Measuring the market power of the United States steel industry

This dissertation measures the market power and the degree of competition in the U.S. steel industry over the period 1960-1985. A single output empirical model consisting of a demand equation, a supply relation, a cost function and two input demand equations is estimated first. An oligopoly index, which reveals the degree of competition in the industry is estimated within this structural model and interpreted as a measure of average firms' conduct. An index of market power, the Lerner index, is also estimated and decomposed into components due to firm conduct and the market demand elasticity. The empirical model is then extended into a two-output model to incorporate the multiproduct nature of steel production and to recognize that the degree of output market competition may differ across product markets. Steel output is disaggregated into sheets and strip products and all other steel products, and a separate index of market power is estimated for each product market.; In the single output model, the Lerner index is estimated to be 0.55-0.63, and the oligopoly index to be 0.3122. The hypotheses of price-taking behavior and monopolistic pricing behavior are rejected. In the two-output model, the oligopoly indices in the market for sheets and strip and for all other steel products are 0.1174 and 0.5799, respectively. The market power indices for the two output markets are estimated to be 0.15-0.39 and 0.69-0.74. The hypotheses of price-taking behavior and monopolistic pricing behavior are rejected in both output markets. The hypothesis of equal pricing behavior in both markets is rejected. Hence, we conclude that the output markets of the U.S. steel industry are oligopolistic and the degree of non-competitiveness is significant and differs across the markets.; The different degrees of competitiveness are interpreted in the context of the dynamic limit pricing model. The relatively high degree of market power in the market for other steel products is interpreted as a result of the superior cost performance of the mini-mills. Thus, we conclude that the estimated industry markup is a consequence of both the superior performance of the mini-mills and noncompetitive pricing behavior.