Many public policies are designed to extend people's lives and to reduce the risk of death. To allocate government expenditures across these different policies effectively, it is necessary to evaluate the policies from a cost-benefit perspective. Making such cost-benefit calculations requires estimates (in dollar units) of the economic value of life-saving technologies. These estimates, also referred to as the value of a statistical life (VSL), are used extensively in health care to determine whether specific medical treatments or health polices are cost-effective. For these cost-effectiveness calculations to be correct, it is necessary for the VSL estimates to be accurate. Previous estimates of the VSL are based on the observation that jobs with higher levels of fatality risk tend to pay more. This empirical regularity has been interpreted to indicate that workers are paid a wage premium to induce them to accept risky jobs;this premium has been used to estimate the VSL. Recently, however, considerable debate has arisen as to what the true VSL is and whether this approach produces accurate estimates. These studies may not effectively control for confounding factors such as worker-specific attributes such as motivation and ability or job-specific attributes such as comfort and autonomy. Moreover, these approaches estimate the VSL for a """"""""marginal"""""""" worker - one who, at the current wage rates, is indifferent between working in a risky job and a safe job. Estimates of this marginal worker's VSL are not necessarily informative about the benefits of risk reductions for the average American. To address these problems, this project will estimate the VSL using government policies regarding air bags as a """"""""quasi-experiment."""""""" Through the government-mandated phase-in of air bags over the 1990s, the fraction of vehicles with air bags increased from less than 40% to 100% over a few years. The jump in prices that occurs when an air bag was added to an automobile model can be used to estimate the valuations of risk reductions for different individuals in the population. In the early years of the policy, the price of an air bag reflected the valuation of risk reductions for an American whose VSL was higher than average. As air bags became more common, the price decreased, and the new price reflected the valuation of risk reductions for an individual whose VSL was lower than average. Using differences in the phase-in schedule across vehicle types, it is possible to disentangle the effects of the policy from economy-wide factors such as economic growth and increases in public acceptance of the technology. One important limitation of existing VSL estimates is that they are based on the behavior of younger individuals, while the primary beneficiaries of medical expenditures are the elderly. One advantage of the approach used here is that automobile purchases are made by individuals at many different ages. Consequently, it is possible to use these data to measure how the VSL varies across age groups.
The proposed study aims to obtain more accurate estimates than currently exist of the social benefits of policies that increase life expectancy or reduce fatality risk. Economists'estimates of the social benefits of such policies are used extensively by the NIH, the EPA, and other government agencies to conduct cost benefit analyses of life-saving expenditures such as medical treatments, medical research, and environmental regulations. By improving the accuracy of these benefit estimates, this study will contribute toward improving the accuracy of government decision-making in these areas.
