Defense Decisionmaking

An investigation of the role of uncertainty in certain approaches to defense decisionmaking, of which the so-called offense-conservative and similar modes of analysis are examples. The decisions suggested by such analyses tend to recommend the allocation of more resources for meeting some threats than prudence warrants, at the expense of adequate preparation for others. The common denominator in these analyses is that they attempt to achieve conservative results by using conservative estimates for the uncertain parameters in their models. This report recommends an alternative mode of analysis — that experts furnishing their judgments to the analyst should supply subjective probability distributions of uncertain parameters rather than conservative estimates. These probabilities of effectiveness are an explicit measure of the risk inherent in a course of action and would be the basis of deciding whether the amount of conservatism achieved is prudent. Nine objections to the recommended alternative mode of analysis are discussed.