Metropolitan Earnings Inequality: Union and Government-Sector Employment Effects


  • *Direct correspondence to Thomas Volscho, Department of Sociology, University of Connecticut Unit 2068, Storrs, CT 06269 〈〉. Data and computer code for replication available from Tom Volscho. We thank Stephen Adair, Andy Baker, Angie Beeman, Ivar Berg, David Brady, Arne Kalleberg, Arthur Sakamoto, and Michael Wallace for providing useful suggestions. We are indebted to Robert Adelman, David Weakliem, the anonymous SSQ reviewers, and the editor for providing constructive comments on various drafts of this article.


Objective. This study examines the effects of union density and government-sector employment on earnings inequality in Metropolitan Statistical Areas (MSAs) in the United States.

Methods. Data on 167 MSAs from the 2000 Census are analyzed using standard regression techniques. Four measures of Atkinson's index (e=0.5, 1.0, 2.0, 3.0) are used as the measure of earnings inequality for full-time, year-round workers.

Results. MSAs with greater union density and greater government-sector employment have lower earnings inequality. The progressive effect of union density is strongest for earners in the middle of the distribution and less beneficial for workers at the bottom of the distribution. Government employment is generally associated with lower levels of earnings inequality, but state and federal government employment have the strongest effects.

Conclusion. Even in the late 1990s, unions and government-sector employment remain effective at reducing earnings inequality.