Non-parametric well-being comparisons

We study the problem of making interpersonal well-being comparisons when individuals have heterogeneous preferences. present a robust, but incomplete, criterion for that states one individual is better off than another if intersection between extended upper contour set better-off and lower worse-off empty. implement in consumption-health space using an online survey 2260 respondents United States to investigate how complete resulting comparison is. To chart sets respondents, we propose new “adaptive bisectional dichotomous choice” (ABDC) procedure based on limited number choices some mild non-parametric assumptions about find fifth all pairs can be ranked with criterion.