Union Contracts and Money Wage Changes in U.S. Manufacturing Industries

This paper analyzes the relationship between wage changes negotiated in selected large collective bargaining agreements and industry-wide changes in average hourly earnings. Multiple regressions are used to examine the relationship in two-digit man-ufacturing industries for the period 195811 to 197811. Negotiated wage changes are found to have a major impact on overall movements in hourly earnings. The estimated equations have potential applications for short-run wage forecasting and for policy analyses. Over the past decade research on U. S. money wage determina-tion has evolved in several directions. Three important parts of the recent literature are the modified or augmented Phillips curve, the Friedman [1968]-Phelps [1970] accelerationist hypothesis, and wage distortion hypotheses. ' A fourth approach stresses the possible role of unions in overall money wage inflation. This is suggested in the contracts literature [Azariadis, 1975; Baily, 1976] and in a series of union-nonunion wage papers by Ashenfelter, Johnson, and Pencavel. 2 There have been some empirical analyses of the aggregate effects of union settlements [Hall, 1974; Flanagan, 1976; Johnson, 1977; Mitchell, 1980a,b], but lack of time series data on union wages has precluded exhaustive testing of this issue. The present paper fits into this fourth general area of money wage research. Rather than addressing broad questions about union con-tracts and wage inflation, however, it focuses on a much narrower topic. The paper employs multiple regressions to examine the em-pirical linkage between wage changes negotiated in large collective bargaining situations and industry-wide money wage movements. Because the paper does not try to explain the size of the bargaining agreements, it cannot address the key question about the possible independent role of contract settlements in overall money wage in-flation. Until the effect of union settlements on hourly earnings i