This paper discusses the obstacles that hamper robust estimations of the cost of nuclear accidents. From an economic standpoint, risks of accidents are often quantified by the assessment of an expected cost; that is the product of a monetary loss by its probability of occurrence. In the case of nuclear power, this definition is inadapted. Estimating the probability of a nuclear disaster is subject to high uncertainties, and so is the assessment of its monetary equivalent. This paper first discusses the two specific challenges in estimating the probabilities of nuclear accidents: these accidents are too sparse to identify frequencies of occurrence with probabilities; they are also dreadful, which makes risk-aversion a complex phenomenon. This paper then focuses on the assessment of nuclear damage. The large discrepancies exhibited in the existing literature arise from three sources: the scope of the assessments, the conflict between the use of past data or PSA studies in the assessment of radiation damage, and the methods that are used to quantify non-monetary welfare losses.

No abstract

No abstract

No abstract