An analysis of oil supply disruption scenarios

Abstract

This report brings the results of simulations of some oil supply disruptions on the M.I.T. Energy Laboratory Energy Macro Model. This model has previously been used to study the macroeconomic effects of the 1973-74 and 1979 oil price shocks, as well as for policy simula- -tions related to these historical events (Mork and Hall, 1980a, 1980b, 1981). Recent extensions of the model allow it to be used for simulation of possible future oil supply disruptions, such as the loss of oil deliveries from Saudi Arabia or the entire Persian Gulf region.

The extensions go mainly in the direction of more explicit modeling of thedomestic energy sector and of the world oil market. Although very important, these extensions are still in an experimental stage. Richard Gilbert of the University of California at Berkeley provided invaluable help in modeling the domestic energy sector. The short-run functioning of the world oil market was modeled along the lines suggested by Chao and Manne (1980). Sanjay Srivastava provided useful research assistance in preparing and programming the solution algorithm for the model. None of these persons are, however, in any way responsible for the contents of this report.

Three basic disruption scenarios are analyzed, of 3, 10, and 18 million barrels per day (mmbd) on the world level, respectively. Disruptions are modeled as leftward shifts in the world supply curve for oil (OPEC's 'price reaction' curve). All disruptions are assumed to last for one year, but another three years are assumed to be needed to restore the lost capacity. Each disruption scenario is simulated under three different policy assumptions. The first case assumes no new policy. The second case assumes a specific tariff on oil imports, rising from $2.50 in 1981 to a long-run value of $10.00 in 1984. In the third case, an ad valorem tariff is introduced gradually in the same way, from 7.5 percent of the world price in 1981 to 30 percent in 1984. As a last exercise, a 10 mmbd disruption is simulated under the assumption that all relative prices adjust perfectly to clear all markets.

Oil supply disruptions are found to add substantially to inflation during the disruption year and to cause substantial real losses. For a 10 mmbd disruption (the possible loss of Saudi Arabia), 6.4 percentage points are expected to be added to the inflation rate in the disruption year, and the net social cost in real terms is projected as $489 billion in 1980 dollars. An 18 mmbd disruption (the possible loss of the Persian Gulf) is expected to lead to an 8 percent drop in real GNP in the year of the disruption, to add 14 percent to the inflation rate and 6.5 percent to the unemployment rate in the same year, and to incur a net loss of $1,010 billion in 1980 dollars.

The two tariffs are found to lead to very similar results. The ad valorem tariff may hold a slight edge over the specific tariff by reducing transfer the real income to oil exporters during a disruption by slightly more than it increases the loss due to unemployment. Both tariffs are, however, clearly inferior to the alternative of no new policy because of the added losses and inefficiencies in normal periods.

If all prices and wages were free to adjust instantaneously, so that full employment were maintained everywhere during a disruption, then the world oil market would be.much higher and the price increase more than three times as high. This effect of unemployment on oil prices provides an automatic stabilizer in the world market and is an important part of the explanation of the apparent resilience of the U.S. economy to large disturbances in the world oil market.