Download the full study here.

Executive Summary

Workers face an important decision when they are asked to vote on unionizing their workplace. A union will represent them in their interactions with their employer, affecting their wages, benefits, promotions, layoffs, work procedures and workplace dispute resolutions. In many states, workers will also owe some form of financial support to the union. If they formally join the union (as opposed to simply being represented by it), they will be subject to legally binding union rules regarding, among other things, whether to strike or cross a picket line. To help workers make an informed decision, this study reviews the effects of unionization, both in the private and public sector. The study’s emphasis, however, is on private-sector industrial unions.

Compensation. Unionized employees generally earn higher wages than nonunionized employees in comparable jobs — approximately 17 percent more on average between 1973 and 2001, according to a study by David Blanchflower and Alex Bryson. In the private sector, however, this percentage has generally decreased since 1973. Thus, private-sector workers joining a union may receive higher wages, but those wages may grow more slowly over time. Private-sector nonunion wages have grown more quickly on the whole from 1979 to 2011; the same is true of private-sector nonunion total compensation.

In the public sector, the wage advantage for unionized employees has increased somewhat, from an average of 13 percent in the period from 1983 to 1988 to an average of 15 percent from 1996 to 2001. Public-sector unions encounter less direct nonunion competition.

Union workers are more likely to receive fringe benefits, and generally, fringe benefits comprise a higher percentage of total compensation for a union worker than for a nonunion worker — 39 percent vs. 28 percent, respectively. Federal statistics gathered in 2006 show that union workers were much more likely to receive health insurance, short-term disability insurance, vision insurance and dental benefits than nonunion workers. They were somewhat more likely to receive paid leave of various kinds, including leave for funerals and jury duty, and they were also much more likely to participate in a traditional pension plan.

The value of many fringe benefits, such as health insurance, tends to be higher for older workers, and Richard Freeman and James Medoff found that unionization may raise the availability of paid leave for older workers more than for younger workers. In addition, the value of traditional pension plans to the worker may be somewhat tempered by the widespread underfunding of the plans, both in the private and public sectors, reducing the plans’ security. Providing retirement benefits can also cause a company to assume long-term financial liabilities that threaten its financial health. In a bankruptcy, pension benefits can be reduced, as can the number of jobs and level of compensation for current workers.

Contracts. Unionization does not guarantee a collectively bargained contract. An employer is compelled only to bargain in “good faith” with the union. One estimate suggests 15 percent of newly formed unions are unable to reach a contract within two years; another places the figure at 45 percent.

Layoffs. Layoffs can also affect compensation. In general, among males, union workers are more likely to be placed on temporary layoff than nonunion workers, based on Census Bureau data for 2002 through 2008. Male union workers are, however, less likely to be placed on permanent layoff than nonunion male workers, unless they are men ages 19 to 24. These young men may face higher permanent layoff rates in unions than outside them due to unions’ emphasis on seniority.

Employment. The higher total compensation that union workers receive — $11.14 more per hour in September 2011, according to federal labor statistics — is a cost to the company. Two separate studies found that the stock market responds to a successful union certification vote by lowering a company’s equity by more than $40,000 per worker. This drop in investor confidence signals an expectation that higher labor costs will lower a company’s profitability. Evidence for this view was on display during the bankruptcies of unionized companies like General Motors and Bethlehem Steel in the auto and steel industries, areas of traditional union strength.

Reduced profits affect a company’s ability to offer new jobs or even retain current ones. Over the past 60 years, nonunion employment gains have considerably exceeded union employment growth. Between 1950 and 2010, the total number of jobs in the U.S. economy nearly tripled from 45 million to 130 million, while union membership fell from 33 percent of the workforce to 11 percent. Despite public-sector union gains, total union membership is at its lowest level since 1950.

This decline reflects the end of the post-World War II dominance of large unionized firms that faced little domestic or foreign nonunion competition. Nonunion competition has accelerated since the 1960s, and although the resurgence of Japanese and European imports has played some role in this, domestic nonunion competition has been particularly vigorous. Changes in government policy toward imports are unlikely to affect this dynamic, nor can these declines be explained by changes in government policy toward private-sector unions.

In the public sector, where union employment rose by 8 percent between 1992 and 2010, nonunion employment still rose more quickly, by 22 percent. Even at specific levels of government — federal, state and local — since 1996, nonunion employment growth has outpaced that of union employment, which actually fell in the federal government, while rising more slowly at the state and local level. Hence, even in the public sector, union employment faces nonunion competition, including an increase in private contracting of government activities.

Strikes. The chances of a strike have declined considerably since 1948. From 1948 to 1979, more than 5 percent of U.S. union members typically went on strike in a given year. This percentage has generally declined since the 1980s and has been particularly low since 2001, with 0.6 percent, 0.5 percent and 1.2 percent of workers striking during 2005, 2006 and 2007, respectively. A 1983 survey of unionized companies indicates that they had become less afraid of strikes and more concerned about high and uncompetitive labor costs.

During a strike, a union can penalize a union member for crossing a picket line and continuing to work, though it cannot punish workers whom it represents but who have not formally joined it. Also during that time, workers do not receive pay from their employer; instead, they receive “strike pay” from the union. This strike pay is typically less than their usual income. During the American Axle strike in 2008, the United Auto Workers union reportedly disbursed strike pay of $200 per week.

Ideally, a union worker will recover any financial losses during a strike by the income from a better contract. The 1970 UAW strike at General Motors, for instance, won workers a milestone contract with generous retirement benefits.

Nevertheless, a better contract is not guaranteed. In the 2005 Northwest Airlines mechanics strike and the 2008 American Axle strike, workers ultimately received substantial wage and benefit cuts similar to what the company had proposed before the strike. The number of jobs available declined as well.

During a strike, a company’s customers may turn to other suppliers. During the 1959 United Steel Workers strike, American manufacturers increased their use of imported steel, and the United States became a net steel importer. This new comfort with imported steel created a weaker market for major U.S. steel manufacturers.

Grievance Systems. Unions frequently implement grievance systems to settle disputes between workers and management. These systems are often seen as an advantage to the worker. In 1984, Richard Freeman and James Medoff found that a union grievance system was equivalent to a 2 percent wage increase in lowering the rate at which employees left a company.

Nonunion companies have grievance systems, too. David Lewin, after surveying the academic literature, has suggested that they exist in about half of nonunion companies. Although data on grievance systems is somewhat limited, case studies suggest that compared to union systems, nonunion systems are about as effective at reducing employee turnover; similar in speed and thoroughness; and about as fair in choosing between workers and management.

Still, a worker using a grievance system may not restore his or her work situation to what it was prior to filing the grievance. Studies indicate that following a grievance, both a union worker and a nonunion worker tend to measure more poorly on a variety of job statistics, such as performance ratings, attendance rates and promotion rates.

Evidence also suggests that union workers are more dissatisfied with their jobs than union workers. This may be due to the problems that led to unionization in the first place, and unions may ameliorate some of the dissatisfaction workers might otherwise feel. Still, the data indicate that even after unionization, measurable dissatisfaction persists.

Work Rules. Unions bargain work rules into the contracts they bargain with management. These rules may limit the company’s ability to subcontract work to other firms, restrict the amount of weight an employee can lift or specify what work nonunion employees may not perform. Work rules in the UAW’s 2007 contract with Ford contributed to the contract’s length of 2,215 pages.

In a 1987 study of union and nonunion companies, Robert Kaufman and Roger Kaufman found that in 44 percent of union companies, employees without work to perform had to be reassigned in keeping with work rules, kept on the job without work or sent home with pay. Only 10 percent of nonunion companies had a similar restriction; the remainder reassigned the worker freely or sent him or her home without pay. The additional protection provided by unions may be attractive to a worker.

There can be costs to these rules, however. Kaufman and Kaufman also found that union companies typically had at least some restrictions on workers’ performing one another’s jobs and prevented supervisors from performing production work for purposes other than demonstration. Nonunion companies had no similar prohibitions. While the absence of such rules may seem to leave a worker unprotected, it also means a nonunion company gains flexibility to adjust to variations in workload and prevent production disruptions.

Without such flexibility, a nonunion company may become less competitive and profitable, jeopardizing line workers’ jobs regardless. A 2007 report by the consulting firm Harbour-Felax found that UAW work rules added hundreds of dollars to the cost of each of the Detroit automakers’ vehicles. A 2007 study by AlixPartners, another consulting firm, found that the three Detroit automakers maintained at least 8,200 more jobs than they would have if they had had the flexibility of Toyota, which is not unionized. Both studies were published not long before the bankruptcy of GM and Chrysler. Work rules regarding staffing levels at Bethlehem Steel also contributed to the shuttering of the company’s plant in Bethlehem, Penn.

Seniority. In union firms, workers with seniority typically receive preference in decisions about promotion and layoffs. This policy can contribute to a sense of job security for the most senior workers, but this security may come at junior workers’ expense. As noted earlier, male union workers ages 19 to 24 are more likely than their nonunion counterparts to face temporary or permanent layoff.

Similarly, Richard Freeman and James Medoff in 1984 cited a survey in which just one-quarter of managers in union companies said a junior employee would be promoted over a senior one if the junior employee were better; 43 percent said the junior worker would have to be “significantly better” and 33 percent said the junior worker would never be promoted over a senior one.

Obviously, junior and senior workers may view such restrictions differently. In the same way, workers who fear favoritism in management decisions or who are especially concerned about senior workers may prefer seniority rules, while workers used to promotions based on merit may find seniority rules frustrating.