Allocating Security Expenditures under Knightian Uncertainty: an Info-Gap Approach

In this paper we demonstrate a methodology aimed at coping with resource allocation under Knightian (non-probabilistic) uncertainty by focusing on the example of competing security measures. The results of this application to security resource allocation also allow us to postulate a possible positivist explanation for the way governments are allocating these expenditures today. We explore the determination of the level and nature of government expenditures that affect security in different ways, and demonstrate that it is better to robust- satisfice the citizen’s expected utility rather than to attempt to maximize it. Moreover, our analysis highlights one rationale for heightened spending on one set of defense measures, when there is less reliable information about threats to national security and their consequences.