Using cross-sectional data for U.S. manufacturing, the author identifies a negative effect of unions on the profits of highly concentrated industries and virtually no effect on the profits of unconcentrated industries. results for concentrated industries are explained in terms of the monopoly model which predicts a negative effect of unions on profits equivalent to the change in surplus. monopoly model also suggests that previous profit studies which omitted the union variable were likely to significantly underestimate the relationship between concentration and profits. This prediction is also supported by empirical evidence. M sANY studies have demonstrated the ability of unions to raise wages above the level of nonunion workers.' Higher union wages in turn may affect the level of prices, employment, capital intensity, productivity and profits. Recent studies by Clark (1982) and Freeman (1983) have found significant negative effects of unions on profit rates in the United States. Using line of business data from 1970 to 1980 for 900 firms, Clark found a negative effect of unions on the rate of return to capital. More specifically, the union effects were found to be significantly negative when market shares were low and insignificant when market shares were high. According to Clark, the absence of a union effect in the latter case can be explained by the ability of firms with market power to pass on higher union wages into higher prices (1982, p. 47). On the other hand, Freeman (1983), using data from the Survey of Manufactures from 1958 to 1976 and the Internal Revenue Service from 1965 to 1976 also found a negative effect of unions on profits. Contrary to Clark's results, however, the union profit effect was negative only in concentrated industries. point of contention appears to be the role of concentration in determining the union profit effect, rather than the overall negative impact of unions. results of Clark and Freeman are important because previous industrial organization research on profit rates generally ignored unions. For instance, of the forty-six profit studies surveyed by Weiss (1974), none included a union variable. More recently, a profit study in this Review by Ravenscraft (1983) using high quality, line of business data from the Federal Trade Commission, also ignored unionization. As will be demonstrated, the omission of the union variable leads to an understatement of the effect of industrial concentration on profits. This means that concentrated industries generate much greater profits than indicated by previous research. In order to measure the size of the understatement, a new concept, employer's surplus, is developed which illustrates the theoretical impact of a union wage increase on monopoly profits. empirical estimates of surplus in this paper indicate how monopoly profits are divided between firms and unions. Most of the data used in this study were obtained from James Medoff and Charles Brown.2 Unlike the previous two studies, each observation corresponds to a state by two digit SIC industry for manufacturing in 1972. Because of the large variations in unionization across states, this data set offers a useful test of the union profit effect. I. Employer's Surplus According to accepted theory, the profit maximizing monopolist will employ labor up to the point where the market wage equals the marginal revenue product, MRP.3 If for simplicity we assume that all other factors are variable, then the MRP curve represents the monopolist's long-run demand for labor. In the absence of unions, employment is determined by the intersection of the MRP curve and nonunion wage (Wn) as pictured in figure 1. If unions raise wages to Wa, employment will decrease from Ln to L, with the reducReceived for publication November 28, 1983. Revision accepted for publication July 6, 1984. *Eastern Washington University. I am very grateful to Dave Bunting, Lisa Brown, Bill Dickens, Claire Brown, and George Strauss for their valuable suggestions on this research. This paper is a revised version of chapter 3 from the author's Ph.D. thesis, The Union Impact on Profits, Productivity, and Prices, University of California, Berkeley. 1 See Parseley (1980) for a recent review of this literature. 2 I would like to thank Brown and Medoff for making these data available to me. 3 Marginal revenue product is equal to marginal revenue multiplied by marginal physical product. See Rees (1979).