Union Relative Wage Effects: A Survey of Macro Estimates

This paper surveys union wage effect estimates drawn from empirical cross-section wage equations fitted to macro (aggregated) data in 34 post-1963 studies containing such equations. Each estimate is the partial derivative of the dependent wage variable (in logarithmic units) with respect to the extent-of-unionism (fraction unionized) variable in the equation. This paper shows that these estimates contain a mixture in uncertain ratio of the union/nonunion wage differential or wage gap, on the one hand, and an "extent-of-unionism" effect, on the other. Therefore the Macro estimates should not be interpreted as wage gap estimates.