There are many situations in which individuals do not expect to find out which outcome will occur. The standard vNM Expected Utility model is inappropriate in these cases, since it does not distinguish between lotteries for which the outcomes are observed by the agent and lotteries for which they are not. This paper provides an axiomatic model which makes this distinction, and which admits preferences for observing the outcome as well as preferences for remaining in doubt. A representation theorem is obtained. Doubt-proneness and doubt-aversion are then defined, and the relation between risk-aversion, caution and doubt attitude is characterized. The model extends the standard vNM framework, but other frameworks, such as nonexpected utility models, can also be extended to allow for preferences for observing the outcomes and preferences for remaining in doubt. A general methodology for extending these well-known frameworks is therefore provided. This framework can accommodate behavioral patterns that are inconsistent with the vNM model, and that have motivated the development of models that differ significantly from the standard vNM framework. In particular, this framework accommodates self-handicapping, in which an agent chooses to impair his own performance. It also admits a status quo bias, without having recourse to framing effects. Several other examples are provided. In one example, voters prefer to remain ignorant, and as the importance of the relevant issues increases, their incentive to acquire information decreases.
