U.S. Gasoline Demand: What Next?

Predicting the demand for motor gasoline over the last ten years has proven a most frustrating experience. Up until 1973, industry analysts felt considerable assurance in applying historical growth rates that averaged approximately 5 percent per year. Who in 1973 would have predicted that gasoline consumption in 1981 would fall below 1973 levels? For example, in 1973, Shell Oil Company predicted annual growth of 4.9 percent per year. Their forecasted value for the year 1981 exceeded the actual level by 42.7 percent (Shell, 1973). Unfortunately, errors of this magnitude are not as benign as predicting the point spread in a pro football game. To the contrary, both private and public policy decisions depend on the accuracy of such forecasts. Because of the importance of gasoline as the major refinery product, refinery expansion plans and retail marketing strategies are conditioned on such forecasts. Similarly, public policy decisions regarding auto efficiency standards, auto pollution controls, and oil import policy depend on gasoline demand forecasts. Current forecasts tend to be extremely pessimistic with respect to gasoline demand in the 1980s. Wharton Econometric Forecasting Associates predicts a 14.7 percent decline in motor-vehicle fuel demand over the decade of the 1980s; Exxon's Energy Outlook predicts a 15.6 percent decline over the decade. Is such pessimism warranted? Are there key assumptions, which if changed, could produce a substantially different picture?