Alternative Types of Ambiguity and their Effects on the Probabilistic Properties and Tail Risks of Environmental-Policy Variables

The concept of ambiguity with respect to decision making about climate change has recently attracted a lot of research interest. The standard approach for introducing ambiguity into this framework is to assume that the decision maker (DM) exhibits ambiguity aversion, with the latter being represented by axioms on DMs preferences different than Savageâ��s (sure-thing principle). As a result, DM is deprived of the property of probabilistic sophistication, since she is faced with either multiple prior probability functions, or a single but incoherent one (capacity). This paper approaches the issue of ambiguity with respect to climate change from a different perspective. In particular, we assume that ambiguity does exists but it does not affect the formation of DMs prior probability function. Instead, it a�¤ects the formation of her posterior probability function. Specifically, we assume that there are n experts, who supply DM with probabilistic input. Hence, although DM has a well defined prior (formed before any expert information on objective probabilities has arrived), she cannot decide which piece of information should conditionalize upon (defer to). We refer to this type of ambiguity as "deferential ambiguity" and show that it affects both DM and the experts. We also introduce a second type of ambiguity, which is solely born by the experts. This type of ambiguity stems from the experts potential inability to discern DMs preferences. This ambiguity is referred to as "preferential ambiguity" in the paper. The main objective of the paper is to analyze the possible interactions between the two types of ambiguity mentioned above and to assess their impact on the probabilistic properties (in particular, tail risks) of environmental-policy variables.