Estimating the residual demand curve facing a single firm

Abstract

This paper presents an econometric technique for estimating the single firm residual demand curve that does not require the estimation of demand cross-elasticities. The technique is particularly suited for the estimation of firm market power in product differentiated industries, where cross-elasticities are notoriously difficult to measure. One estimated parameter is directly related to a firm's markup of price over marginal cost when product differentiation is extensive, when an industry consists of a dominant firm and a price-taking fringe, when a Consistent Conjectures Equilibrium is the correct oligopoly solution concept, when firms are Stackelberg leaders relative to their environment, or when firms are perfect competitors. The estimates can indicate power over price in other circumstances. An instrumental variables technique is employed, using firm-individuated factor prices to identify firm-specific residual demand. Yet even when instruments are unavailable, the technique produces elasticity estimates biased in the conservative direction of disproving market power. This methodology is applied to estimate the market power of three firms in an industry characterized by product differentiation: brewing. Over our 1962–82 sample period, Pabst and Coors had similar national market shares and each had high shares in several states. Yet Pabst was virtually a price-taker while Coors possessed substantial market power. Anheuser-Busch had market power in the early part of the sample, but little after 1975 when Miller Brewing changed the competitive nature of the industry.