Abstract
In 1975 the U.S. Department of Transport issued a forecast of the 1985 U.S. traffic safety situation. The forecast was based on projected 1985 values of several parameters. This forecast proved to be overly pessimistic, overestimating fatalities by 65%. Post-hoc predictions from three alternative models provided better fit to the 1985 data. These models included the degree of motorization model, the economic model and the model of the author.