ESTIMATING BACKWARD INTEGRATION IN A PRIMARY INPUT MARKET: THE CASE OF U.S. HOG INDUSTRY

The U.S. pork sector is evolving from an industry of small, independent firms vertically linked by spot markets to one of substantially larger firms vertically connected through contractual agreements and integration. Potential benefits to this tighter vertical arrangement include lower consumer pork prices, although the true nature of this benefit is still under debate. At the same time, there is concern of market foreclosure because highly vertically integrated industry may prevent independent hog producers from having access to open markets in which to sell their output. Boehlje underscores the need for empirical answers to questions related to the above structural change in the pork industry. The objective of this paper is to estimate econometrically the extent of backward integration by pork processing firms into the upstream hog production stage, taking into account the oligopsonistic nature of the processors, and to simulate the effect of vertical integration on consumer and producer prices and welfare. Following Perry, backward integration is defined as the fraction of the upstream limiting production factor (e.g., farm land, feedlot facilities and water supply) that the downstream processors own. This economic measure of vertical integration is richer than the traditional measure which is a ratio of internal intermediate input production (by the upstream subsidiary) to the total intermediate input usage (by the downstream processor). Rather than an economic measure of backward integration, this accounting ratio reflects the resolution of vertical integration and, by itself alone, does not give insight into the price and welfare effects of the structural change. The procedure of this study is as follows. An individual processor's profit maximization problem consists of maximizing revenue from a variety of pork products and minimizing the hog input expenditures, where the latter sub-problem involves a division of the total hog input procurements between open market purchases and internal production by the upstream subsidiary. From the optimization, one derives the optimal pork output supply, total hog input demand, and open market hog input purchases. These behavioral equations of the individual processor are aggregated and estimated in conjunction with the pork product demand equations of consumers and the open market hog supply equation of independent hog producers. The aggregate model is then used as a basis for simulating price and welfare effects of vertical integration. Possible estimation problems arising from the complexity of the optimal solutions are anticipated and coping strategies devised. In addition, data are unavailable on open market hog quantities; as a result theoretical restrictions derived from the model are used to transform this variable so as to utilize the slaughter data published by the USDA.