The True Value of, the Value of a Statistical Life (VSL), in Benefit-Cost Analysis: How Might the VSL Be Different from What We Have Been Told?

Abstract

The Value of a Statistical Life (VSL) serves as the definitive metric for allocative efficiency in Health Economics and Benefit-Cost Analysis (BCA). However, the production and application of VSL estimates have become stratified into four distinct groups-econometric producers, practitioner appliers, scholarly supporters, and scholarly objectors-that largely operate in isolation. This paper argues that this lack of cross-disciplinary discourse has allowed foundational assumptions to persist without adequate scrutiny, compromising the scientific validity of the VSL. This study provides a critical methodological re-evaluation of the hedonic wage models used to derive VSL. The central contribution is a microeconomic examination of labor demand elasticity. Standard VSL regression analyses presume that observed wage differentials between "safe" and "risky" sectors accurately reflect the worker's risk premium. This paper demonstrates, through a partial equilibrium framework, that this equality holds true only under the highly specific condition of perfectly inelastic labor demand. By relaxing this assumption to reflect the reality of elastic labor demand, the analysis shows that observed wage differentials systematically underestimate the true compensation workers require for risk. Furthermore, the paper identifies an "information paradox" in current VSL theory. Standard models assume workers possess perfect knowledge of occupational risks to price them into their wages . Yet, regulatory agencies commission epidemiologists and engineers to identify these same risks, premised on the fact that they are unknown to the layperson. If workers already possessed the knowledge assumed by VSL producers, such expert analysis would be redundant. The analysis also challenges the "job-desperation effect," where low-income workers with limited bargaining power accept high-risk positions at lower wages. This "missing variable" of risk awareness occasionally results in data showing negative risk premiums that defy economic intuition. The paper further critiques the assumption of risk neutrality, arguing that extrapolating small risk probabilities linearly to a full statistical life creates logical inconsistencies. If preferences were truly linear, individuals would logically accept wagers involving high probabilities of death for proportional monetary gains-a conclusion that contradicts observed human behavior. Additionally, the paper introduces the "Regulatory Indifference Paradox." If standard VSL assumptions were strictly accurate-specifically that wages perfectly compensate for risk-then workers, on average, should be economically indifferent to safety regulations. According to the model, any safety improvement would be offset by an exactly proportional reduction in wages, neutralizing the net benefit to the worker. Finally, the paper addresses the issue of false precision. It argues that the "statistical" label is a semantic shield used to deflect ethical scrutiny regarding life valuation. Moreover, the reliance on calculating the median of widely divergent study results creates an illusion of scientific consensus that masks the unreliability of the underlying estimates. The paper concludes by proposing a framework that integrates insights from behavioral economics to restore scientific integrity to the valuation of life.