Strategic Pricing in a Differentiated Product Oligopoly Model: Fluid Milk in Boston

In an imperfectly competitive industry, differentiated products compete with each other with price rather than quantity as the strategic variable. Several previous studies have employed a generalized Nash Bertrand model: Liang (1989), Cotterill (1994), Cotterill, Putsis and Dhar (2000), and Kinoshita, Suzuki, Kawamura, Watanabe and Kaiser (2001); however, only Liang has explored the theoretical foundations of that model. This paper generalizes the Liang two good model to three goods. A surprising and important result follows. Price conjectural variations do not exist in models with 3 or more goods. Price reaction functions, however, exist in multiple good models. We estimate them jointly with a brand level demand system to evaluate the total impact of a brand manager's price change on own quantity. In a differentiated product market this is a useful addition to a partial demand elasticity approach because a change in one brand's price typically engenders a price reaction by other brands which affects own quantity via substantial cross price elasticities among substitutes. Strategic pricing in the Boston fluid milk market was also influenced by the existence of a raw milk price support program, the Northeast Dairy Compact. We find that the advent of the Compact was a focal point event that crystallized a shift away from Nash Bertrand to more cooperative pricing. If the downstream market is not competitive one needs to consider strategic price reactions when designing and evaluating agricultural price programs. Key Words: oligopoly, price conjectural variations, brand level demand elasticities, focal point collusion