This paper contributes to the literature on economic aspects of highway safety in four ways. First, it is based on a county-level data set for the United States for 1970 and 1980, affording more than 2.600 observations each year in a consistent panel, many more observations than previous studies have used. Second, this study uses different and more appropriate estimation procedures than most previous studies allowing for the count nature of highway fatalities and correcting for omitted-variable bias otherwise possible in cross-section analysis. Third, the model used here controls for more variables in a single model than previous studies. Finally, this study allows for important differences in the estimated coefficients between urban and rural driving environments. The first section of this paper is concerned with appropriate specification of the equations to be estimated and the data set used. The second section considers issues of estimation. The third section presents the results, and the forth section sets forth conclusions. (Author/publisher)
Abstract