|V. Cost-Benefit Studies of Helmets and Helmet Laws
By Jonathan Goldstein
Cost-benefit studies attempt to estimate the dollar value of the costs and benefits associated with either (1) the repeal of a helmet law; (2) the non-use of helmets, or (3) the cost of motorcycle accidents. Typically direct (hospital care costs, rehabilative costs and professional fees) and indirect costs (lost earnings/productivity) are estimated. Most studies have concentrated on the costs associated with fatal injuries (Watson et al. (1981), Hartunian (1983)), while others consider both fatal and non-fatal accident costs (Mueller (1980) and Rivara (1988)). These latter studies also estimate the percentage of direct costs paid for by public funds. Typical findings are that mandatory helmet use laws could save between 500- 600 lives , and between $60-$180 million annually. In addition it is estimated that between 50%-63% of direct costs are paid for by public funds.
There are two fundamental problems with these studies. First , they rely on the overstated (biased) statistical estimates of helmet effectiveness from correlation and regression analyses , reviewed above, to establish the "benefit" of wearing a helmet. Clearly, this approach substantial overstates the costs associated with the non-use of a helmet because it inappropriately assigns the effects of speed and alcohol to the non-use of a helmet. Second , these studies do not consider all of the costs associated with the use of helmets or with mandatory helmet laws. Typically the only cost considered is the purchase price of the helmet. The costs associated with the neck injury inducing aspects of helmet use (established in Goldstein (1986)) and the costs generated by risk compensating behavior (discussed in next section) are never considered. Thus, the benefits associated with helmet use are overstated and the costs are understated. It is conceivable that the costs of helmet use could offset or outweigh the benefits, but this never surfaces in the literature.
1. Watson, G. S., Zador, P. L., and Wilkes, A. W. (1981). "Helmet Use , Helmet Use Laws, and Motorcycle Fatalities." American Journal of Public Health, 71:3.
This article is a response to criticisms of Watson et al. (1980) reviewed above. In this paper, the authors rely on alternative data sets to corroborate their original finding that on average helmet law repeal increases fatalities by 38%. In addition they estimate the cost of helmet law repeal or enactment in terms of additional lives lost or saved. Using before-after correlations and comparisons of helmeted-non helmeted accident victims (the type of studies criticized in Section I) Watson et al. find "support" for their original finding. They conclude that: (1) helmet use in accident situations could reduce unhelmeted rider fatalities by 52%; and (2) that repeal of mandatory helmet use laws increase fatalities by 40% while enactment of such laws decrease fatalities by 30%. Using the latter result, they estimate the cost of repealing all existing helmet laws as an additional 1,100 fatalities and the benefits of all states enacting laws as a reduction of 600 fatalities.
As argued above , this cost-benefit analysis is based on estimates of helmet effectiveness that are overstated and fails to consider the neck injury inducing effects of helmets and the risk compensating behavioral responses as costs of helmet legislation.
2. Mueller, A. (1980). "Evaluation of the Costs and Benefits of Motorcycle Helmet Laws." American Journal of Public Health, 70:6.
Relying on the estimates of helmet effectiveness generated from correlation analyzes of accident victim data from Colorado, Oklahoma and South Dakota (some of which are reviewed above), the author compares the costs of motorcycle helmets with the medical costs averted due to helmet use. The author concludes that $61 million could be saved annually if all motorcyclists were to use helmets. The author limits his analysis to a comparison of the cost of helmets and the cost of medical care expenses averted. Lost lives and lost productivity are not considered. The medical care benefit of helmet use is based on estimates of (1) the probability of an accident; (2) the effectiveness of helmets; (3) medical care expenses for injuries of different severity class; and (4) the percentage increase in helmet use resulting from mandatory helmet use.
This cost-benefit analysis is subject to the two main criticisms advanced above. In addition the author recognizes these problems: "the cost-benefit analysis . . . assumes that differences in injury severity between helmeted and unhelmeted riders is due entirely to the effectiveness of helmets. It is conceivable, however, that part or all of the difference is due to confounding factors. For instance, if helmeted motorcyclists travel at lower speeds... then they would be less likely to sustain injuries..."
3. Hartunian, N. S., Smart, C. N., Willemain, T. R., and Zador, P. L. (1983). "The Economics of Safety Deregulation: Lives and Dollars Lost Due to Repeal of Motorcycle Helmet Laws." Journal of Health Politics, Policy and Law, 8:1, Spring.
Following the approach of Watson et al. (1980) (reviewed above), the authors estimate the number of excess deaths attributable to helmet law repeal and the associated costs to society. No attempt was made to estimate excess non-fatal injuries and costs. The authors findings indicate that 516 excess deaths (24% of all fatalities) occurred in 1980 in the 28 states with no helmet laws. The associated economic costs were estimated to be at least $180 million. Economic cost estimates include direct costs -- hospital, doctor, legal, and funeral costs -- and indirect costs -- cost of foregone earnings.
The estimates in this study are subject to the above criticisms advanced of the statistical methods employed by Watson et al. (1980). In addition these estimates fail to consider any of the cost inducing effects (increased neck injuries) of helmet use. Finally, the portions of the statistical analysis that differ from Watson et al. (1980) are subject to the same criticisms of the other regression studies analyzed in Section III above. Thus, this study produces cost estimates of helmet law repeal that are systematically overstated. It is conceivable that correctly generated estimates could show no costs or net benefits associated with helmet law repeal.
4. Rivara, F. P., Dicker, B. G., Bergman, A. B., Dacey, R., and Herman, C. (1988). Journal of the American Medical Association. 260: 221-223.
This study estimates the direct and indirect costs and the percentage of direct costs paid by public funds associated with 105 motorcycle accidents at a level I trauma center in Seattle. The study finds that average direct costs per patient was $25,764 and that 63% of these costs were paid by public funds. While the study does not compare the costs incurred by helmeted riders to those incurred by non-helmeted riders, it concludes on the basis of previous helmet effectiveness studies that mandatory helmet use could dramatically reduce the direct and indirect costs and thus the costs to society.
While the study calculates direct, indirect and societal costs in a fairly standard manner, the results are unrepresentative of the costs associated with the average motorcycle accident and the
conclusion that mandatory helmet use could dramatically reduce these costs is unfounded.
The study analyzes atypical motorcycle accidents. The types of accidents sent to a level I trauma center involve the most seriously injured patients and thus dramatically biases estimates of average direct, indirect, and societal costs in an upward direction. In addition, the percentage of direct costs paid for by public funds is also overestimated. If the average (less severe) accident is considered, excessive bills that go beyond the economic means of accident victims are less likely to occur and thus accident victims are less likely to rely on public assistance.
Finally, the inference that mandatory motorcycle helmet use could dramatically reduce such costs is unfounded. This inference is based on the results of helmet effectiveness studies that systematically overstate the effectiveness of helmet use and that totally ignore the negative impacts (costs) of helmets which occurs through an increase in neck injuries, potential increase in the probability of a motorcycle accident and risk compensating behavior.
© Copyright Jonathan P. Goldstein Ph.D. 1986. All Rights Reserved.