Regression results

TO DETERMINE THE RELATIONSHIP BETWEEN light truck use and

roadway fatalities, we estimated the model using state level data for the four years, 1994 through 1997. There are three equations and one identity. Endogenous variables, or variables determined within the model, are marked. Separate equations for single- and multiple-vehicle fatalities are specified because of the aforementioned concerns over light truck rollover accidents and light truck "aggressiveness" when in accidents with cars.

In Table 1, the right-side exogenous variables in the equations in the left two columns are standard in the highway safety literature. (See, for example, Theodore E. Keeler's "Highway Safety, Economic Behavior, and Driving Enforcement.") Researchers have found motor vehicle fatalities to be inversely related to education levels, seatbelt laws, and drunken driving enforcement. Fatalities have been found to be positively related to miles traveled, the ratio of rural to urban miles traveled, inclement weather, young drivers, suicide, and to the average speed of travel.