The above review of safety regulation in Japan illustrates the recommendations of economic analysis on how an authority should be organized, and how such prescriptions are justified. But economic analysis also focuses on regulatory goals and instruments. In this section we shall draw on the examples of regulation in the United States and France. Naturally they both have their shortcomings and they are very different, yet both may be taken as good models.

Let us start with a reminder of the economic principles guiding the setting of goals and the choice of instruments. Economics is useful for discarding hollow slogans such as ‘nuclear safety is our absolute priority’ or ‘nuclear safety is priceless’. Catchphrases of this sort suggest that there is not, or there should not be, any limit to safety efforts.

The first limit established by economic theory relates to determination of the optimal level of safety efforts on the basis of marginal costs. The incomplete nature of data reduces the practical value of this response to the question of when safe is safe enough. Very little is known about the marginal costs of safety efforts and even greater uncertainty affects the marginal costs of avoided damages. However, this is not specific to nuclear accidents. Economists have difficulty determining the optimal level of carbon-dioxide emissions or safety on oil-rigs. It is equally tricky to estimate the right discount rate, or put a value on loss of human life or damage to the natural environment.