RISING WAGE INEQUALITY: Allowing for Differences Across Skill Groups 2

The figures illustrate several interesting features of the union and nonunion wage structures. First, the rotation of the 1993 lines relative to the 45 degree line indicates that both union and nonunion wage outcomes have “widened” across skill groups in the past two decades. This is particularly true for women: the difference in predicted wages between the highest and lowest skill groups rose by about 18 percentage points in the nonunion sector and about 13 percentage points in the union sector from 1973-4 to 1993. Second, the flattening effect of unions on male wage outcomes may have moderated over time. Third, unions exert a more modest flattening effect on the female wage structure than the male structure.

An important caveat to this conclusion is the potential role of unobserved heterogeneity. If union workers in lower skill groups are positively selected and those in higher skill groups are negatively selected, then the “flattening effect” observed in Figure 2 is exaggerated. Figure 3 compares the unadjusted union-nonunion wage gaps across skill groups in 1973-4 and 1993 with a series of adjusted wage gaps based on the pattern of union wage effects across skill quintiles estimated in Card (1996). In the absence of longitudinal estimates of true union wage effects for women I have simply assumed that the adjusted effects are the same for men and women. As it happens, the unadjusted union wage gaps for women are fairly similar to the adjusted gaps. For men, on the other hand, the differences are more noticeable. In particular, the raw wage gaps are negative for the highest skill group whereas the adjusted gap is positive.

Tags: , ,