This paper develops a theoretically consistent technique for valuing non-marketed local attributes using compensating income differentials in the absence of housing market data. The individual's indirect utility function is identified with aggregate data describing equilibrium location decisions, and this function is used in place of the unidentified equation describing how housing prices are determined. The model is used to value climate amenities in Brazil, where such data problems are prevalent. Similar problems arise in other developing countries, particularly when one looks outside of the largest cities.
wage-hedonics discrete-choice analysis climate amenity global warming