Union Membership and Wage Variation

07/01/2003
Summary of working paper 9473
Featured in print Digest

For men, unions tend to have an equalizing effect on wages across skill groups. However, unions do not reduce wage inequality for women. These trends were remarkably similar in all three countries over the past twenty years.

Union membership rates have been declining in the United States, the United Kingdom, and Canada. Since the mid-1950s, for example, union membership in the United States has slipped to under 15 percent. During the same period, wage inequality has risen in these three industrialized nations.

In Unionization and Wage Inequality: A Comparative Study of the U.S., the U.K., and Canada (NBER Working Paper No. 9473), co-authors David Card, Thomas Lemieux, and W. Craig Riddell explore the impact of unions on wages in light of the rise in earnings inequality in several industrialized nations. They find that, for men, unions tend to have an equalizing effect on wages across skill groups. However, unions do not reduce wage inequality for women. These trends were remarkably similar in all three countries over the past twenty years.

The authors studied these three countries because the institutional arrangements governing unionization and collective bargaining are relatively similar there and all three now collect comparable data on wages and unions. Bargaining is highly decentralized in the three countries and the fraction of the workforce covered by collective bargaining is relatively modest, allowing for comparisons of workers covered by union contracts with those who are not.

Union coverage tends to be concentrated in the middle of the skill distribution for men in the three countries and union wages tend to be compressed relative to non-union wages. This implies that unions systematically reduce the variance of wages for men in all three countries.

Three factors appear to account for the fact that unions do not reduce income inequality among women. First, unionized women are more concentrated in the upper end of the wage distribution than their male counterparts. Second, the union wage gap is larger for women than for men, resulting in a larger "between-sector" effect. Third, unions have about the same impact on wages of women in different skill groups, whereas they tend to raise wages more for men at the bottom of the income distribution.

The authors also find that the decline in unionization of women has been much smaller than that of men. As a consequence, unionization rates of men and women are nearly equal in all three countries, marking a sharp departure from the historical pattern. However, the modest decline in union coverage for women had little impact on their wage inequality.

During the 1980s and 1990s, unionization rates fell in all three countries, with the most rapid decline in the United Kingdom, followed by the United States and Canada. These trends contributed to rising male wage inequality, particularly in Britain. The authors estimate that the precipitous fall in unionization in the United Kingdom explains up to two-thirds of the difference in the trend in male wage inequality between Britain and the United States.

-- Les Picker