Are America's unions committing suicide? It seems an odd question to ask in the wake of last week's settlement of the United Auto Workers' prolonged strike against General Motors. After all, GM failed in one of its chief objectives: doing away with outdated piecework rules that, combined with improvements in assembly-line efficiency, have resulted in workers often toiling just half a day for full pay. The automobile manufacturer won only modest increases in required productivity, so that a hard-working welder will still be able to complete his quota in as few as five hours.
Such expensive labor practices imperil the long-term health of GM and other unionized companies--and therein lies the rub for organized labor. Companies and industries in which unions have significant power have been steadily declining, and union membership has declined along with them. According to data from the Bureau of Labor Statistics, the percentage of the private work force that is unionized declined to less than 10% last year from about 27% in the early 1950s. (Public-sector union membership, however, is growing.)
Unions drive up labor costs. Unionized workers earn about 15% more on average than nonunionized workers, according to the National Bureau of Economic Research. And rigid work rules like the ones at GM force companies to hire more workers. If a company has a monopoly on its product, it can simply pass the higher costs on to consumers. But if other, nonunionized firms sell exactly the same product, then the unionized firm may well go bankrupt quickly. In the real world, most unionized companies are, like GM, somewhere in between. They have significant market power, but not enough that they don't have to worry about competition.
Studies of unionized companies find that they are generally unable to overcome the costs of unionization. Economists have investigated two potential avenues of adjustment: Unionized firms could purchase more machines, or they could expand by buying nonunionized firms outright. Neither approach seems to work.
Buying more machines has an intuitive appeal. Since unions increase wages, unionized companies could, in theory, automate their way to lower costs and thereby at least partly overcome the higher costs of union wages. In a forthcoming article in the Journal of Labor Economics, Federal Reserve economist Bruce C. Fallick and I show that this strategy has not helped U.S. companies. When more efficient machines increase a firm's profits, unions increase their wage demands commensurately, canceling out the benefits from the machines. Because of this, unionized firms are in practice less willing to purchase new machines than nonunionized firms. We found that firms generally invest about 30% less in automation when their workers are unionized.
Alternatively, unionized companies have been accused of acquiring nonunion firms to get around higher wages and union work rules: If your plant in Massachusetts organizes, buy a firm that operates in Mississippi, and threaten to transfer work to the Southern plant the next time you bargain with the Massachusetts union.
In fact, this doesn't happen often either. In another recent paper, Mr. Fallick and I looked at a large sample of U.S. mergers. We found that unionized companies almost always merge with other unionized companies, while nonunion companies almost always merge with nonunion companies.
A simple example illustrates why. Consider a convenience store with a nice location, nonunionized work force, and a market value of $1 million. If the workers unionized, that market value would go down, since higher labor costs would mean lower current and future profits. If a unionized company thinks about acquiring a nonunionized convenience store, it will worry that its union will spread to the new store--so that, shortly after the company has paid $1 million for the new store, its value would go down. If the company buys a unionized store instead, there is less of a chance the asset will decline in value.
Another way a company can adjust to higher union costs is to outsource production to nonunion or overseas suppliers. This strategy may well help the company survive, but it doesn't do much to ensure the survival of unions in America.
So it is easy to see why, in retrospect, the number of union employees in America has gradually declined. There is little that companies can do to overcome the cost disadvantage imposed by higher union wages. American workers seem to understand this, which is the final piece of the declining unionization puzzle. After witnessing such disruptive absurdities as the GM strike, most Americans want little to do with unions, and efforts to unionize new plants are rarely successful. A recent paper by economists Henry Farber and Alan Krueger documents a striking decline in American workers' demand for unionization.
Messrs. Farber and Krueger hypothesize that Americans have become more satisfied with their jobs and less convinced that unions provide valuable services. When workers were working 12 hours a day, seven days a week over hot furnaces, the union movement had the moral high ground. But today unions pursue old strategies that are unreasonable, confrontational and ultimately self-destructive. Unless unions change their approach, they will continue to dwindle--and their demise will have been self-inflicted.
Kevin A. Hassett is a resident scholar at AEI.