Gains from Contracting for US Hog Growers
Publication or External Link
Vertical coordination through contracts between farmers and other stages of the agro food chain have been of growing importance in US agriculture. Production contract arrangements between contractors and individual growers have been one of the major vehicles of this emerging system of vertical coordination. Despite the unprecedented success of production contracts as claimed by many through risk reduction, income stabilization, use of improved managerial inputs, and know-how transfer from contractors to growers, contract growers dissatisfied with existing contract payments complain that contractors are extracting too much of contract benefits while growers gain only small, or even negative, returns from contract production. Thus, measuring growers' gains from contracting, and understanding what determines the returns to contracting, is important for evaluating the policy issues associated with contracting in agriculture. This study examines hog growers' gains from contracting and explores the distribution of the gains from contracting among contract hog growers.
The purpose of this dissertation is threefold. The first purpose is to review the major issues that have been examined in the literature on principal-agent theory, with special attention to the issues that are important in the agricultural sector in general and hog production in particular. Some further extensions of the basic theories are developed to enable solving the empirical puzzles. Some implications for agents' gains from contracting in both static and dynamic settings are derived. Related discussion shows how hog contracts relate to standard principal-agent theories. The main finding is that for the most plausible information structure, that is, when growers have partial but better knowledge of their ability than contractors, some low ability growers with below average productivity receive negative gains from contracting on average. This conclusion holds even when renegotiation-proof long-term contracts are in place for each ability distribution. In contrast, none of the growers receives negative gains from contracting when they have complete knowledge of their ability before signing the contract.
The second purpose is an explicit theoretical modeling of hog contracts to theoretically analyze optimal incentive structures for hog contracts. A principal-agent model allowing reservation profit to vary with ability is developed to explore whether some contract growers receive negative gains from contracting on average. The results of this theoretical development suggest a rich set of alternative conditions where negative average gains from contracting are possible for growers with below average productivity of any particular ability level discernible by the contractor. These losses are likely to be repeated under long-term contracting when ability is a permanent random draw for the grower that is different than expected. Even low-ability growers with above average productivity can experience an ex post loss from contracting.
The third purpose of this dissertation is to test the main theoretical findings on contract growers' gains from contracting using revealed preference data from the well-known Agricultural Resource Management Survey (ARMS) for 2004. In order to do this, contract growers' gains from contracting are measured using standard impact evaluation methods. By going beyond typical estimation of how contracting affects average growers' profits, estimates are developed to show how high-profit growers are affected differently from low-profit growers, and whether some growers are worse off with contracting. The results are especially relevant for policy analysis regarding hog contracting because it shows what share of contract growers lose from contracting and identifies their characteristics. The impact distribution of contract growers' gains is also explored using quantile regression. The estimated growers' gains from contracting are then used to evaluate theoretical predictions of the hog model.
The main empirical findings of this research can be summarized as follows. First, both risk reduction and limited credit are important motivations for hog contracting. Second, the sorting effect is positive, implying that contract growers tend (because of the effect of unobservables) to choose contracting because of a comparative advantage in doing so. A positive selection bias is estimated, which tends to give contract growers a comparative disadvantage from independent operation. Third, high ability growers earn higher profits on average than low ability growers as predicted by the hog contracting model. Fourth, the mean effect of contracting for contract growers (ATET) is positive for all contract growers. However, when contract growers are divided into quartiles by size, the ATET is positive only for the lower three quartiles whereas it is negative for the highest. Fifth, the ATET decrease over quantiles of the profit distribution for contract growers and the ATNT decreases over quantiles of the profit distribution for independent growers. Sixth, one third of the contract growers receive negative gains from contracting. Below average productivity growers lose from contracting as predicted by the hog contracting model. Seventh, the mean effect of contracting for independent growers (ATNT) is negative. Eighth, the ATET exceeds the ATNT, meaning that independent growers would gain less than contract growers had they contracted. Ninth, contract and independent growers are different with respect to the productivity of the variable factors of production but unilateral technological superiority of one group to the other is not found. Finally, the results suggest that small growers will be forced either to exit the hog business or expand operations regardless of their contracting status.