Berkeleyan
Study finds little union impact on company survival, wages
| 22 September 2004
Despite popular claims to the contrary, labor unions in recent years have had little impact on either company survival or average wages in private sector manufacturing, according to researchers from UC Berkeley and the University of Michigan.
In a working paper for the private, nonprofit and nonpartisan National Bureau of Economic Research, UC Berkeley Assistant Professor of Economics David Lee and University of Michigan Professor of Economics John DiNardo analyzed firms where union-organizing drives narrowly won or lost from 1984 to 1999.
Detailed examination of data about these businesses — about 10 percent of them in California — showed little union impact in terms of worker productivity or hours, manufacturing output, total value of shipments, sales volume, the number of company employees, or a firm’s ability to stay in business.
On a related note, the researchers also found no evidence that an employer facing increased risk of unionization will boost wages to undermine the effort.
“The private sector doesn’t paint a pretty picture for the labor movement, and nothing indicates a rebound,” said Lee.
But the researchers’ results are a double-edged sword. They also refute what Lee and DiNardo said is a frequent employer stance: that a plant or business operation will have to close to absorb additional costs if a union drive is successful.
While Lee and DiNardo do not identify a single explanation for why new unions have had trouble achieving wage gains, possible factors they list include:
• union inability to win significant wage gains through collective bargaining;
• a decline in union membership and new organizing activity;
• increased managerial opposition to unions;
• increased use of replacement workers;
• outsourcing.
“The ’80s were a tough period for the unions,” said Lee, a labor economist. “Some Gallup polls show people are less concerned about ‘Big Labor,’ and maybe that reflects this idea that unions are weaker than they used to be.”
For example, one major union blow came with the 1981 strike by 13,000 members of the Professional Air Traffic Controllers Association. After 48 hours, President Reagan fired 70 percent of the controllers who walked out, and decreed that they could never be rehired by the federal government. Lee and DiNardo said many union leaders felt this had a chilling effect on other organized labor groups’ ability to play their biggest card — threatening to strike.
“During the 1980s, prominent unions were accepting wage cuts, facing the pressures of the opening of international competition,” according to the Lee-DiNardo report.
Historically, authorities have reported a gap of 15 percent to 30 percent between union and non-union wages, but Lee said he and DiNardo found a difference of closer to 1 percent. Why?
Typical surveys about union wages, they said, tabulate random households and how much an individual union worker makes compared to a non-union worker. But they don’t take into account that unionized establishments may be very different from non-unionized workplaces.
Lee and DiNardo concentrated on business information and union efforts over the last 20 years. They looked at data for more than 25,000 establishments, about 5,000 of them manufacturing. The manufacturing information included wage data while the rest did not.
Their sources included the National Labor Relations Board electronic election database, Federal Mediation and Conciliation Service records, and information about employment, sales volume, wages, output, hours, shipments, and assets from the Census of Manufacturers and the Annual Survey of Manufacturers.
Lee noted the findings don’t address many benefits of union membership — such as the 40-hour work week, seniority systems, safety regulations, and the ability to file grievances — that have been largely institutionalized since unionization began in the 1930s.
Their paper, “Economic Impacts of Unionization on Private Sector Employers: 1984-2001,” can be found online at www.nber.org. The study also will be published in the November issue of the Quarterly Journal of Economics.
They hope to expand their investigation, looking backward at the effect of new unions established before the 1980s, as well as assessing the effect of unions on workplace safety and other important benefits not related to wages.