Traffic accidents are responsible for the greatest number of deaths each year for many age groups. At present, authorities rely heavily on policing and prosecutions to control accidents. The authors of this work examine the effectiveness of these and other techniques, and suggest alternatives that may provide better results.They particularly favour an epidemilogical approach that takes driver conduct as a given and looks for other ways to control the frequency and severity of accidents. They examine the use of rewards to encourage good driving and the use of licensing to control the exposure of high-risk drivers. The deterrent effect of civil liability and the question of no-fault insurance are also considered, as are various methods used to control drinking and driving.The authors conclude by asking for greater evaluation of the interventions used. Traffic safety research, they argue, has barely begun to confront the central policy issue: how can society get the greatest payoff from the marginal dollar spent to prevent accidents?
