Mutual Sharing or Insurance? An Experimental Study of Decisions at Various Levels of Uncertainty

Decision theory assumes that agents making choices assign subjective probabilities to outcomes, also at choices where information on probabilities is obviously absent. Yet, Skogh and Wu (2005) show that risk averse agents may gain by risk sharing also at unknown (and unassigned) probabilities of losses, as long as the agents presume that the risks are equal. Their Restated diversification theorem is tested by an experiment where the players may lose half their endowments in each of five risky rounds. The probability of loss, and the information about this probability, varies with the rounds. The result supports the hypothesis of beneficial risk sharing at genuine uncertainty. Moreover, the result tentatively supports an evolutionary theory of the development of the insurance industry starting with mutual pooling at uncertainty, turning into insurance priced ex ante when actuarial information is available.