Identifying and measuring market power with residual demand elasticity: Evidence from the United States aluminum market, 1954--1984

The empirical methodology in identifying and measuring market power always seeks to orient a satisfactory basis; among these is the residual demand analysis. The residual demand curve facing a firm or an industry is the firm\u27s or industry\u27s demand curve taking into account the supply response of all other competitors. A firm\u27s or an industry\u27s market power is determined largely by the firm-specific or industry-specific demand curve so that the demand curve estimation potentially gains an insight into the firm\u27s or industry\u27s market power. This study presents various econometric techniques to demonstrate the existence or absence of Alcoa\u27s and the primary aluminum industry\u27s market power in the U.S. aluminum market based on the residual demand analysis. Alcoa\u27s and the primary industry\u27s estimated inverse residual demand elasticities imply the corresponding market power. The structural model is applied to estimate Alcoa\u27s market power within the primary industry so that the product is homogeneous. Based on the DF/CF model, Alcoa\u27s residual demand elasticity is calculated through the estimated market demand elasticity, fringe supply elasticity, and Alcoa\u27s extant market share. The structural model is also applied to estimate the primary industry\u27s market power with the aluminum market delineated with primary and secondary aluminum so that the product is heterogeneous. By simultaneously estimating the primary and secondary industries\u27 inverse residual demand curves, the associated inverse residual demand elasticities correspond to the industries\u27 market power in the aluminum market. By simultaneously estimating the industries\u27 inverse partial residual demand curves, one can also infer the industries\u27 market power when the tacit collusion is engaged across the industries. The time series linear feedback model is applied to estimate Alcoa\u27s and the primary industry\u27s market power, based on the bi-directional causality between the market demand and market power and the bi-directional causality between Alcoa\u27s and the primary industry\u27s market power. The former measures Alcoa\u27s and the industry\u27s abilities to adjust price given demand fluctuation, while the latter measures Alcoa\u27s and the industry\u27s conjectural variation given the inter-firm rivalry in the primary industry. The estimations of frequency feedback method and impulse response analysis further distinguish the causality into the long-run and short-run effects