from Health Economics at http://bit.ly/2v7v4Wo on July 30, 2017 at 09:31AM
We applied Smeed fs Law to Japanese prefectural data from between 1988 and 2016. We found that the coefficient for the number of vehicles was stable over the estimation period, but that the constant term decreased gradually. We decomposed fatalities per capita into fatalities per accidents and accidents per capita, and applied regression equations to the data. We conclude the following from this study. First, the relationship between fatalities per capita and the number of registered vehicles per capita was stable, which is consistent with Smeed fs Law. Second, the effects of technological advances have changed the estimated coefficients for time dummies. The role of hospitals may be difficult to incorporate into Smeed fs Law because of the complicated relationship between the distance to hospital and fatalities per capita.