I. INTRODUCTION More than 40 years ago, Arnold Harberger [1954] estimated the welfare loss due to monopoly in the United States. Harberger's finding that the welfare loss caused by monopoly is surprisingly small sparked a lively debate that enriched the economic literature enormously. The articles by David Schwartzman [1960], David Kamerschen [1966], Frederick Bell [1968], Dean Worcester [1973], and John Siegfried and Thomas Tiemann [1974] constitute but a small sample out of the vast number of studies of this issue. Later, Keith Cowling and Dennis C. Mueller [1978] found some higher values of deadweight losses - ten times as large as Harberger's. In a subsequent study I was able to lend support to Harberger's by analyzing the interactions between the minority of large firms in an industry, that act as independent oligopolists, and most of small firms that behave as price takers [1986]. The exploration of the dynamic relationship between concentration and prices started with Schumpeter [1949], who hailed the entrepreneur who can produce more cheaply, make a profit that remains in his pocket until others, who follow him, copy his model. The process eventually results in the annihilation of the surplus over costs, when the new business form has become part of the circular flow. [1949, 133]. In a related set of studies, Demsetz [1973], Peltzman [1977] and Lustgarten [1979] explored the relationship between concentration ratios and increased efficiencies in manufacturing industries and found them to be positively correlated. I could lend support to this conclusion by postulating a dynamic process of cause and effect cast in a price-leadership model [1984]. In a recent study, Salinger [1990] concluded from a cross-sectional study that large changes in concentration, whether negative or positive, are associated with price reductions. The present paper builds on past research to demonstrate that, while innovation may lead to increases in market share and market concentration, the price increases that might be expected to follow from increased concentration are likely to be offset by the price decreases that bring the increased concentration into being.(1) If so, estimating deadweight losses by subtracting the static consumers' losses from producers' gains is misleading because it ignores intertemporal consumers' gains. The dynamic link among efficiency, falling prices, concentration, and welfare losses suggests that a very useful method of evaluating an industry's contribution to welfare would be to study its economic history. If today an industry is heavily concentrated, an appropriate starting point for a historical study would be an earlier year when the industry was still relatively less concentrated. This paper presents, first, a welfare theory of dynamic evolution of an industry that becomes progressively more concentrated over time; second, an empirical study - based on the history of the beer industry - that focuses on the calculus of dynamic social gains from technological change; and, third, an examination of a small group of industries that became more heavily concentrated between 1963 and 1982. The last section sets forth the conclusions to be drawn from this research. II. DYNAMIC GAINS AND STATIC LOSSES Suppose a concentrated industry was unconcentrated 40 years ago. The industry's firms were small but not atomistic, and they sold their product in nearly perfectly competitive markets practically causing no deadweight losses. The dynamic process begins when, initially, a group of pioneering firms invests in the development of a new technology that reduces costs and by lowering the price to consumers succeeds in increasing its market share. The gain to consumers flows directly from this intertemporal decline in price. The increase in the share of the innovating firms can initially mitigate the downward pressure on prices, but in the long run reversing it is not likely. …