Welfare calculations in models of the demand for sanitation

Abstract

Empirical and theoretical research on the demand for health inputs has been, and still is, extensive. However, it is not only important to analyse how household choice is affected by alternative and household specific attributes, but it is equally important to analyse the welfare implications of, for example, a change in the price of different inputs. Policymakers and donor agencies involved in sanitation programmes will typically be interested in the welfare implications of different policy alternatives. This paper analyses the welfare implications of a change in the price of certain sanitation related inputs. In addition, a comparison of the welfare estimates derived from the ‘standard’ approach, i.e. the conditional logit model and the more novel approach i.e. the random parameter/mixed logit model is undertaken. The findings suggest that model choice indeed has implications for the results since the welfare estimates from the two models differ quite remarkably.