Price transmission and market power in the vertically separated markets of fluid milk
This dissertation analyzes price transmission, market power and vertical relationships at the processing and retailing levels for the fluid milk market in the New England area. The New England fluid milk market is particularly interesting due to its prior minimum farm price policies. This study contributes to previous market power and market conduct literature in three ways. First, it incorporates time series properties to test for asymmetry in price transmission by employing an error correction model and a revised version of the Engle-Granger cointegration test. Second, it proposes a method based on previous tax rate analysis research to test for market power without requiring detailed cost data. Third, it utilizes a natural experiment of minimum price policies to simulate retail price responses to changes in marginal cost so that different vertical relationship models can be evaluated and ranked.
Most previous asymmetric price transmission research is based on the Houck model, which might produce spurious regressions when price series are nonstationary. This dissertation exploits the time series properties by employing an error correction model. Furthermore, the revised version of the standard Engle-Granger cointegration test (Enders and Siklos, 2001) was employed to improve the efficiency of cointegration testing and accomodate asymmetry in convergence to long run equilibrium. Asymmetry in price transmission is suggestive of market power. New empirical industrial organization models have been widely used to estimate market power, but they require input price data. This study tests the market power hypothesis by estimating a reduced form model and analyzing the extent to which farm prices are transferred to retail prices. Since only farm and retail prices are available, the market power hypothesis can only be tested for the combined processing and retailing markets. It is unclear how market power is distributed between the two market levels due to lack of wholesale milk price data. The Northeast Interstate Dairy Compact provides a natural experiment to observe changes in marginal costs. Through simulations of retail price responses to marginal cost changes, this study ranked different vertical relationship models according to their fit to the observed data.
The results show that farm and retail prices are cointegrated and provide empirical evidence for asymmetry in farm to retail price transmission. Under some reasonable assumptions on processing and marketing costs, the minimum price policy, which was intended to protect farmers from market power of larger processors, actually increases market power exercised by processors and retailers. Finally, retailers are found to exercise more market power than processors.