An examination of the wage contour hypothesis: Wage leadership in the united states construction industry

One of the assumptions contained in John Dunlop's wage contour hypothesis is that wage leadership positions among bargaining units in the construction industry remain constant over a period of time. A peripheral issue concerns the dominance of craft or regional effects in the formation of wage contours. This study tests for stability through the systematic identification of hypothetical wage leaders and the comparison of annual results over a seven year period. Dominance effects are examined through the formation of contours around the identified wage leaders and the subsequent determination of the dominating effect in each contour membership. Discriminant analysis is used in the examination of stability of wage leaders in order to identify those bargaining units most likely to be wage leaders given the values of certain relevant variables. Seven groups of wage leaders and non-leaders are identified, one for each year of the study (1973 through 1979). The resulting comparisons of the memberships of these groups indicate that there is much variation among wage leadership positions in the construction industry, at least during the decade of the 1970s. The dominance of craft or regional effects is tested by forming contours using the type of cluster analysis that allows for prespecification of the cluster centers. The centers used are the wage leaders identified in the first part of the study. The results of this analysis indicate that the regional effect dominates contour formation more frequently than the craft effect in the construction industry.