Provided that linearity assumption is reasonable. As I understand, the regression equation your cited was determined from historical data, but the projections go beyond the range of the historical data i.e. the price did not go beyond $2.30 at the pump or $50 crude. The argument that the 1970s prices were higher in inflation adjusted $$ is not applicable because it was basically a one -time spike, an outlier if you will, whose impact on the regression coefficient outweighed by the multitude of observations within the "normal" range - its influence shows mainly as lowering the R-squared or percent of variance in the price at the pump explained by variations in the prices of crude.
However, it is not too farfetched to believe that when crude prices continue going up, the coefficient will change i.e. linearity assumption will be violated. Stated differently, the coefficient 0.03211 (3.2 cents pump price increase for every $1 crude price increase) may keep growing as the prices of crude keep increasing.
There are several possible reasons:
- supply/demand structure is different than that used to calculate the equation in question; that is to say, supply is about to hit a ceiling - a condition that did not obtain in the past, while the demand keeps growing;
- increasing demand may push producers to explore less 'economical' fields, which will increase the cost of crude production
- increasing demand may require new technologies which also may push the production up
- uncertainty of supply may trigger hoarding behavior, which will further push the demand and thus prices of crude and at the pump as well.
Consequently, the 3.2 cent coefficient may no longer hold if the crude prices keep rising beyond $55.
Wojtek