Lessons from Applying Value of Statistical Life and Alternate Methods to Benefit–Cost Analysis to Inform Development Spending

Estimating value of statistical life (VSL) is an important input to many benefit-cost analysis (BCA) approaches, but for many low- and middle-income countries, there are limited or no data estimating VSL. Current guidance relies on extrapolation of results from high-income settings, which may be unreliable, leading to low confidence applying VSL. During 2019, we surveyed 1,820 low-income individuals (average consumption per capita USD329) across four diverse regions in Ghana and Kenya, to inform recommendations about effective spending in the development sector. We elicited VSL using a stated-preference approach, capturing the willingness-to-pay to reduce the risk of death for themselves and their children. Additionally, we conducted multiple “policy choice experiments” (PCEs) in which we asked respondents to choose, from the perspective of a decision-maker, between programs that save lives of different ages, and save lives and provide cash transfers. VSL estimates for this population fell in the range of USD66,795–USD90,453 (PPP-adjusted). We found similar results in the PCE but uncovered much stronger preferences for saving younger lives. Overall, our results suggest that VSL in low-income countries may be higher than estimates based on extrapolations from wealthy countries and that within these communities, policymakers should place more weight on saving the lives of young children. We also explore methodological learnings about how to apply and collect data for BCA in particularly low-income, low-education settings. We find that through careful training and gatekeeping, it is feasible to elicit complicated preferences in this population, and both approaches have their benefits and drawbacks.