تجربه گرایی در اقتصاد محیط زیست: یک چشم انداز از نظریه سیستم های پیچیده

Economies are open complex adaptive systems far from thermodynamic equilibrium, and neo-classical environmental economics seems not to be the best way to describe the behaviour of such systems. Standard econometric analysis (i.e. time series) takes a deterministic and predictive approach, which encourages the search for predictive policy to ‘correct’ environmental problems. Rather, it seems that, because of the characteristics of economic systems, an ex-post analysis is more appropriate, which describes the emergence of such systems’ properties, and which sees policy as a social steering mechanism. With this background, some of the recent empirical work published in the field of ecological economics that follows the approach defended here is presented. Finally, the conclusion is reached that a predictive use of econometrics (i.e. time series analysis) in ecological economics should be limited to cases in which uncertainty decreases, which is not the normal situation when analysing the evolution of economic systems. However, that does not mean we should not use empirical analysis. On the contrary, this is to be encouraged, but from a structural and ex-post point of view.

Ecological economics deals with, and is related to, policy generation and, in order to do this needs numerical data about both human and natural systems. It is the goal of this paper to analyse the role of empiricism in the framework of neo-classical environmental economics and ecological economics. After doing that, the paper defends a phenomenological and ex-post analysis to deal with the complexity of modern economies, by giving some examples of empirical work already done under this view. The concepts underlying ecological economics and neo-classical environmental economics will be outlined, to emphasise that the latter makes some strong implicit assumptions about the working of systems under its analysis (i.e. economic systems). These assumptions are compatible neither with the main characteristics of present complex environmental systems nor with the nature of economies. This is why ecological economics deals with both the problems and the systems in an alternative way. The structure of the rest of the paper is as follows: Section 2 focuses on the conceptual structures in ecological economics and in neo-classical environmental economics from an evolutionary perspective based on the concept of time. Section 3 presents the debate about the role of policy for sustainability purposes. Section 4 presents the position of these two schools of economic thought on empirical analysis, focusing on time and evolution. With this background, Section 5 mentions some of the latest developments in empirical analysis that have been published in the field of ecological economics, and that are an example of what could be empirical analysis when dealing with complexity in ecological economics. Finally, Section 6 reaches the conclusion that a predictive use of econometrics in ecological economics should be limited to cases in which uncertainty decreases. This leads to presenting the way ahead regarding empirical analysis in ecological economics, and its relationship to policy formulation.

The criticism presented here on the use of the positivist version of empirical analysis does not mean that we cannot conduct some forecasts about the future behaviour of the variables. We can do it, provided that we are analysing the variable or the system when they are near or at, one attractor point (i.e. they are meta-stable) or when they are following a well-established trend identified historically. In these cases, when the level of uncertainty decreases, prediction is possible, under certain limitations (a sudden change is always possible). However, when the system is at a bifurcation point (i.e. in the very moment of shifting from one attractor point), prediction is not possible because we might have novelty expressed either by an external shock or by internal causality, which will drive the system towards one attractor or other. For example, internal causality may be caused by feedback loops between the different hierarchical levels of the system. We should bear in mind that when the differences in scale are too large, it is almost impossible to relate the non-equivalent information obtained from the different levels, making prediction almost impossible. This is a reflection of the unavoidable indeterminacy of the representation of these systems across scales (Mandelbrot, 1967). So, if a basic characteristic of complex systems is that “they can only be approximated, locally and temporarily, by dynamical systems” (Rosen, 1987, p. 134), but we still try to control them by using predictive dynamic models, we may face a “global failure” (Rosen, 1987, p. 134, emphasis in the original) in the form of a growing discrepancy between what the system is doing and what the model predicted. This is one of the reasons why normal science is losing credibility among citizens, and why post-normal science, with its interest not in finding ‘truth’ but on giving good quality information for the decision-making process, is viewed as a way out of that difficulty. When analysing data, in order to tackle complexity we can adopt the idea of triangulation (Ramsay, 1998) or parallel non-equivalent descriptions (Giampietro and Mayumi, 2000a). This idea consists of using more than one source for the data, analysing the data with different theories or models, or using different hierarchical levels at the same time, in order to gain robustness in our analysis and give more credibility to scientific analysis. This will bring redundancies, which are rather positive since they will reinforce the argument or the regularities that we may find. This is thus an argument in favour of a inter-disciplinary approach to sustainability, in which the different readings of the different disciplines are seen as compatible in generating the overall understanding of the structure of the system, and its development. If the use of empiricism for prediction by econometrics (i.e. time series analysis) is very limited, what kind of empiricism can we use? In ecological economics we are interested in evolution, the process of becoming, structural change and the emergence of novelty; therefore, first we have to bear in mind that since stochastic processes are dominant in nature, scientific theories should be more down-to-earth, based in direct observations. Then, we should use empirical analysis not to give the exact values of the parameters in future, but to discriminate between those theories, which are consistent with reality and those which are not. In cases of high uncertainty we should, therefore, describe and understand instead of seeking to explain and predict. Hence, we would be closer to the generation of scenarios by using narratives (i.e. soft modelling) rather than to forecasting (i.e. hard modelling). This is so because the nature of evolutionary complex adaptive systems, characterised by irreversibility and stochasticity, with their numerous possible trends, their uncertainty, the emergence of novelty, makes them largely unpredictable. That is, ex-ante modelling is often not useful for policy. We have to admit that there are no deterministic explanations (universal and a-historical). Rather we can describe and understand these systems by finding historical and spatial, regularities, and by looking at the emergence of such systems’ properties. This leads us to admit that the knowledge we can obtain from complex systems is context dependent (Clark et al., 1995); it is dependent on the time window considered and also on the spatial context. This is the reason why, as pointed out by Boulding (1987), the failure in our predictions are not the responsibility of human knowledge itself. Rather, it reflects an inherent property of complex systems, that of unpredictability. Therefore, our failure might come either because we do not know the parameters of the system (ignorance) or because they change very rapidly (emergence of novelty, evolution) reflecting structural or genotypical change caused by external shocks or by internal causality within systems (e.g. chaotic behaviour). Science applied to the decision-making process under the post-normal science framework would then be limited to assessing the consequences of the different policies, and to providing a phenomenological narrative or interpretation of how the future might unfold (Kay et al., 1999). This is part of the process of guaranteeing transparency and fairness in the process of decision-making, by promoting a continuous dialogue with stakeholders and policy makers. Thus, “these narratives focus on a qualitative/quantitative understanding which describes: • the human context for the narrative; • the hierarchical nature of the system; • the attractors which may be accessible to the system; • how the system behaves in the neighbourhood of each attractor, potentially in terms of a quantitative simulation model; • the positive and negative feedbacks and autocatalytic loops and associated gradients which organise the system about an attractor; • what might enable and disable these loops and hence might promote or discourage the system from being in the neighbourhood of an attractor; and • what might be likely to precipitate flips between attractors” (Kay et al., 1999, p. 728). The implication of the argumentation presented before is that the evolution of complex systems is not fully predictable. This fact leads us, when dealing with sustainability, to the issue of incommensurability of values as a key characteristic that should distinguish ecological economics from environmental economics (Martı́nez Alier et al., 1998). Thus, the fact that the future is open has some repercussions from a policy perspective. This openness asks for what has been called ‘soft management’ by Haken and Knyazeva (2000). This has to be understood as encouraging flexibility in response to changing boundary conditions. This flexibility can be achieved by enhancing the diversity in the system. The more diversity, the more responses we will have to changing conditions, with more chances that one, or some of these responses, will be successful and will bring the system ahead in its development. That is, diversity increases the adaptive capacity of the system. This is something that Holling (1996), following Walker et al. (1969), called ‘ecological resilience’, in contrast to ‘technical resilience’ which would be closer to finding and returning to the initial equilibrium. In conclusion, in complex systems prediction is very often not possible not only because the parameters defining the relationships between variables may change (phenotypic evolution), but also because the functional relation itself may also change (genotypic evolution) since they are involved in the process of becoming of the system, generating, therefore, more novelty. Consequently, a predictive use of econometrics (i.e. time series analysis) in ecological economics when dealing with complex systems should be limited to those non common situations in which uncertainty decreases. Rather, the phenomenological approach presented here, and exemplified by the papers mentioned in Section 5 dealing with an ex-post analysis, seems more suitable for ecological economics to deal with the issue of evolution of complex systems such as economies, involving novelty in the form of structural change. This may also include, as stated above, the use of econometric analysis to account for past developments and trends. At the end, history does count.