STANFORD GRADUATE SCHOOL OF BUSINESS—What do Disney, AT&T, Exxon, and Verizon have in common? Based on economic performance and what they paid their CEOs from 1991-2002, a new academic study argues that all these firms were headed by CEOs who were paid too much.
These firms, say the researchers, are among a group of companies headed by CEOs whose pay is negatively related to job skill: The CEOs seem to be rewarded—in most cases, quite amply—for their bad performance. Disney’s Michael Eisner, for example, was paid $38 million above the industry average when for three out of six years the company’s performance actually declined in relation to other firms in the entertainment industry.
“Lately there have been legitimate concerns about CEO pay,” said Stanford Professor Robert Daines. In 1992, the average CEO of an S&P 500 firm earned $2.7 million. By 2000, average pay for these CEOs had increased more than 400 percent, to more than $14 million. When compared to the pay of average workers, the increase is even more dramatic: In 1992, CEOs were paid 82 times the average of blue-collar workers; in 2004, they were paid more than 400 times those salaries.
“Barry Bonds makes a lot of money because he’s a great baseball player. In general, the best-paid players are also the most skilled,” said Daines. “The main question is: Is the CEO labor market working in the same way? Do you make more money if you are better at it? Or is the market for CEO pay broken, in that CEOs receive high pay for something besides skill—like having friends on the board?”
Daines, the Pritzker Professor of Law and Business in the School of Law, who holds a courtesy appointment at the Business School, studied the relationship of executive pay to executive skill. In a recent paper, he and colleagues from the University of Pennsylvania and New York University concluded that particularly in big firms, a high salary doesn’t necessarily mean that a CEO is more competent than his or her peers.
The question first arose when Stanford Law School was doing a search for a new dean. Some faculty members wondered whether the choice of dean actually made a difference to the institution’s performance. “We wondered: Maybe the school will do roughly as well, no matter who is chosen as dean,” said Daines, adding it was a natural step from that question to ask whether individual CEOs mattered to the companies they lead.
First, the researchers defined a way to measure CEO skill. Although many studies have examined the relationship between CEO pay and company performance, Daines and his colleagues wanted to isolate specific evidence of executive competence. After all, company performance can depend on a number of factors beyond the power of an individual CEO: the economy, regulatory constraints, or industry conditions, to name just a few.
They decided that a firm run by a skilled CEO should consistently do better than its industry peers. More specifically, a skilled CEO should continue a firm’s prior good performance and reverse poor performance. Conversely, a bad CEO would be more likely to continue a poor showing and to reverse the firm’s prior successes.
The results were instructive: The study, which reviewed CEO pay and economic performance between 1991-2002, found that in small firms, highly paid CEOs generally are more skilled than their industry counterparts. The correlation is even stronger if the firm has a large shareholder or if the CEO has been paid largely in stocks and options. Conversely, pay is more likely to be negatively related to skill in larger firms. “In many large firms, the highest paid executives actually performed the worst,” said Daines.
This was especially the case when the business was bound by what Daines called “environmental constraints”: regulations, limits on capital spending, or a very competitive climate. “Some CEOs have a lot of options to consider while making investments and other strategic decisions. Others are more constrained,” Daines said. The lesson there? “When managers can do less to affect firm outcomes, there’s less reason to pay them high salaries,” said Daines.
In effect, the study identifies the kinds of firms where high pay would make sense, as well as those where the money is perhaps wasted. In larger firms, especially those with a lot of constraints on CEO choices and no large shareholder to keep an eye on things, it doesn’t appear to make much sense to pay exorbitantly high salaries. In smaller firms—;particularly companies with a large shareholder—;extra money paid to the CEO appears to be a much better investment, and highly paid CEOs appear to do better, suggesting that pay and skill are linked in such firms (as they are generally linked in professional baseball). In such firms, high compensation does not indicate a breakdown in the firm’s governance.
Indeed, one very interesting result of the study is that incentive pay matters enormously in cases where there is CEO turnover in poor performing companies. If the new CEO is paid more than his or her predecessor and if the pay is largely incentive-based, the new boss is more likely to reverse prior poor performance.
Finally, the study found that investors would do well to pay attention to which firms are paying their CEOs commensurate with skill. If, in fact, a company has hired a good CEO worth the money paid to him or her, then that firm represents a good investment opportunity. A portfolio that holds onto the stocks of firms with effective CEOs who are also highly paid—;and sells the stocks of companies run by poorly paid CEOs—;generates an annual rate of return of 8 percent above the market and other portfolios that control for risk characteristics. Thus, understanding which firms are run by highly skilled CEOs could be highly lucrative for investors deciding which stocks to hold or sell.
Daines suggested that this might be explained by the “cockroach” theory: Where you see one cockroach, you can assume that there are many. “Maybe paying the CEO too much in relation to his or her skill level is a sign of other, bigger problems with the company,” he said. “There’s evidence that information contained in CEO pay can potentially be used to understand investment in companies.”
One theory is that such CEOs are overpaid because they have too much influence over the board that should be monitoring them on shareholders’ behalf and too much influence over the committee that sets their pay. Thus, CEOs are effectively able to set their own pay and distort their compensation contract. In this view, CEO pay is the product of badly functioning corporate governance.
In addition, having a single large block holder of stock tends to ensure that CEO skill and pay are linked. “If you have a large shareholder, things seem to work better. Someone watching over management’s shoulder makes a difference,” said Daines. And in those cases where skill and pay were linked, incentive pay tended to strengthen that link. “Incentive pay does seem to work in many cases,” said Daines.
The Good, the Bad, and the Lucky: CEO Pay and Skill, Robert Daines, Vinay B. Nair, Lewis Kornhauser, forthcoming, 2005
“Compensation and Incentives: Practice vs. Theory,” George P.Baker, Michael C. Jensen, and Kevin J. Murphy, Journal of Finance, 1988, 63:3, pp. 593-616
“Are CEOs Rewarded for Luck? The Ones Without Principles Are,” Marianne Bertrand and Sendhil Mullainathan, Quarterly Journal of Economics, 2001, pp. 901-932
Strategic Leadership: Top Executives and Their Effects on Organizations, Sydney Finkelstein and Donald C. Hambrick, South-Western: 1996
Also on Stanford Knowledgebase: