The econometric literature on program-evaluation and optimal treatment-choice takes functionals of outcome-distributions as `social-welfare’ and ignores program-impacts on unobserved utilities, whereas the utility-based welfare-analysis tradition in public-finance ignores unobserved heterogeneity in individual preferences. This paper reconciles the econometric and public-finance approaches to welfare-analysis in the practically important setting of discrete-choice. We show that under unrestricted preference-heterogeneity and income-effects, the distribution of individual indirect-utility is nonparametrically identified from average demand. This enables cost-benefit analysis of non-marginal policy-interventions and their optimal targeting based on planners’ re-distributional preferences. Our methods are illustrated via empirical analyses of an experimental and an observational dataset.