In health technology assessment (HTA), “value for money” is frequently conceptualized as incremental cost-effectiveness (CE), with effectiveness measured as (quality-adjusted) life years gained (LYG). Commonly used CE thresholds have been subject to controversial debate. We reviewed and analyzed the worldwide economic literature on the value of a statistical life (VSL), which reported empirical estimates during the last 25 years.

We conducted an extended systematic literature search in the EconBiz and EconLit databases, spanning the period from January 1995 to December 2020. We transformed VSL data into the implied values of a statistical life year (VSLY) and analyzed the results by regional origin, elicitation method, and annual gross domestic product (GDP) per capita. Then, we performed a regression analysis using the ordinary least squares (OLS) model after log-transforming the VSLY estimates.

We identified 156 empirical economic studies with 169 unique VSL estimates. Overall, the median VSLY was EUR168,367 (mean, EUR256,701) or 6.3 times annual GDP/capita. The median VSLY (expressed as multiples of GDP/capita) differed by regional origin (North America: EUR288,994; 7.2; Europe: EUR168,367, 5.2; Asia: EUR45,260, 4.5) and by elicitation method (revealed preference/wage risk studies: EUR274,625, 9.1; stated preference/contingent valuation studies: EUR113,246, 4.4; stated preference/discrete choice experiments: EUR178,130, 5.2). Regression analyses confirmed that studies with data originating from North America resulted in significantly higher VSLY estimates. The difference remained statistically significant after adjusting for GDP/capita.

The empirical economic literature, while showing heterogeneity by elicitation method, suggests that the adoption of a preference-based or “demand-side” perspective leads to willingness-to-pay estimates for a LYG that exceed commonly used CE thresholds in the context of HTA.

Among economists, there is widespread agreement that the monetary valuation of health gains should reflect the preferences of those who will be affected by resource allocation decisions. In the context of Health Technology Assessments (HTAs), this view implies a need for reliable empirical estimates of the value of statistical life year (VSLY), which should provide a useful point of reference for cost benefit analyses.

We conducted a systematic review of the literature on the economic value of a statistical life (VSL). We searched in the EconBiz and EconLit databases for studies, which reported VSL estimates based on original research and were published between 1995 and 2015. We classified studies by methodology, that is, revealed preference (RP) or stated preference (SP; that is, CV, contingent valuation, or DCE (discrete choice experiment) approach, and by regional origin of data. We transformed VSL estimates into VSLY expressed in year 2014 Euros, using life expectancy tables for the populations studied, a real discount rate of 3 percent, national Consumer Price Indices for inflating, and purchasing power parities for currency conversion. In addition, we calculated ratios of VSLY to gross domestic product (GDP) per capita.

Our search yielded 120 studies appropriate for inclusion. From these, we extracted a total of 132 VSL estimates (RP, n = 60; SP, n = 72). The median VSLY was 6.4 times GDP/capita. Transformed into Euro (2014), the median VSLY was EUR165,000 (mean, EUR217,000). We found significant differences by regional source of data (North American, median EUR272,000; European, EUR158,000) and by method (RP, EUR241,000; SP: CV, EUR117,000; DCE, EUR187,000). VSLY estimates were sensitive to discount rate.

Our data indicate that VSLY estimates based on empirical data exceed benchmarks commonly used in the context of HTAs. However, inter-study variability, methodological limitations, and normative considerations, all suggest to exercise caution before translating this observation into actual policy.

Evaluation of “value for money” is an important component of Health Technology Assessments (HTAs). It is often conceptualized as “cost effectiveness” or cost per (quality-adjusted) life year gained. Whether used in isolation or alongside further drivers of social value (such as priority for younger or more severely impaired patient groups, or for access to effective treatment, even if costly), for example within a multi-criteria decision analysis framework, any reference “value of a statistical life year” (VSLY) should be supported by empirical data capturing the preferences of the population(s) in question. Here we report results based on a systematic review of relevant European economic studies, which were published during the last two decades, that is, from 1995 to 2015.

Our literature search (using the EconBiz and EconLit databases, supplemented by an analysis of relevant reviews) identified fourty-one European studies providing original data, yielding a total of fourty-eight average estimates for the value of a statistical life (VSL, or fatality prevented). We classified studies by methodology, for example, revealed preference (RP) or stated preference (contingent valuation, CV; discrete choice experiment, DCE) approach. We transformed VSL estimates into VSLY expressed in year 2014 Euros, using the life expectancy of the populations studied, a real discount rate of 3 percent, the national Consumer Price Index (CPI) for inflating, and purchasing power parities for currency conversion. We calculated confidence intervals by means of nonparametric bootstrapping.

The median VSLY was EUR158,000 (for RP studies, EUR218,000; DCE, EUR188,000; CV, EUR143,000); we did not identify studies using the human capital approach. Our VSLY estimates showed large heterogeneity, both by methodology and regional origin; thus the differences that we observed did not reach statistical significance.

Our results suggest that the empirical willingness-to-pay for a statistical life year might be substantially higher than benchmarks currently used by the international HTA community.