This paper offers a new approach to the estimation of marginal net perceived benefits from public goods, It is designed to utilize information from survey data which consist of answers to the question: Should more, less or the same amount of tax money be spent on program X? Widespread availability of such data, and their relative underutilization, motivated this research,. Considerable disparities are found in perceived benefits over various age groups, educational attainments and geographical locations, emphasizing that deciding which public goods to supply and how much of them has important distributive, as well as allocative, implications.