Market power and efficiency impacts of concentration: Evidence from U.S. manufacturing

This dissertation estimates the price, cost and welfare impacts of industrial concentration in U.S. manufacturing industries in order to assess the desirability of increases in concentration. A “new empirical industrial organization” (NEIO) model that separates the price and cost effects of concentration is estimated for each of 255 industries at the 4-digit SIC level. The estimation of the five-equation model (pricing, output demand and three input demands) for each industry resulted in over 4,000 parameters including the price elasticities of demand, conjectural variation, and economies of size. From the econometric results, the elasticities of market power and cost efficiency with respect to the Herfindahl index are computed. Finally, changes in consumer and producer surplus as well as cost savings are computed for each industry in the sample. ^ Empirical findings point out that concentration leads to lower prices only in roughly one-third of the industries and, therefore, to higher prices in two-thirds of them. The trade-offs between market power and efficiency are as follows: efficiency overweighs market power in 36% of the industries resulting, in lower prices; efficiencies are overweighed by the market power effect in 42% of the industries, resulting in higher prices; and inefficiencies reinforce the market power effect in 22% of the industries, resulting in even higher prices. These results imply that from a “consumer surplus” standard, concentration is socially desirable only in 36% of the cases, those where prices decline. Consumer losses are more likely to take place in high-concentration markets. These results lend support to the current guidelines used by the U.S. Federal Trade Commission that closely scrutinize mergers of highly concentrated industries but generally allow mergers in low-concentration industries. ^ From a “total welfare” standard, however, concentration is socially desirable in 177 industries (69.41%). Concentration is found to be beneficial for low levels of concentration, still beneficial for moderate levels and detrimental in highly concentrated markets, in which the cost savings generated from the efficiency effect are not enough to compensate for the deadweight losses from the market power effect. ^ With a one-percent across-the-board increase in the Herfindahl index, consumers would lose $1 billion while industries would gain $1.5 billion, resulting in an aggregate welfare increase of approximately $473 million, due to mainly efficiency gains pocketed by the industries.