Incentives and income taxation: the implementation of individual revenue requirement functions

Abstract

Consider an economy where a fixed amount of revenue is to be raised, perhaps to finance a public good, by a tax on income. The only two goods in the economy are labour (or capital) and a consumption good, and workers are differentiated by only one parameter. A map describing ability to pay, from the agent types to tax liability, is given. Under what conditions on preferences can such a map be implemented by an income tax, in the sense that each agent pays exactly what is desired, after all behavioural adjustments? We find sufficient conditions on preferences and the map, as well as necessary conditions. The only restriction on tax functions is that they be measurable; bunching and gaps are allowed. Applications to both normative and positive public finance are discussed.