Valuing Mortality Risk in China: Comparing Stated-Preference Estimates From 2005 and 2016

Abstract

We estimate the marginal rate of substitution of income for reduction in current annual mortality risk (the “value per statistical life” or VSL) using stated- preference surveys administered to independent samples of the general population of Chengdu China in 2005 and 2016. We evaluate the quality of estimates by the theoretical criteria that WTP for risk reduction should be strictly positive and nearly proportional to the magnitude of the risk reduction (evaluated by comparing answers between respondents) and test the effect of excluding respondents whose answers violate these criteria. For subsamples of respondents that satisfy the criteria, point estimates of the sensitivity of WTP to risk reduction are consistent with theory and yield estimates of VSL that are two to three times larger than estimated using the full samples. Between 2005 and 2016, estimated VSL increased sharply, from about 22,000 USD in 2005 to 550,000 USD in 2016. Income also increased substantially over this period. Attributing the change in VSL solely to the change in income implies an income elasticity of about 2.5. Our results suggest that estimates of VSL from stated-preference studies in which WTP is not close to proportionate to the stated risk reduction may be biased downward by a factor of two or more, and that VSL is likely to grow rapidly in a population with strong economic growth, which implies that environmental-health, safety, and other policies should become increasingly protective.