STOCKPILE policies represent one of the few non-military options
available to offset the political turbulence of the Persian Gulf region.
Two key uncertainties affect stockpile policies: (a) the influence of U.S.
(and in general, OECD) import demands upon the OPEC price; and (b)
the duration and the severity of oil disruptions induced by political
disturbances among the OPEC nations.
This paper will analyze these issues using models in which U.S. policy
choices are restricted to just two instruments-stockpiles and disruption
"tariffs." There are many other policy options that resemble these two
instruments. Oil stockpiles could be held either in the public or in the
private sector. Similarly, there are many potential policies that resemble
a tariff. Quotas, subsidies, gasoline rationing and energy efficiency stand-
ards would all tend to reduce oil imports, but differ from a tariff in terms
of their economic efficiency and equity.
For analyzing these problems, Balas [1979] has used a two-person
gaming model. This paper is based upon a somewhat different view-
point-a probabilistic view of disruptions analyzed by applying the tools
of dynamic programming. Dynamic programming is particularly appro-
priate for those sequential decision analyses in which inventory levels and other "state variables" are employed to summarize the history of a
probabilistic process. Teisberg [1981] has employed similar approach.
Here however, we adopt a macroeconomic framework based on the
maximization of the expected utility of consumption. The numerical
assumptions are similar to those employed in Plummer [1982].
