Labor unions have been criticized for decades because of a belief that they reduce overall employment: By mandating higher wages for current workers, they leave companies with less cash for new hiring. Now, new research is challenging that assumption.

Economics course books teach that for ordinary goods, an increase in prices tends to reduce demand. The same goes for workers. As the cost of labor rises in a particular area, employers gain incentive to move, outsource or use technology to replace jobs.

Unions certainly raise the cost of labor. According to the U.S. Bureau of Labor Statistics, the average weekly wage for a union worker last year was 28% higher than for a non-union one.

“Right-to-Work States ”

“* Alabama ”

“* Arizona”

“* Arkansas”

“* Florida”

“* Georgia”

“* Idaho ”

“* Iowa ”

“* Kansas ”

“* Louisiana ”

“* Mississippi”

“* Nebraska”

“* Nevada ”

“* North Carolina”

“* North Dakota”

“* Oklahoma”

“* South Carolina”

“* South Dakota”

“* Tennessee”

“* Texas”

“* Utah”

“* Virginia”

“* Wyoming”

There's plenty of anecdotal evidence that unions cost jobs. Heavily unionized firms like airlines, car makers and steel producers have struggled financially in recent years and even succumbed to bankruptcy. Union membership has declined from 20% of the workforce in 1983 to just 12% last year. The government now employs more union members than the private sector seemingly because it needn't worry about profits. Just this month, Harley Davidson told the Associated Press it might have to leave Milwaukee after 107 years if it can't being down labor costs (although some call that threat a tactic in the company's ongoing union negotiations).

Yet it's difficult to prove a cause-and-effect relationship between unions and struggling businesses. Unions might simply have thrived in industries that are subject to boom and bust cycles.

State comparisons should offer clearer evidence of whether such a relationship exists because their rates of unionization vary dramatically from a high of 25% in New York to a low of 3% in North Carolina. States also can be roughly divided into two groups, union-friendly and not, thanks to the 1947 Taft-Hartley amendments to the 1935 National Labor Relations Act. The amendments allow states to pass so-called right-to-work laws, which prohibit agreements between unions and employers making union membership or the payment of dues a condition of employment. Twelve states passed such laws soon after Taft-Hartley, and 10 more have done so since, for a total of 22. Without compulsory fees, unions in RTW states have a difficult time growing or even maintaining their memberships.

Union advocates prefer the term "right to fire" to "right to work," referring to what they see as an erosion of worker protection. RTW proponents use the term "forced unionism states" to describe states that haven't yet adopted the legislation. (Technically, workers in pro-union states may become objecting non-members. They must still pay fees, but only the portion that unions can prove goes toward collective bargaining, and not, say, political campaigns. Those are commonly referred to as Beck rights, after the 1975 Supreme Court ruling in Communication Workers v. Beck that created them.)

The National Right to Work Legal Defense Foundation, an advocacy group, has for years reported that economic trends in RTW states are healthier than in pro-union states. For example, it finds that "earnings and disposable income, adjusted for the cost of living, are highest in low-unionization states." It also finds that "tax-paying families are fleeing forced-unionism states." (The latter point is made in a recent Cato Journal paper, although the author, Ohio University professor Richard Vedder, also wrote that the evidence is "not a rigorous analysis of the relationship between right to work laws and human behavior." Six RTW states have no personal income tax, he notes, which could easily explain their appeal to migrating workers.)

Recently, a new dissenting voice on the relationship between unions and employment has emerged. It comes from a statistician.

In July 2007, Lonnie Stevans, a Hofstra professor who teaches quantitative study methods, was vacationing in Oregon when he picked up a hotel copy of USA Today. The headline, "Unions Try to Hang On as Open Shop Laws Gain Ground," caught his attention. (An "open shop" is a workplace that doesn't require union fees as a condition of employ. An "agency shop" is one that does.)

"After reading the article and living in Oklahoma for a number of years, it did not appear to me that workers were better-off in right-to-work states," says Stevans. "Yet, there were numerous assertions about the superior economic position of these states vis- -vis those with agency shops, so I wanted to see what the empirical evidence was."

Stevans says he found little such evidence, so he gathered studies and data and wrote his own paper, which was published last year in the Review of Law and Economics. Crucial to the process was controlling for state differences that aren't related to RTW laws, he says. For example, many Southern and Western states that have adopted such laws have a lower cost of living, higher rates of farm and service employment and greater population growth than Northern states that haven't adopted them, which can affect economic indicators.

After controlling for such differences, Stevans found that RTW states had lower per-capita income for workers, but higher income for business owners, than non-RTW states. He found that RTW states also had fewer business bankruptcies. He found no link between RTW legislation and either capital formation or rates of employment.

According to Stevans, there's little evidence of a "trickle down" of benefits in RTW states from business owners to workers.

So how is it that higher labor costs associated with unions in some states haven't reduced their employment, all else held equal? One possible explanation is that workers with higher pay spend more within the local economy, supporting employment nearby, Stevans says.

Asked for comment, a spokesman for the National Right to Work Foundation said the group would need time to examine Stevans's research, but that he didn't see anything that contradicts the group's statements. "Our argument doesn't rest on the financial benefits of workers being able to decide whether to pay fees to a union," he says. "Ultimately, it's a matter of liberty, not economics."