WASHINGTON - January 25 - For the first time in U.S. history, union membership rates were lower in manufacturing than in the rest of the economy. According to the Bureau of Labor Statistics annual union membership report released today, union membership declined sharply in 2006, from 12.5 percent of all workers in both 2004 and 2005, to just 12.0 percent of all workers last year. The following analysis by CEPR economist John Schmitt and researcher Ben Zipperer discusses the annual union membership data:

Overall, the number of U.S. workers in a union fell last year by 326,000 workers, to 15.4 million workers in 2006. The largest decrease in union membership rates occurred in manufacturing, where union membership dropped 1.3 percentage points to just 11.7 percent of manufacturing workers. For the first time since the BLS began tracking these trends, and likely for the first time in U.S. history, union membership rates were lower in manufacturing (11.7 percent) than in the rest of the economy (12.0 percent).

In addition to losses in manufacturing, very few segments of the private sector reported gains in unionization. Union membership in the private sector slid in 2006 to only 7.4 percent. Among public-sector workers, membership also fell (down 0.3 percentage points), but, at 36.2 percent, remained at levels consistent with those over the last two decades. Public-sector union jobs in 2006 accounted for almost half of union members, even though public-sector employment comprised less than one-fifth of the economy. (For a discussion of trends in illegal firings in the private sector during union organizing campaigns, see CEPR’s report, Dropping the Ax: Illegal Firings During Union Election Campaigns.)

Workers of all races saw declines in union membership. At 14.5 percent, African-Americans remained more likely to be in a union than white, Asian, or Hispanic workers, but union membership among blacks in 2006 still fell by 0.6 percentage points. Since 1983, the earliest year for which directly comparable data are available, union membership has decreased by 12.6 percentage points among blacks (from 27.1 percent in 1983), but dropped only 7.5 percentage points among whites (from 19.2 percent in 1983). (For longer-term trends in African-American unionization, see CEPR’s report, The Decline in African-American Representation in Unions and Auto Manufacturing, 1979-2004.)

Membership declines were roughly the same -- down about 0.5 percentage points -- for both men and women. In 2006, men (13.0 percent) were more often union members than women (10.9 percent), but over time, the unionization rates have been converging. In 1983, the earliest year for which directly comparable data are available, men (24.7 percent) were much more likely to be in a union than were women (14.6 percent).

The decline of unions within manufacturing was severe and will likely persist. In 2006, the number of unionized workers in manufacturing was nine percent lower than in 2005, a loss of 190,000 union members. Buyouts and early retirements of unionized auto workers throughout 2007 will lead to additional losses in union members, as will continued weakness in the manufacturing sector. Because of these declines, it is no longer accurate to view manufacturing work as a “union job.” Manufacturing workers are now less likely to be in a union than is the average U.S. worker.

The latest numbers continue a long-run decline in union membership. In 1983, about 1 in 5 workers in the United States was a member of a union, including almost 1 in 3 black men. By 2006, only 1 in 8 workers was a union member, and only about 1 in 6 black men.

In 1983, about 1 in 6 private-sector workers was in a union. Twenty-three years later, the share has fallen to about 1 in 14.

John Schmitt is a senior economist and Ben Zipperer is a research assistant at the Center for Economic and Policy Research in Washington, D.C. CEPR’s Union Membership Byte is published annually upon release of the Bureau of Labor Statistics Union Membership report. Data for years before 2005 are from the authors’ analysis of Current Population Survey data.

###