Abstract
This paper summarizes the state-of-the-art for assessing the value of a statistical life (VSL) as a component of the costs of road accidents. It focuses on the most popular approaches for assessing the VSL, with respect to its theoretical foundations, current state-of-research and empirical evidence. Our paper also provides a first (to our knowledge) compendium of results for the VSL based on Stated Choice (SC) methods. Among the analysed alternatives, the willingness-to-pay (WTP) appears to be the leading approach for assessing the VSL and the SC methods represent the current state-of-the-art for determining the WTP for non-market goods. We conclude that the SC approach overcomes some of the most important shortcomings of the alternative approaches and offers a significant flexibility that can be used to address its own limitations. We also identify a significant need for research, as a gap between the methods employed in research (SC methods) and the state-of-the-practice (other methods) has emerged.
Acknowledgements
We gratefully acknowledge the support of BecasChile given by the Chilean Council for Scientific and Technological Research (CONICYT) and the funding by the German Federal Highway Research Institute. The authors would also like to thank Prof. Luis Ignacio Rizzi for his useful comments and insights. All errors are the authors’ sole responsibility
Disclosure Statement
No potential conflict of interest was reported by the authors.
Notes
1. See Landefeld and Seskin (Citation1982) for a good discussion about gross and net losses to society.
2. For even earlier research see Andersson and Treich (Citation2011).
3. Formally, in the context of the WTP analysis, the correct notation is to refer to the value of risk reductions (VRR) and to the WTP for risk reductions (Jones-Lee, Citation1994; Rizzi & Ortúzar, Citation2006b), but in favour of terminological consistency, the term VSL is preferred in this review.
4. Technically this problem is common to every WTP approach, but is severally increased when no other variables provide a context for the decision or a different trade-off to evaluate.
5. This approach is often confused (particularly in marketing) with the conjoint analysis. However, there are significant conceptual differences between both methods (Louviere, Flynn, & Carson, Citation2010). Although both approaches have similarities, especially in relation with the alternatives presented to respondents, both methods differ in the analysis. Conjoint analysis is based purely on mathematical algebraic ranking algorithms (the alternatives are previously ordered by the individuals) and not on solid microeconomic theories, such as the random utility theory (Thurstone, Citation1927; McFadden, Citation1974). Conjoint analysis is not treated in this paper.
6. The way in which travel time savings (especially small ones) should be considered in the utility function is a hot topic nowadays. See for example, Metz (Citation2008).