Risk and uncertainty

The collective principle for risks outlined in Chapter 2 can be applied to many situations inside and outside commerce. But in some circumstances no immediate relevant collective springs to mind. This chapter looks at such situations where probability estimates cannot be made by the collective approach on a basis that would necessarily command overall agreement. Such situations are sometimes labelled ‘uncertainty’ and differentiated from ‘risk’ situations.

The separation of uncertainty from risk is a frequent practice in industrial circles. Examples cited in this context are political uncertainties such as nationalization, economic uncertainties such as forthcoming rates of inflation, or changes in the rates of interest. Many statisticians argue cogently that there is no real distinction between risk and uncertainty as defined in this way. The distinction really being made, it is argued, is between repeatable and non-repeatable events. Thus games of chance such as roulette are repeatable, just as the actuary of an assurance company regards death as a repeatable event, in the sense that a large population is at risk and alternative events ‘death’ or ‘no death’ are repeated for each person in the population each year. Although one cannot be dogmatic about a single person dying, one can be reasonably specific about the number of persons dying from among a large group in a specified period. But the outcome, ‘the next President of the United States will be a woman’ is not of the same category. To count up the number of eligible voters who are women (W) and men (M) respectively and express the chance as the proportion W/(M+W)