Evolution relevant for environmental science

Transcription

1 Evolutionary Modelling for Environmental Policy Jeroen C.J.M. van den Bergh Dept. of Spatial Economics Faculty of Economics and Business Administration & Institute for Environmental Studies (Vrije Universiteit) Evolution relevant for environmental science (Un)desirable technology: progress, unlocking, transitions. Humans exert selection pressure: genetics of biological populations (e.g., pesticide use). Environment and resources in human-economic history: (co)evolution. Policy and management: behaviour, trade-off efficiency-diversity, common pool resources. Penn, D.J. (2003). The evolutionary roots of our environmental problems: towards a Darwinian ecology. The Quarterly Review of Biology 78(3): Department of Spatial Economics & Institute for Environmental Studies 2 1

2 Evolution in environmental economics Boulding (1982): ecological-evolutionary analogies. Norgaard (1985): coevolution. Faber and Proops (1990): growth & technology. Gowdy (1994): macroevolution economic growth. Sethi & Somathan (1996): norms in CPR. Munro (1997): selection by pesticides. J.C.J.M. van den Bergh (2006). Evolutionary thinking in environmental economics. Working paper, VU,. Department of Spatial Economics & Institute for Environmental Studies 3 Evolutionary methods for specific questions Aggregate models: Evolutionary game theory (both theory and method). Dynamic probability distributions. (Adaptive) multi-agent models. Explicit interaction between agents (real micro). Structure of interaction and environment: spatial lattice, network or gaseous cloud. Department of Spatial Economics & Institute for Environmental Studies 4 2

3 Population thinking essential: Distinguishing feature from neoclassical economics and systems modelling Real microeconomics: no representative agent. Variety of characteristics: population of agents. Distribution of characteristics matters: affects system dynamics. Systems perspective: coevolution versus co-dynamics of subsystems. Winder, N., B.S. McIntosh and P. Jeffrey (2005). The origin, diagnostic attributes and practical application of co-evolutionary theory. Ecological Economics 54(4): Department of Spatial Economics & Institute for Environmental Studies 5 Models of local interaction relevant? Ecological: water, air, fauna, vegetation (biodiversity) - landscape ecology. Economic relevance: communication, cooperation, learning, imitation (spatial economics). But economic processes involve very often non-local interactions: phone, fax, (e)mail, internet, international trade, international cooperation/agreements. Some (important) environmental processes also nonlocal: uniformly mixing pollutants - GHG. Department of Spatial Economics & Institute for Environmental Studies 6 3

4 Windrum paper Very rich discussion of technology, innovation, path-dependence and modelling. Emphasis on path-dependence/lock-in appropriate: relevant to transition debate NL. Demand and supply side factors. Short-term focus on efficiency/cost-effectiveness: e.g., energy conservation versus renewable energy reinforcing fossil fuel lock-in. Also standards face this trade-off. Complex selection environment: should policy equalize selection for old & new technologies. Non-compatible vs complementary ( interoperable ) technologies; or hybrid technologies: how to model these? Modelling interaction among phases? RDD&D. Lock-in at individual or across levels? Are there currently models of lock-in that are policy relevant (enough detailed mechanisms to link policy instruments to)? Evolutionary aspects sometimes explicit but not always clear modular innovation similar to recombination? Policy highly cumulative: path-dependency makes policy change difficult (public transport infrastructure; some government branches reinforce lock-in car). Policy maker boundedly rational: future election vs opinion on problems and alternative options. Suggested model section 5.2: nice, complex, but also ad hoc? Numerical results? Department of Spatial Economics & Institute for Environmental Studies 7 Saint-Jean paper Dyn. efficiency effect of env. regulation: via innovation and buyer-supplier relations. Environmental innovations are innovations: general innovation theories/models apply. But large diffusion of environmental innovations is socially desirable (not just niche). Environmental innovation competes with old technology: do the same but better. Negative + positive externality: env. regulation + innovation policy (= transition policy). Impact regulation on technology depends on instrument; innovation and diffusion require perhaps different incentives (theoretical evol. models desirable Jaffe et al., 2001). Model of industrial dynamics: similar aspects as Windrum supply-demand & path-dep. But model is quite different from Windrum (e.g., stochastic elements, inter-firm interactions). Why? No convergence of ideas (yet)? BTW: are analytical models feasible? Is industrial clients needed? Or wider interpretation possible (consumers)? What are the core modules of an evolutionary model to address system innovations (e.g., Iwai on diffusion, Arthur on increasing returns due to imitation & information externalities, Silverberg/Verspagen on selection-cum-accumulation)? Policy: product vs process standard; different impacts on competition and innovation? What RDD&D phases (invention/innovation/diffusion) do two instruments influence? Emission standards leave more diversity/flexibility; product standards enforce lock-in? Paradigm shift (section 3.2) unclear: under which policy conditions? Jaffe, A.B., R.G. Newell and R.N. Stavins (2001). Environmental Policy and Technological Change. Environmental and Resource Economics 22: Department of Spatial Economics & Institute for Environmental Studies 8 4

5 Jager paper Very clear and convincing set of suggestions to perform agent-based modelling. Motivation: for all environmental policy issues (i.e. general approach)? Spatial and network structure useful or even needed? Evolutionary outcomes may be very sensitive to these (local interactions)? Data collection on this realistic? Not too simple rules : if a simple model can (conditionally or temporally) predict should one not be satisfied? In addition, go slowly from simple to complex models has proven to be a robust research strategy. Relate to other studies, not ad hoc. In subjective well-being or happiness research by economists, psychologists and sociologists two phenomena are commonly reported: adaptation (preference drift) and status rivalry (reference drift). What role do you assign them? Other aspects: temporal and spatial myopia (linked to networks/space). Diversity of boundedly rational strategies and social preferences (envy, altruism, cooperation, punishment, reward, communication, imitation) makes choice of model difficult (ad hoc). Lack of firm empirical basis further complicates matters. Decision-making under uncertainty special issue; much research on by economists and psychologists (rank dependent utility theory, prospect theory). Utility separated from (subjective) probability weighting. Does this fit into your approach. Department of Spatial Economics & Institute for Environmental Studies 9 5