technocratic branch. They are coming out on top in market economies, too. It is the MBAs on the road to class power.
Under communism, the rise of the intelligentsia was undeniably a political process. But the academic theory underpinning the rise of the MBA class in the West is all about market forces. The goal of the pay-for-performance revolution, after all, wasn’t to raise CEO compensation, although that was certainly one of its consequences. The point was to get the managerial aristocrats to do a better job by more closely tying their paychecks to their impact. On this reading, the soaring salaries of CEOs, and the growing gap between them and their senior lieutenants, is one chapter in the broader story of superstar economics. Once companies began to do a better job of tying pay to performance, they discovered that some managers were more talented than others, and those stars, like the best singers or lawyers or chefs, could command a significant financial premium.
For the CEO to be a superstar—and to be paid like one—he has to stop being a company man. The executives of the postwar era were corporate lifers. They were the creations and the servants of their companies, and a great deal of their value came from their knowledge of the particular corporate cultures that had created them and the specific business they did. The superstar CEO cannot be tied to a single corporation and, ideally, not even to a single industry. He must be an exemplary talent whose skill is in “management” or in “leadership.” He is more likely to have an MBA—28.7 percent of CEOs did in the 1990s, compared to 13.8 percent in the 1970s—and less likely to be a loyalist of a specific firm. If these are the general skills we believe it takes to lead successful businesses, the world’s companies will engage in a bidding war to secure the services of the men and women who are the world’s best managers and leaders.
That is exactly what has happened. The surge in CEO salaries coincided with a rise in bosses hired from outside the firm. In the seventies and eighties, when CEOs were paid less than they had been in the 1930s, 85.1 percent and then 82.8 percent of chief executives were company men. But in the 1990s, as CEO compensation came to vault upward by 10 percent a year, more than a quarter of chief executives were appointed from outside their firm. Jumping to a new company was a good way to get a raise—external hires, according to research by Kevin Murphy, one of the leaders of the pay-for-performance school, made 21.6 percent more than chiefs appointed from inside. In sectors where these portable general managers are most valued, all CEOs earn more—a premium of 13 percent.
One of the drivers of superstar incomes in other professions is the economics of scale—singers who can perform for millions, designers whose styles can be sold around the world. Size can be a reason to pay CEOs more, too. As companies get bigger thanks to the globalization and technology revolutions, the economic impact of good management increases. The world’s very best CEO may be only marginally better than the hundredth best. But if your company’s annual revenues are, say, $10 billion, then just a 1 percent difference in performance is worth $100 million. Sure enough, as economists Xavier Gabaix and Augustin Landier found in a 2008 paper, “The six-fold increase of U.S. CEO pay between 1980 and 2003 can be fully attributed to the six-fold increase in market capitalization of large companies during that period.”
But there is one very big problem with the superstar CEO model, and it goes back to the challenge posed by the rise of the managerial aristocracy that first Berle and Means and later the pay-for-performance school grappled with. It is what economists call the agency problem, and it means that CEOs are a very special sort of superstar: the one who is in charge of the company that pays his salary. Superstar athletes are paid by the owners of sports teams, superstar chefs by their diners, and even superstar hedge fund managers are paid by their investors. But CEOs are paid by the companies they run. Their compensation, to be sure, is determined by the board of directors, but, particularly in the United States, the chairman of the board is often the CEO.
“In the U.S., you can more or less do whatever you want, without having the support of the owners,” Mats Andersson, the chief executive officer of the Fourth Swedish National Pension Fund and critic of corporate governance in the United States, told me after speaking at a conference on the issue convened in Washington by the Securities and Exchange Commission. “Because of the composition of the boards in Sweden, the company’s big decisions all have to be based on the mandate or the support of the owners.
“Who is actually responsible for executive remuneration in U.S. companies?” Mr. Andersson said. “If I could decide on my own salary, I would certainly love that system.”
Adam Smith forthrightly warned that the consequence of the agency problem was “negligence and profusion.” Academic economists today use a more delicate term: “skimming.” A decade ago, two young economists, Marianne Bertrand and Sendhil Mullainathan, came up with an original way to investigate whether CEOs were superstars, being rewarded by their firms for exceptional performance, or whether they were stewards who were rigging the rules of the game in their own favor. The test was to see whether performance-based CEO pay responded as strongly to external good fortune as it did to managerial prowess. Two of the examples of outside luck were changes in the price of oil and changes in the exchange rate. Bertrand and Mullainathan found that luck matters: “CEO pay is as sensitive to a lucky dollar as to a general dollar,” which is to say for overall good company performance. They found, for instance, that a 1 percent increase in the revenues of oil companies because of an increase in the price of oil led to a 2.15 percent increase in CEO pay. Better still, from the perspective of the oil chief, while an increase in the price of oil always correlated with an increase in the CEO’s paycheck, when the price fell, the CEO’s salary didn’t necessarily decline: “While CEOs are always rewarded for good luck, they may not always be punished for bad luck.”
Thanks to the financial crisis and the global recession it triggered, public opinion and politics in much of the world are catching up to this ivory tower critique. Consider Britain, where a Conservative-dominated coalition government began 2012 with a proposal to rein in executive pay. “We cannot continue to see chief executives’ pay rising at 13 percent a year while the performance of companies on the stock exchange languishes well behind,” Vince Cable, the business secretary, told parliament as he announced the new measures. “And we can’t accept top pay rising at five times the rate of average workers’ pay, as it did last year.”
Cable’s reference to the gap between CEO salaries and those of average workers is telling. We may frame our complaints about rising executive compensation with arguments about skimming—that the millions are unearned. But part of our unease stems from something entirely different—that the final outcome, the gap between CEOs and the rank and file, is wrong.
This second concern may very well not be solved by doing a better job of coping with the agency problem. Bertrand and Mullainathan’s finding that there is a lot of skimming going on in the corner office doesn’t, it turns out, make them complete skeptics of the pay-for-performance revolution. Pay for performance actually works, but only in companies where the board is strong enough to truly oversee the chief executive. Boards are best able to do that, Bertrand and Mullainathan discovered, when they have a large shareholder. “An additional large shareholder on the board reduces pay for luck by between 23 and 33 percent”—a big number, especially when you consider how tricky it is in real life and in real time to distinguish between lucky profits and hard-earned ones.
There’s a reason for CEOs to position themselves as superstars—highly talented people being paid for their skill—that goes beyond getting a great deal from the comp committee. Even in an age of tension between the 99 percent and the 1 percent, we love our superstars. That’s because, as the
FOUR
RESPONDING TO REVOLUTION
