Whose Employment is Affected by Unions?

12/01/2002
Summary of working paper 9043
Featured in print Digest

More union involvement in wage setting significantly decreases the employment rate of young and older individuals relative to the prime-aged group.

In 1973, unemployment in most European countries was modest, ranging between 2.0 and 3.2 percent, compared to 4.8 percent for the United States. By 1995, the unemployment situation for the European countries had changed dramatically, rising to an average of 10.7 percent. But in the United States, the unemployment rate rose only to 5.6 percent by 1995, roughly half that of European countries. This reversal of fortunes was concentrated on youth, older individuals, and women, rather than on prime age males: while the employment-to-population ratios (employment rates) of all groups rose in the United States relative to other Western countries, the increases were especially large for youth and older individuals, and somewhat larger for women, according to new research by Giuseppe Bertola, Francine Blau, and Lawrence Kahn. At the same time, the authors note that unionization fell in the United States compared to these other countries.

In Labor Market Institutions and Demographic Employment Patterns (NBER Working Paper No. 9043), the researchers investigate the cross-country impact of labor market institutions on the relative employment rates of youth, women, and older individuals, relative to prime age males. This study overcomes the drawbacks of earlier research by examining data from a wider base of countries, 17 in all, and over a longer time frame, 1960-96. The authors focus on the wage-employment tradeoffs faced by different groups of workers. The countries included in the study were part of the Organization for Economic Cooperation (OECD): Australia, Belgium, Canada, Denmark, Finland, France, Germany, Japan, the Netherlands, Italy, Norway, Portugal, New Zealand, Spain, Sweden, the United Kingdom, and the United States. The researchers control for overall demographic factors, country effects, and institutional factors, such as collective bargaining coverage, labor tax rates, employment protection, unemployment insurance, and rates for retirement benefits, as well as the overall unemployment rate (although results were similar when the unemployment was not included as an explanatory variable).

The researchers find that, for both men and women, more union involvement in wage setting significantly decreases the employment rate of young and older individuals relative to the prime-aged group (with no significant effects on the relative unemployment of these groups). In contrast, a larger role for unions has little impact on male-female employment rate differentials but raises female unemployment relative to male unemployment.

The authors conjecture that unions, in determining their wage-setting policies, balance out the gains from higher wages against the losses from resulting reductions in employment. Groups with the most extensive non-market opportunities to use time productively suffer the least when they lose employment. These groups are likely to be youth, older individuals, and women, all of whom have more extensive non-market uses of time than prime age males do: household production for women (under a traditional division of labor in the family), schooling for youth, and retirement for older individuals. As a result, unions negotiate the highest wage increases for these groups, leading to larger employment reductions for them. Alternatively, it may be more socially acceptable in highly industrialized societies to concentrate employment losses associated with unions on women, youth, and the elderly.

The authors' findings suggest that union wage-setting policies price the young and elderly out of employment and drive affected individuals in these groups to non-labor-force activities, leaving unemployment rates unchanged. A probable scenario for women, according to the authors, is that high union wages encourage them to enter the work force, pushing up their unemployment rates. However, the expected employment declines for women do not materialize, because women who otherwise would not be employed because of the high union wage floors find work in an unregulated work field or in the public sector.

-- Marie Bussing-Burks