrous domain shift (say in the climate system), even though this does not really constitute an increase in system resilience. And a careful distinction between resilience and adaptability would, for instance, prevent the climate change debate from being hijacked by the adaptationists.
Second, a key result of resilience research (mentioned in passing by Levin et al.) is that one should attempt to work with natural variations. Small perturbations should be utilized to build resilience rather than be suppressed in order to reduce variability of the state variable. For instance, it is better to leave mild illnesses untreated so as to build immunity rather than ingest antibiotics at the mildest sneeze. It follows that Levin et al.’s general statement that “effective feedback is necessary for resilience” needs to be qualified, because taking medicines at the mildest sneeze is in fact a sign of a system with a very effective negative feedback! Normal negative feedbacks reduce short-term variability, what resilience needs is a non-linear negative feedback: nil at low values of deviation and high at values close to the edge of the stability domain.
Third, competition—and the positive feedback it provides—may lead to greater efficiency, but it is not a necessary condition for resilience: most traditional resource management systems, now viewed as resilient and ecologically well-adapted (Berkes et al., 1994), have evolved in non-competitive communitarian settings. Fourth, the link between risk-spreading and resilience is more complex than suggested by Levin et al.. If risk is adaptively internalized, by, for example, a person developing multiple skills, the results are different than if it is externalised, by the person continuing to specialise but hooking up with a much larger system: the regional or national job market. The latter approach increases the connections between the elements (connectance) of the overall system. It has been shown in a number of cases, ranging from food webs (May, 1973; Siljak, 1978, who worked on “connective stability”) to trade networks (Siljak, 1978), electrical systems (Fink, 1991), and (qualitatively) even for stock markets (Rochlin, 1991), that indiscriminately increasing connectance may increase efficiency and even asymptotic stability, but reduces resilience to structural perturbations in the system. Furthermore, “dynamical systems composed of interconnected subsystems are stable [with respect to disruptions in connectance] if the subsystem are self-contained and the interdependence between the subsystems is properly limited” (Siljak, 1978, p.2): something for us to ponder over in this era of indiscriminate globalization, free trade, and networking.
Fifth, resilience may in fact require some “slack capacity” (Rochlin, 1997) that is relatively “unplugged” from the larger system (such as a stash of gold jewelry at home as against simply a diversification of one’s stockmarket investment portfolios), so that this capacity is unaffected by shocks in the larger system. “Exploiting all opportunities for mutual gain”, an attribute valourized by Levin et al., may in fact reduce resilience by leaving no such “slack”. On the other hand, adaptability may require a different kind of slack: a store of as yet unexplored, unvalued resources (such as biodiversity) and of course the ability to learn.
Clearly, since resilience and adaptability are defined with respect to a stability domain, the notions of equilibrium and stability continue to be relevant even after incorporating nonlinearity. In general, for a system to be able to persist (in some desired form or with some desired properties) over time, i.e., to sustain, requires that it be at or around some equilibrium, have some stability in the face of small (“normal”) perturbations, some resilience in the face of large (“abnormal”) perturbations, and some adaptability to domain shifts. Thus, rather than stretch and pull the concept of resilience (sometimes beyond all recognition) or cast the debate in terms of short-term stability versus long-term resilience, it would be more appropriate to think of these properties as different attributes of sustainability—a concept general enough to serve in most discourses as one of society’s meta-objectives (Lélé, 1988; Lélé, 1993).
Particular situations would require more emphasis on particular attributes: systems characterised by high degree of environmental variability (such as semi-arid regions or turbulent business conditions) must give primacy to resilience (hence the domination of r-selected grasses or small, loosely structured, opportunistic firms), while those characterised by low environmental variability (such as moist tropical regions or stable economies) permit the neglect of resilience (hence the domination of k-selected trees and complex rainforests or large, complex firms). Each specialisation comes at the cost of some other qualities, and has an associated productivity gain under certain conditions. Research may no doubt have hitherto focussed disproportionately on stable ecosystems. But to insist that resilience is somehow more “fundamental” would be akin to insisting that k-selected species are unfit to survive.
Finally, it needs to be pointed out that the critique of conventional socio-environmental science is much broader than just the problem of not incorporating non-linearities. From a scientific perspective, the study of complex socioecological systems also suffers from the problem of reductionism—the tendency to look at tree growth only as a function of age rather than thinking of succession, disease and pollination in modelling forest stands (Holling et al., 1995)—and the lack of methodological pluralism (Norgaard, 1989). From a social perspective, conventional (primarily Western) environmental science is also characterized by narrow value systems. Both resilience as a goal in itself and sustainability as an over-arching goal essentially pertain to the temporal dimension of human well-being. There is, however, also a simultaneous or primarily spatial dimension to many environmental problems, where current actions of one group affect the current well-being of another group. Typically, these spatial externalities (and the political power of the actors involved) are asymmetrical, if not entirely unidirectional. Unless one takes a clear position that in addition to our concern for the future, intra-generational justice is also a fundamental value, these environmental problems will get short shrift. For instance, when upstream factories pollute rivers that are the primary source of drinking water for downstream communities, the argument that the pollution is reducing “long-term river ecosystem resilience” or “water use sustainability” will not cut much ice with the factory owners: one has to invoke the notion of intragenerational justice. Similarly, a concern for the climate we may pass on to our future generations will not in itself prevent unfair arm-twisting by the North. The North would prefer to increase the resilience of the global climate system by buying up forestlands and emission rights from the South at historically biased exchange rates, rather than reduce its own level of fossil fuel consumption. The lessons from the history of environmental policy are unambiguous: no amount of “trust”, “clever institutional design” (e.g., the already faltering Montreal Protocol), or “epistemic consensus” (e.g., the IPCC: but see Jasanoff, 1992) can compensate for major asymmetries in the interests and powers of the different actors. And no amount of concern for long-term resilience of the human ecosystem can by itself ensure a fair environmentalism or a just development. Levin et al. do not claim that it does, but in a world of limited attention spans and paltry environmental budgets, we must ensure that resilience does not knock more pressing but politically inconvenient matters off the agenda.
Financial support from the John D. and Catherine T. MacArthur Foundation’s World Environment and Resources Program is gratefully acknowledged.
Angermeier, P. L. and J. R. Karr (1994), “Biological integrity versus biological diversity as policy directives”, BioScience, 44(10): 690-697.
Berkes, F., C. Folke and M. Gadgil (1994), “Traditional Ecological Knowledge: Biodiversity, resilience and sustainability”, in C. A. Perrings, K.-G. Maler, C. Folke, C. S. Holling and B.-O. Jansson, eds., Biodiversity Conservation, Stockholm: Kluwer Academic, pp. 281-300.
Brown, B. J., M. Hanson, D. Liverman and R. Merideth Jr. (1987), “Global sustainability: Toward definition,” Environmental Management, 11(6): 713-719.
Costanza, R., B. G. Norton and B. D. Haskell (1992), Ecosystem Health: New Goals for Environmental Management, Washington D. C.: Island Press.
Fink, L. H. (1991), “Introductory overview of voltage problems,” in L. H. Fink (ed), proceedings of Bulk Power System Voltage Phenomena II: Voltage Stability and Security held at Deep Creek Lake, Maryland, U.S.A., ECC, Inc.
Holling, C. S. (1973), “Resilience and stability of ecological systems,” Annual Review of Ecology and Systematics, 4: 1-23.
Holling, C. S., F. Berkes and C. Folke (1995), “Science, sustainability and resource management,” Beijer Discussion Paper Series no. 68, Beijer International Institute of Ecological Economics, Stockholm.
Jasanoff, S. (1992), “Pluralism and convergence in international science policy,” in Science and sustainability: selected papers on IIASA’S 2Oth anniversary, Laxenburg, Austria: International Institute for Applied Systems Analysis, pp. 157-180.
Lélé, S. (1988), “The concept of sustainability,” in A. T. Charles and G. N. White III, eds., Natural Resources Modelling and Analysis, Halifax, Nova Scotia, Canada: Centre for Resource Systems Analysis, pp. 46-48.
Lélé, S. (1993), “Sustainability: A plural, multi-dimensional approach,” Working Paper, Pacific Institute for Studies in Development, Environment, & Security, Oakland, U.S.A.
Levin, S. (1993), “Forum: Science and sustainability,” Ecological Applications, 3(4): 545-6.
Levin, S., S. Barrett, S. Aniyar, W. Baumol, C. Bliss, B. Bolin, P. Dasgupta, P. Ehrlich, C. Folke, I. Gren, C. S. Holling, A. Jansson, B.-O. Jansson, K.-G. Maler, D. Martin, C. Perrings and E. Sheshinski, 1998, “Resilience in natural and socioeconomic systems”, Environment and Development Economics, 3(2): 222-235.
Lubchenco, J., A. M. Olson and L. Brubaker, et al. (1991), “The Sustainable Biosphere Initiative: An ecological research agenda,” Ecology, 72(2): 371-412.
May, R. M. (1973), Stability and Complexity in Model Ecosystems, Princeton, NJ: Princeton University Press. Norgaard, R. B. (1987), “Economics as mechanics and the demise of biological diversity,” Ecological Modelling, 38: 107-121.
Norgaard, R. B. (1989), “The case for methodological pluralism,” Ecological Economics, 1(1): 37-57.
Regier, H. A. (1993), “The notion of natural and cultural integrity,” in S. Woodley, J. Kay and G. Francis, eds., Ecological Integrity and the Management of Ecosystems, Delray Beach, Florida, U.S.A.: St.Lucie Press,
Rochlin, G. I. (1991), “Jacking into the market: Trans-national technologies and global securities trading,” paper presented at the 3rd International Conference on Large-Scale Systems, Sydney, Australia.
Rochlin, G. I. (1997), Trapped in the Net: The Unanticipated Consequences of Computerization, Princeton, NJ: Princeton University Press.
Schaeffer, D. J., E. E. Herricks and H. W. Kerster (1988), “Ecosystem health: I. Measuring ecosystem health,” Environmental Management, 12: 445-455.
Siljak, D. D. (1978), Large-Scale Dynamic Systems: Stability and Structure, New York: North-Holland.
Westman, W. E. (1986), “Resilience: concepts and measures,” in B. Dell, A. J. M. Hopkins and B. B. Lamont, eds., Resilience in Mediterranean-type Ecosystems, Dordrecht, Netherlands: Dr. W. Junk, pp. 5-19.
If you have come to this page from an outside location click here to get back to mindfully.org