The Impact of Incentives on Valuing Health Risks

Recent research has found that health risk values elicited using Discrete Choice Experiments (DCEs) may be inadequately sensitive to the magnitude of the risk reduction, a phenomenon referred to as insensitivity to scope. This paper investigates the use of DCEs to estimate the value of a statistical life (VSL) under different experimental conditions. In particular, we use an experimental design where in one experimental arm we carry out a standard unincentivised DCE as in the existing literature, while in the other experimental arm we carry out an incentivized version of the DCE with real payments. Our findings suggest that the incentive has an impact on the results, in the sense that the VSL estimates are higher in the unincentivised arm of the experiment. However, we find evidence of external insensitivity to scope in both experimental arms and only weak evidence of stronger internal sensitivity to scope in the incentivized arm. Hence, our results suggest that a lack of scope sensitivity is unrelated to the hypothetical nature of the payments in stated-preference valuations of health risks.