What is a life worth? Robustness of VSL values from contingent valuation surveys

The value of a statistical life (VSL) is a key input for estimating the benefits of policies that save lives. Several recent studies have obtained estimates of the VSL from contingent valuation surveys, i.e., by asking people to say how much they would pay to reduce their risk of dying. This article examines statistical factors that may influence the estimates of the VSL obtained from such surveys. We examine the importance of distributional assumptions, the choice of the welfare statistics of interest, the procedure for computing them, outliers, undesirable response effects, and internal validity of the willingness-to-pay (WTP) responses. We illustrate the importance of these factors using dichotomous-choice and open-ended WTP data from four recent contingent valuation surveys.