Five Essays in the Economics of Climate Engineering,Research, and Regulation under Uncertainty

Abstract

This thesis revolves around two of the most prominent strategies for tackling environmental problems. One is technological innovation with a focus on Climate Engineering technologies, mostly Solar Radiation Management (SRM) (Crutzen 2006; Keith 2013). The other is regulatory decision-making under fundamental uncertainty. Research and learning are intimately linked with both strategies, thus playing a connecting role in this dissertation. Methodologically, this thesis takes a theoretical approach, combining modern environmental economics with recent developments in decision theory and the literature on regulation. The first part of this thesis advances the current state of knowledge on technological solutions to environmental problems, taking Climate Engineering technologies as an illustration. The focus here is on the implications of specific strategic conflicts on the incentives to develop SRM technologies with costly RD the latter occurs if one country chooses high levels of SRM and thus imposes an externality on other countries (Weitzman 2012). The second part of this thesis focuses on regulatory decisions under uncertainty for which the standard expected utility framework is inadequate. This may happen if the matter of regulation involves complex processes or novel substances and thus requires a description of knowledge that goes beyond a unique probability distribution formulation. A well-known alternative are multiple prior models (static and dynamic axiomatizations were provided by Gilboa and Schmeidler 1989 and Epstein and Schneider 2003/2007, respectively). The third and fourth paper in this thesis overcome shortcomings in the existing decision-theoretic literature on multiple prior by establishing a consistent notion of the value of information (“Informativeness of Experiments for MEU – A Recursive Definition”) and well-behaved learning dynamics (“Learning Under Ambiguity – A Note on the Belief Dynamics of Epstein and Schneider (2007)”) for maxmin expected utility (MEU) preferences, a well-established ambiguity averse decision rule widely used to model precaution (Vardas and Xepapadeas 2010; Heal and Millner 2013). These decision-theoretical contributions stand for themselves, but also build the ground for the main paper in this part (“Information acquisition under Ambiguity – Why the Precautionary Principle may keep us uninformed”). This paper connects learning and technology choices by focusing on regulatory settings like the approval of a new pesticide in which ambiguous scientific knowledge can be reduced by the regulator by means of (costly) research, for instance with animal testing. In decision-theoretic terms, this paper analyzes active learning under ambiguity and is, to our knowledge, the first model to do so. We find a complex and surprising interplay of the maxmin rule and the research behavior of the regulator: Our results suggest that, despite its notion of precaution, the maxmin rule often leads to an underinvestment in research relative to a standard expected utility regulation, giving rise to a counterintuitive increase in erroneous regulatory decisions (for instance the approval of harmful pesticides). Jointly, the five papers in this thesis contribute to theoretical environmental economics by furthering our knowledge on the role of learning when science is uncertain, on the role of technologies, and on the interplay between technological solutions and uncertainty.