A precursor of modern public choice theory was Knut Wicksell (1896),^[1] which treated government as political exchange, a quid pro quo, in formulating a benefit principle linking taxes and expenditures.^[2]

Some subsequent economic analysis has been described as treating government as though it attempted "to maximize some kind sort of welfare function for society" and as distinct from characterizations of economic agents, such as those in business.^[3] In contrast, public choice theory modeled government as made up of officials who, besides pursuing the public interest, might act to benefit themselves, for example in the budget-maximizing model of bureaucracy, possibly at the cost of efficiency.^[4][3]

Modern public-choice theory has been dated from the work of Duncan Black. In a series of papers from 1948, which culminated in The Theory of Committees and Elections (1958),^[5] and later, Black outlined a program of unification toward a more general "Theory of Economic and Political Choices" based on common formal methods, developed underlying concepts of what would become median voter theory, and rediscovered earlier works on voting theory.^[6][7][3]

Kenneth J. Arrow's Social Choice and Individual Values (1951) influenced formulation of the theory. Among other important works are Anthony Downs (1957) An Economic Theory of Democracy and Mancur Olson (1965) The Logic of Collective Action .^[8]

James M. Buchanan and Gordon Tullock coauthored The Calculus of Consent: Logical Foundations of Constitutional Democracy (1962), considered one of the landmarks in public choice. In particular, the Preface describes the book as "about the political organization" of a free society. But its methodology, conceptual apparatus, and analytics "are derived, essentially, from the discipline that has as its subject the economic organization of such a society" (1962, p. v). The book focuses on positive-economic analysis as to the development of constitutional democracy but in an ethical context of consent. The consent takes the form of a compensation principle like Pareto efficiency for making a policy change and unanimity or at least no opposition as a point of departure for social choice.

Somewhat later, the probabilistic voting theory started to displace the median voter theory in showing how to find Nash equilibria in multidimensional space. The theory was later formalized further by Peter Coughlin.^[9]