Abstract
This dissertation considers a country that faces uncertain and intermittent disruptions of a single import. It focuses on the possibility of stockpiling the good to hedge against the adverse effects of interrupted supplies. The following normative questions are addressed: How much of the good should be stockpiled in the light of the perceived threat of supply disruptions? How should inventories be used (priced) during a supply disruption? How much worse off is the country when it is subject to interrupted supplies? The optimal allocation of emergency reserves is contrasted with the market allocation. Certain specific conditions have to be specified if the market economy is to be able to replicate the optimal stockpiling and stock withdrawal policies. We investigate the bias that arises in an actual market economy lacking a complete set of risk and future markets. Obviously, an omniscient plannr can do better than an actual market economy. But what about the "real world" planner? To make a fair test of the welfare properties of private stockpiling we restrict the planner to operate only on the markets that do exist. That is, we place the planner and the market in a symmetric position with respect to the transaction technology. When we discuss practical planning problems an alternative restriction is imposed on the planner. In dealing with the problem of allocating a government controlled strategic reserve we consider the limitations on the information available to the planner about the demand conditions for the good.