Citation

Blattenberger G, Fowles R, Loeb PD. Res. Transp. Econ. 2013; 43(1): 112-122.

Copyright

(Copyright © 2013, Elsevier Publishing)

DOI

10.1016/j.retrec.2012.12.004

PMID

unavailable

Abstract

Motor vehicle crashes continue to result in large numbers of fatalities each year and represent the leading cause of death for young persons. In 2006, for example, there were over 42,700 fatalities associated with these crashes. Understanding the causes of these crashes and methods to reduce them continues to be of great interest to economists, public health officials, and policy makers. We present in this paper statistical models using a rich set of panel data covering the period 1980 to 2007 by state and the District of Columbia. Our choice of variables is based on an extensive literature highlighting the importance of policy, safety, demographic, and economic determinants of fatality rates.

The estimation techniques used in this paper takes cognizance that standard econometric inference focuses on parameter uncertainty. Models are estimated conditional on the assumption that the model to be estimated and reported is the "true" model. Tests are then made on a multitude of alternative models, each sequentially assumed to be the "true" model. Model uncertainty is manifested in this procedure, but it is often ignored in practice. Recent Bayesian statistical methods speak directly to the issue of both model choice and variable selection. This paper utilizes three Bayesian techniques: Extreme Bounds Analysis, Bayesian Model Averaging, and Stochastic Search Variable Selection to address model and parameter uncertainty in models estimating the determinants of motor vehicle crash fatalities.