9. Risk, Uncertainty, and Insurance

Entrepreneurship deals with the inevitable uncertainty of the future. Some forms of uncertainty, however, can be converted into actuarial risk. The distinction between “risk” and “uncertainty” has been developed by Professor Knight.39 “Risk” occurs when an event is a member of a class of a large number of homogeneous events and there is fairly certain knowledge of the frequency of occurrence of this class of events. Thus, a firm may produce bolts and know from long experience that a certain almost fixed proportion of these bolts will be defective, say 1 percent. It will not know whether any given bolt will be defective, but it will know the proportion of the total number defective. This knowledge can convert the percentage of defects into a definite cost of the firm’s operations, especially where enough cases occur within a firm. In other situations, a given loss or hazard may be large and infrequent in relation to a firm’s operations (such as the risk of fire), but over a large number of firms it could be considered as a “measurable” or actuarial risk. In such situations, the firms themselves could pool their risks, or a specialized firm, an “insurance company,” could organize the pooling for them.

The principle of insurance is that firms or individuals are subject to risks which, in the aggregate, form a class of homogeneous cases. Thus, out of a class of a thousand firms, no one firm has any idea whether it will suffer a fire next year or not; but it is fairly well known that ten of them will. In that case, it may be advantageous for each of the firms to “take out insurance,” to pool their risks of loss. Each firm will pay a certain premium, which will go into a pool to compensate those firms which suffer the fires.

As a result of competition, the firm organizing the insurance service will tend to obtain the usual interest income on its investment, no more and no less.

The contrast between risk and uncertainty has been brilliantly analyzed by Ludwig von Mises. Mises has shown that they can be subsumed under the more general categories of “class probability” and “case probability.”40 “Class probability” is the only scientific use of the term “probability,” and is the only form of probability subject to numerical expression.41 In the tangled literature on probability, no one has defined class probability as cogently as Ludwig von Mises:

Class probability means: We know or assume to know, with regard to the problem concerned, everything about the behavior of a whole class of events or phenomena; but about the actual singular events or phenomena we know nothing but that they are elements of this class.42

Insurable risk is an example of class probability. The businessmen knew how many bolts would be defective out of a total number of bolts, but had no knowledge as to which particular bolts would be defective. In life insurance the mortality tables reveal the proportion of mortality of each age group in the population, but they tell nothing about the particular life expectancy of any given individual.

Insurance firms have their problems. As soon as something specific is known about individual cases, firms break down the cases into subaggregates in an effort to maintain homogeneity of classes, i.e., the similarity, as far as is known, of all individual members in the class with respect to the attribute in question. Thus, certain subgroups within one age group may have a higher mortality rate because of their occupation; these will be segregated, and different premiums applied to the two cases. If there were knowledge about differences between subgroups, and insurance firms charged the same premium rate to all, then this would mean that the healthy or “less risky” groups would be subsidizing the riskier. Unless they specifically desire to grant such subsidies, this result will never be maintained in the competitive free market. In the free market each homogeneous group will tend to pay premium rates in proportion to its actuarial risk, plus a sum for interest income and for necessary costs for the insurance firms.

Most uncertainties are uninsurable because they are unique, single cases, and not members of a class. They are unique cases facing each individual or business; they may bear resemblances to other cases, but are not homogeneous with them. Individuals or entrepreneurs know something about the outcome of the particular case, but not everything. As Mises defines it:

Case probability means: We know, with regard to a particular event, some of the factors which determine its outcome; but there are other determining factors about which we know nothing.43

Estimates of future costs, demands, etc., on the part of entrepreneurs are all unique cases of uncertainty, where methods of specific understanding and individual judgment of the situation must apply, rather than objectively measurable or insurable “risk.”

It is not accurate to apply terms like “gambling” or “betting” to situations either of risk or of uncertainty. These terms have unfavorable emotional implications, and for this reason: they refer to situations where new risks or uncertainties are created for the enjoyment of the uncertainties themselves. Gambling on the throw of the dice and betting on horse races are examples of the deliberate creation by the bettor or gambler of new uncertainties which otherwise would not have existed.44 The entrepreneur, on the other hand, is not creating uncertainties for the fun of it. On the contrary, he tries to reduce them as much as possible. The uncertainties he confronts are already inherent in the market situation, indeed in the nature of human action; someone must deal with them, and he is the most skilled or willing candidate. In the same way, an operator of a gambling establishment or of a race track is not creating new risks; he is an entrepreneur trying to judge the situation on the market, and neither a gambler nor a bettor.

Profit and loss are the results of entrepreneurial uncertainty. Actuarial risk is converted into a cost of business operation and is not responsible for profits or losses except in so far as the actuarial estimates are erroneous.

• 39Knight, Risk, Uncertainty, and Profit, pp. 212–55, especially p. 233.
• 40Mises, Human Action, pp. 106–16, which also contains a discussion of the fallacies of the “calculus of probability” as applied to human action.
• 41See Richard von Mises, Probability, Statistics, and Truth (2nd ed.; New York: Macmillan & Co., 1957).
• 42Mises, Human Action, p. 107.
• 43Ibid., p. 110.
• 44There is a distinction between gambling and betting. Gambling refers to wagering on events of class probability, such as throws of dice, where there is no knowledge of the unique event. Betting refers to wagering on unique event about which both parties to the bet know something—such as a horse race or a Presidential election. In either case, however, the wagerer is creating a new risk or uncertainty.