Labor unionism’s decline in the United States is chronicled in the Bureau of Labor Statistics’ annual report on union membership, but the report only scratches the surface.
A far more detailed look is available on the “unionstats.com” Web site, where it is reported by state, metropolitan area, industry, and occupation.
The source of the data for both of these is the Census Bureau’s Current Population Survey (CPS) for the Bureau of Labor Statistics. The CPS involves about 60,000 interviews per month and at the national level is regarded as highly accurate.
The national public sector unionism numbers in the BLS report indicated that union density—the percent of the workforce in unions—in the public sector fell from 37.4 in 2009 to 36.2 in 2010. There were 21,133,000 public employees and 7,896,000 union members in 2009, which dropped to 21,033,000 public employees and 7,623,000 union members in 2010.
On www.unionstats.com, Barry Hirsh and David Macpherson, using data from the CPS, include state-by-state listings showing employment and union density for total, private, public, private construction, and private manufacturing jobs. It also includes data from selected metropolitan areas and detailed national data by industry and occupation.
The Public Service Research Foundation has taken the state data and developed a set of tables and charts showing employment, union membership, and union density from 1983 to 2010. Those are available on request by writing to [email protected]
Public sector union density varies greatly from state to state, from New York, at about 70 percent, to Mississippi, at about 9.5 percent. Between 2009 and 2010 the five-year average of public sector union density declined in some states and rose in others. The big gains and losses were in low-population states, leading to the conclusion that even the five-year average isn’t enough to correct for the distortions due to sampling error. In most of the larger population states, union density increased marginally.
When the data involve a subset of the workforce such as the public sector, in a lower-population state the sample size is much smaller and the degree of confidence in the findings suffers. Applying a five-year average to the data substantially improves the confidence in the figures and tends to smooth out apparent large differences from year to year.
A five-year average does tend to backload trends, so if the trend is downward the figures will be a bit higher than is actually the case, and if the trend is upward the figures will be a bit lower.
The coming year will reveal much about the future of public sector unionism. If unions fail to thwart efforts to roll back their monopoly bargaining privileges, will public employees be more or less inclined to support them?
David Denholm ([email protected]) is president of the Public Service Research Foundation, a not for profit research and education organization that studies labor unions and union influence on public policy.