arrow checkmark code cross email facebook magnifier pdf phone plus twitter user youtube


Publicado en: Documento de trabajo

The Value of a Statistical Life for Risk Averse and Risk Seeking Individuals

Marcela Parada.

Documento de Trabajo IE-PUC, N° 501, 2017.

Abstract: This paper estimates the Value of a Statistical Life (VSL) under the hedonic wage method accounting for individual risk preferences. This estimation is based on a unusual approach; it empirically uses two approaches to account for risk preferences in the estimation of the wage-risk trade-off and VSL estimation. First, risk aversion is directly measured using survey-measures of preferences toward hypothetical gambles, and second observed individual behaviors, that may proxy for risk preferences such as smoking status are used. The empirical results and the used of observed measures of risk aversion are reconciled with a theoretical model of economic behavior that predicts how the wage-risk trade-off changes as risk aversion differ across individuals. It also captures the effect of risk aversion across the wage distribution using quantile regressions to test for the modeling assumption that safety is a normal good. The VSL estimates range between 0.81 and 9.16 million dollars. The results using smoking behavior as proxies for risk attitudes are consistent with previous findings. However, directly measuring risk aversion corrects the wage-risk trade-off estimation bias in the opposite directions than when using smoking status. This suggests that, consistent with the literature that connects smoking behavior with labor market outcomes, smoking status could be capturing productivity effects due to poor health in addition to purely risk preferences. This paper contributes to the literature by mitigating potential bias that may appear due to
individual rusk preferences using a novel approach.

Keywords: value of a statistical life, risk attitudes, elicited risk aversion, quantile regression.
JEL Classification: J17, J28, D81, C21


Descargar documento