DPG Phi
Verhandlungen
Verhandlungen
DPG

Dresden 2017 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

SOE: Fachverband Physik sozio-ökonomischer Systeme

SOE 18: Economic Models II

SOE 18.2: Talk

Thursday, March 23, 2017, 15:15–15:30, GÖR 226

A pure optimization paradigm is not sufficient to account for sustainable policies — •Wolfram Barfuss1,2, Jonathan F. Donges1,3, Steven Lade3, and Jürgen Kurths1,2,41Potsdam Institute for Climate Impact Research, GER — 2Humbold University, Berlin, GER — 3Stockholm Resilience Centre, Stockholm University, SWE — 4University of Aberdeen, UK

Optimization is a widely used paradigm to deduce the course of action in many sustainability contexts, from integrated assessment models to natural resource management. Simultaneously, a wide range of criticisms and refinements of the optimization approach exist. These include aspects involving the discounting of future rewards and the treatment of multiple kinds of uncertainty. Here we demonstrate by a counterexample that a pure optimization of accumulated discounted rewards is not sufficient to reach a sustainable policy. This is done by introducing a conceptual model example based on a Markov decision process, formalizing a social-ecological tipping interaction. We translate the notion of sustainability into a definition of sustainable policy, which is capable of 'meeting the needs of the present without compromising the ability to meet those of the future' by introducing a minimum acceptable reward value. We further introduce a general return function, unifying a discounted with an average reward setting. The simplicity of our model allows a full analytical treatment, including a discussion of the discount factor as a free parameter. Overall, this suggests that care should be taken under what conditions an optimization approach is used to not result in undesired outcomes.

100% | Mobile Layout | Deutsche Version | Contact/Imprint/Privacy
DPG-Physik > DPG-Verhandlungen > 2017 > Dresden