Working Paper: NBER ID: w29185
Authors: H. Spencer Banzhaf
Abstract: The Value of Statistical Life (VSL) is arguably the most important number in benefit-cost analyses of environmental, health, and transportation policies. However, agencies have used a wide range of VSL values. One reason may be the embarrassment of riches when it comes to VSL studies. While meta-analysis is a standard way to synthesize information across studies, we now have multiple competing meta-analyses and reviews. Thus, to analysts, picking one such meta-analysis may feel as hard as picking a single "best study." This paper responds by taking the meta-analysis another step, estimating a meta-analysis (or mixture distribution) of six meta-analyses. The baseline model yields a central VSL of $7.0m, with a 90% confidence interval of $2.4m to $11.2m. The provided code allows users to easily change subjective weights on the studies, add new studies, or change adjustments for income, inflation, and latency.
Keywords: Value of Statistical Life; Meta-Analysis; Benefit-Cost Analysis; Environmental Health; Transportation Policies
JEL Codes: I12; I18; J17; J31; K32; Q51
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| choice of VSL (J17) | outcomes of benefit-cost analyses (H43) |
| VSL estimates (J17) | policy evaluations (H43) |
| VSL choice (J17) | monetized benefits from regulations (G18) |
| variation in VSL estimates (J17) | conclusions regarding benefits of policies (J18) |
| choice of VSL (J17) | mortality risk assessments (J17) |