Interactions among transportation fuel substitution, vehicle quantity growth, and national economic growth

Hypotheses concerning the micro- and macroeconomic effects of historically important fuel switches in the vehicles operating on the dominant U.S. transportation modes are statistically tested using nineteenth and twentieth century data. Estimates of annual rates of change of fuel switching are constructed using logistic curve models fitted to often sparse data on fuel use. The estimated annual rate of loss of market share of an old fuel is then shown to be positively and consistently correlated with five-year averages of declines in the rate of growth of the fleets of vehicles affected by the ongoing fuel switch. Other statistical tests show that five-year average changes in the rate of growth of the quantity of vehicles positively correlate with five-year average changes in the rate of growth of macroeconomic activity when the affected vehicles are directly (through production and sales) responsible for a large share of macroeconomic activity (real GNP). The vehicle types shown to have this effect are locomotives from 1885-1915 and automobiles thereafter. The third set of tests supports an interpretation that the indirect effects of fuel switches, acting through their effects on the costs of transporting factors of production (both labor and physical inputs), have a significant and consistent GNP effect throughout the 1880-1980 period, even during the WWII-influenced period when the test for direct effects of automobiles fails. In case study discussion of each fuel switch it is shown that a pronounced drop in GNP growth occurs at the time when the important fuel switches are most rapidly being accomplished. Results of the statistical and case study analyses are compared to the author’s prior research on vehicle engine innovation. The timing of peak rates of fuel switching and vehicle engine innovation are graphically shown to coincide for locomotives and automobiles from the 1880s to the 1980s. Further, the peak rates of fuel switching are graphically shown to occur when current real GNP relative to trend real GNP is declining and/or below the trend.