Source: Ozkan Eren, Industrial and Labor Relations Review, Vol. 62 no. 3, April 2009
From the abstract:
This paper examines the union membership wage premium among private sector employees covered by collective bargaining agreements. Using Current Population Survey data for 2000-2003, the author not only estimates the conditional mean wage premium—the metric on which most previous research has focused—but also employs recently developed instrumental quantile regression techniques to estimate the wage effect of union membership across the wage distribution. Union members enjoyed, on average, a wage premium of 9% over comparable covered nonmembers. Further analyses find no evidence that this mean premium is explained either by unobserved differences or by measurement error. The author also finds that a narrow focus on the mean impact partially masks heterogeneity in the impact across groups of workers. Notably, union wage effects were considerably more pronounced for low wage earners than for high wage earners.