United Nations economists recommend to governments a rough guideline of **2-3 times per capita GDP as a rough threshold per [quality-adjusted life year](https://en.wikipedia.org/wiki/Quality-adjusted_life_year) (say 1 year of perfect health)**, so at a high level that is what can be used as a cost effectiveness benchmark. (via [Updating Cost-Effectiveness — The Curious Resilience of the $50,000-per-QALY Threshold](http://www3.med.unipmn.it/papers/2014/NEJM/2014-08-28_nejm/nejmp1405158.pdf)) >Governments and other prominent actors often use "value of a statistical life" estimates to compare the value of improving health relative to raising incomes. These estimates often imply that a year of healthy life is roughly 2-3x as valuable as a year of doubling someone's income. (via [Approaches to Moral Weights: How GiveWell Compares to Other Actors](https://www.givewell.org/how-we-work/our-criteria/cost-effectiveness/comparing-moral-weights)) But this has many complexities, notably: - Should additional years in life from preventing death be counted or only value from immediate suffering release be counted? - If additional years are counted, this would bias towards young people and away from old people or say even chickens? - I think the suffering should be measured in the moment while it is happening and the suffering related to death can only happen to family members so it should be estimated for them only (say via Time Trade Off method from [[Measures of suffering]]) - Should only suffering of the sufferer be counted or even the associated suffering of family or caretakers be taken into account? - This means for conditions like schizophrenia where family is significantly impacted would count more for cost-effectiveness v/s a condition where family is relatively unimpacted - How should other externalities be counted (for example, someone with reduced suffering ends up increasing the net cost on society as a whole) #inbox