Agricultural economics devotes much attention to analyzing market interventions and their distributional implications. Recent research emphasis has shifted from the analysis of policy effects toward explaining the underlying causes of farm policy (Gardner 1983). One popular research tool in analyzing the latter issue is the metaphor of aggregate rationality: policies are the outcome of a single optimization problem that balances the welfare of conflicting pressure groups. The criterion function that rationalizes policies is often called a political preference function (e.g., Rausser and Freebairn). In the present paper we investigate two assumptions underlying agricultural economic models of the political economy of wealth transfers: that the political process is indeed characterized by aggregate rationality (as questioned by Bates 1989), and that governments' objectives are limited to fiscal concerns and political rewards from pressure groups. Many studies posit that individual rationality in rent seeking leads to a form of aggregate rationality in policy outcome (Gardner 1983; Rausser and Freebairn). Other studies describe policy formation as a bargaining game leading to the maximization by a single policy maker of a criterion function of groups' welfare measures (Zusman). Rent-seeking activities by interest groups determine the weights given individual surplus measures in the criterion function (Roe and Graham-Tomasi, Rausser and de Gorter). A surplus-possibility (or transformation) set constrains this maximization of a weighted sum of surpluses. One infers the relative weights on group's welfare from the marginal rates of transformation of one surplus measure into another. Why might this model of aggregate rationality make sense? The political process often leads to enforceable contracts (e.g., the U.S. Farm Bill), which are outcomes of a recurring and predictable process (e.g., five-year farm bill cycle). Enforcement mechanisms and the repeated nature of the political-economic process make plausible the use of cooperative-game-theoretic arguments, i.e, the assumption of a Pareto efficient political equilibrium (Shubik). Studies typically treat the government as a monolithic actor (Fafchamps, Sadoulet, and de Janvry; Lopez; Oehmke and Yao), and some studies additionally posit cooperation of all players in the political-economic game (Love, Rausser, and Burton). The aggregation of individual rational behaviors of rent seekers and government agencies may not lead to a cooperative game equilibrium (Bates 1990, North 1991). The Folk theorem admits many solutions to a bargaining game and does not necessarily lead to Pareto optimality (Fudenberg and Maskin). By contrast, the original rent-seeking literature, as opposed to the cooperative/efficient view of government choice, posits government inefficiency as a cornerstone (Tullock): the existence of deadweight losses associated with transfers is prima facie evidence of inefficiency. In addition, enforcement mechanisms are often incomplete and costly (North 1990), and governments may have a vested interest in incompletely-enforced property rights, which exacerbate rent seeking and inefficient outcomes. This problem of aggregate rationality we term the integrability problem of policy choice. That is, can one recover a well-behaved criterion function from a set of equations describing observed policy outcomes? The second assumption, regarding government behavior, has a strong and weak form. The strong assumption excludes government as a fullfledged player in the political-economic game. Algebraically speaking, the government's obJohn Beghin and William Foster are assistant professors, Department of Agricultural and Resource Economics, North Carolina State University. The authors thank D. S. Bullock for his comments, D. Burton, L. Karp, A. Love, and G. Rausser for helpful discussions.
Read full abstract