Union Wage Gap in the U.S. Construction Sector: 1983–2007

• DOI: http://doi.org/10.1111/irel.12029
• Date: 01 July 2013
• Published date: 01 July 2013

Union Wage Gap in the U.S. Construction

Sector: 1983–2007

CIHAN BILGINSOY*

Wage gap decomposition shows that declining union power was the principal

force behind the shrinking union wage premium in the U.S. construction industry

between the 1980s and the 2000s. This decline was largely offset by changes in

returns to workforce attributes and workforce compositions. Without these moder-

ating effects, the decline in the wage gap would have been twice as large (in log

points). The patterns were similar in the basic and mechanical trades, but magni-

tudes of change were larger in the latter.

Introduction

OVER THE LAST THREE DECADES,THE U.S. CONSTRUCTION INDUSTRY,UNLIKE MANU-

FACTURING, did not experience a major dislocation. It is primarily a local activity

insulated from international competition. Nor did construction, unlike services,

expand. Its shares in total output and employment remained around 4 to 5 per-

cent. However, fundamental transformations of industrial relations took place

within the sector (Allen 1994; Philips 2003; Weil 2005). The open-shop sector

gained control of residential, small commercial, and remodeling work, and also

made inroads into industrial construction, which was historically dominated

by the union contractors. The political and legal environment turned increasingly

in favor of the open-shop sector. The share of unionized workers in the blue-

collar construction workforce shrank by about one third since the 1980s. The

relationship between declining unionization and the union wage premium has

been of considerable interest in the literature. Thieblot (2001) argued that the

primary reason behind falling unionization is the success of building trades

unions in achieving high wages for their members. Belman and Voos (2006),

however, presented econometric evidence disputing the claim that unions priced

themselves out of the market.

*The author’saffiliation is Department of Economics, University of Utah, Salt Lake City, UT, USA.

Email: bilginsoy@economics.utah.edu. I am grateful to G€

unseli Berik and two reviewers of this journal for

insightful comments and suggestions. I am responsible for all remaining errors.

INDUSTRIAL RELATIONS, Vol. 52, No. 3 (July 2013). ©2013 Regents of the University of California

Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

677

Research on the changing structure of the industry did not generate a paral-

lel line of dedicated inquiry into the evolution of construction wages. The most

recent studies that focus on the sources of union wage differentials in the con-

struction industry date back to the 1980s (e.g., Allen 1988, 1994; Perloff and

Sickles 1987). This void exists despite interesting developments in wages.

Union wage premium declined since the 1980s by only about 10 percent.

Once adjusted for workforce characteristics, however, reduction in the union

premium was 23 percent. The discrepancy between the unadjusted and

adjusted wage premiums suggests that while the construction sector was under-

going transformation, concurrent changes were taking place in the workforce

that largely offset the impact of declining union effect on the union wage gap.

This article focuses on the union wage gap in the blue-collar construction

workforce. It uses determination of the wage gap as a window to peek into the

relative transformation of union and nonunion workforces, and complements

existing studies of industrial relations in construction by gauging contributions

of changes in union bargaining power and workforce characteristics to the

over-time change in union wage gap. I carry out the analysis by comparing

two samples of pooled observations from 1983 to 1988 and 2002 to 2007.

First, I estimate union and nonunion wage equations for each period and

decompose the wage gap into union–nonunion mean endowment and remuner-

ation (or return) differences. The remuneration effect is the adjusted union

wage gap, which is customarily interpreted as the union effect. Second, I quan-

tify the sources of the change in wage gap from the 1980s to 2000s . This sec-

ond decomposition finds that the declining union effect was the major

contributor to the drop in the union–nonunion wage gap. Its impact, in fact,

exceeded the change in the wage gap by a factor of more than two. However,

there were substantial changes in returns to workforce attributes and the distri-

bution of workforce endowments in favor of the union sector, which buoyed

the relative union wage and partially offset the negative impact of the drop in

union power. When decompositions are applied separately to the basic and the

mechanical trades, similar patterns are observed, but estimated magnitudes of

over-time changes in the wage gap, union effect, and endowments effect were

larger in the mechanical trades.

Transformation of the Construction Industry and the Union Wage Gap

The central problem in gauging the effects of unions on wages is that the

wage distribution that would have prevailed in the absence of unions is not

observable. While it is a flawed measure of this counterfactual—because it

overlooks the general equilibrium (e.g., spillover and threat) effects of the

unions on wages—the nonunion wage is used often as a proxy for the compet-

678 / CIHAN BILGINSOY