THIS SERIES has described ways to address inequality: Increase tax progressivity; invest more in education; reform health care. But there's pressure to reach beyond that: to tackle inequality where it apparently originates, meaning the workplace. This pressure can be dangerous. Companies are not instruments of social policy; their first duty is to make money by serving customers, and they can provide for their workers only so long as they do that. Nevertheless, two sorts of corporate reform are warranted. It should be easier for labor unions to organize. And it should be harder for top executives to pay themselves outlandish sums.
Union membership has fallen from 20 percent of the workforce in 1980 to 13 percent in 2005, and part of this decline is inevitable. It reflects attrition in the manufacturing industries that are most easily organized. It reflects the rise of sophisticated human resource departments that provide workers with training, savings plans and grievance procedures -- usurping some of unions' traditional functions. And it reflects the deregulation of domestic industries such as trucking and airlines, plus tougher foreign competition. These forces spur businesses to innovate, but they also constrain their ability to make wage concessions to unions. In competitive markets, companies will pay workers what it takes to prevent them from being lured away by rivals -- and not more.
Yet the decline of organized labor also reflects a legal climate that is neither inevitable nor desirable. The way labor law is enforced now, employers can block attempts to establish unions by intimidating workers; a supervisor can summon an employee to daily meetings to discuss the dangers of unions or ban discussion of a union during work hours. If these tactics are not enough, employers can fire union organizers; although this is supposed to be illegal, the penalties are too feeble to serve as a deterrent. Meanwhile, a series of decisions from the National Labor Relations Board has narrowed the definition of workers who are eligible for union membership. Two months ago, for example, the three board members appointed by President Bush outvoted the two appointed by President Bill Clinton in ruling that relatively junior workers can be defined as "supervisors," thus restricting their right to join a union.
A fairer legal climate might reduce inequality slightly. According to David Card of the University of California at Berkeley, de-unionization explains about 15 percent of the increase in wage inequality among men over the past quarter-century. But the larger gain from reforming labor law would be political. Freedom of association is a core democratic right, and polls suggest that between 30 and 50 percent of nonunion workers would choose union representation if they had a chance to vote for it. The suppression of freedom of association is wrong in itself, and it fosters the suspicion that the rules of the economy are rigged against workers. Setting aside the debate over how much union membership can improve wages or benefits, the option of union membership is crucial to the legitimacy of capitalism.
The same goes for rules on executive compensation. Since 1970, the pay of chief executives has jumped from less than 30 times the average wage to almost 300 times that level. This helps explain why the richest 1 percent of Americans pocketed 21.6 percent of all the gains in national income between 1996 and 2001, according to Ian Dew-Becker of the National Bureau of Economic Research and Robert J. Gordon of Northwestern University. As with the decline of labor unions, some of the rise in executive compensation reflects market forces and is inevitable. Yet similar market forces are at work in other advanced nations, where executive pay has grown more modestly. In 2003, the ratio of U.S. chief executives' pay to that of manufacturing workers was more than double the norm in 13 other rich countries.
This reflects the way that bosses' pay is often set in the United States. Chief executives negotiate with a committee of board members whose independence is sometimes suspect, whose personal interests (particularly if they are CEOs of their own companies) may be served by rising executive-pay scales and who see little upside in risking a fight with the chief executive. In the absence of real discipline from compensation committees, CEOs can get away with pointing to the typical pay rate in their industry and asserting that they deserve a little more. The result is an inflationary spiral in executive compensation, unhinged from CEOs' real contribution to firms' performance.
What proportion of bosses' pay should be regarded as excessive? In a paper published last year, Harvard's Lucian Bebchuk and Cornell University's Yaniv Grinstein take a careful look at this question. They begin by noting that executive pay was already raising eyebrows back in 1993 and that it has nonetheless grown mightily since then. Then they observe that sales and profits of top companies have risen, which would tend to cause the bosses' pay to rise in tandem; and that an increasing share of the top companies are new-economy outfits, which tend to pay more. By analyzing the statistical relationship between executive pay and firms' size, profits and product mix, Mr. Bebchuk and Mr. Grinstein calculated how much compensation could have been expected to rise between 1993 and 2003. Their result: In 2003 the top five executives at the average public company could have been expected to earn a collective $6 million -- but they actually received almost twice that.
Overall, that means that the 1,500 companies studied "overpaid" a total of $8.7 billion in 2003 -- and this number is an understatement because it leaves out executive pensions, which are thought to have grown especially dramatically. If corporate governance reforms reestablished discipline over executive compensation, that excessive pay might shrink a bit. Inequality would decline, though only slightly -- the money would flow to shareholders, and more than three-quarters of all stocks are owned by the richest 10 percent of the population. But, as with labor law reform, the chief gain from corporate governance reform would be political. Executive overpayment running into the billions sends a terrible signal about the justice of the capitalist system.
Most critics of business are misguided. It is wrong to denounce managers who relocate factories to other countries or who fight to control wages; they are responding to market signals, as indeed they should. But when managers distort market forces by rigging the legal environment, that is a different matter. An entire industry of consultants exists to advise companies on how to avoid recognizing a union; a second industry of consultants exists to legitimize super-sized executive pay. Until this changes, the growing material inequality in the nation will be compounded by the corrosive perception that the rules are unequal, too.
This is the ninth editorial in an occasional series on inequality. Previous editorials in this series can be found athttp:/