Just Deserts

High Executive Pay Makes Economic Sense

The AFL-CIO has a Web site, www.paywatch.org, dedicated to condemning what it considers to be “runaway CEO pay” in private corporations. It claims that high executive pay damages all other “stakeholders” in corporations, especially workers and stockholders. In other words, the AFL-CIO asserts that excessively paid executives and other stakeholders are locked in a zero-sum struggle.

The Web site had 250,000 hits in 1997. William Patterson, director of the AFL-CIO Office of Investment, boasts that “We put a face on runaway pay in 1997.” Betsy Leondar-Wright, an official of United for a Fair Economy, a consortium of religious and activist organizations in league with the AFL-CIO, asserts, “There’s nothing like this issue. This one just galvanizes people instantly.”[1] Visitors to the site are invited to “fight back” by sending messages to boards of directors, using their clout as shareholders, staging workplace actions, and e-mailing members of Congress.

The site promotes enraged envy. Visitors are told that on average, CEOs make 326 times the pay of ordinary factory workers. The site offers a calculator for envious visitors to compute how long they would have to work to take in what their CEOs took in during the previous year.

The AFL-CIO’s agenda is clear: to arrest and reverse the declining market share of unionized labor. Unions represented only 9.8 percent of the private-sector work force in 1997. That figure has been declining since 1953 when it was 36 percent. The site asserts that “Workers have less power in today’s economy, which means that executives can take far more for themselves. That wasn’t always the case. When unions represented a third of the workforce in the 1950s and 1960s, workers had an effective counterweight to corporate greed.” A more powerful AFL-CIO will be able to seize the ill-gotten gains of CEOs and restore them to their rightful owners.

What’s really going on with executive compensation? In Chapter 10 of their new text for MBA students, Richard McKenzie and Dwight Lee survey the literature on executive compensation and conclude that, in general, the high compensation packages paid to executives can be explained by some simple economic theory.[2] Far from exploiting workers and stockholders, high executive compensation is a means for all corporate stakeholders to benefit. It is part of a positive-sum game. The authors’ explanations fall into two categories, which I call productivity and meta-productivity arguments.

The Productivity Argument

Top executives are the ones who control most resources in firms. Their decisions affect the performance of all other managers and employees. Excellent executives are scarce, and their talents are economized by putting them at the top where their productivity is multiplied through its effect on many others.

General Electric’s CEO, Jack Welch, made 1,003 times the pay of the average U.S. factory worker in 1996. His total compensation was $27.6 million in that year. United for a Fair Economy submitted a shareholder resolution that asked the GE board to set “a cap on CEO compensation expressed as a multiple of pay of the lowest-paid worker at GE.” GE responded by arguing that Welch’s compensation “is appropriate in light of the value that his superior leadership, vision and dedication provided to the share owners during his 17-year tenure as chief executive officer.” The total value of GE shares rose by more than $225 billion during that period.[3] Welch earned what he was paid. Moreover, increasing shareholder wealth is good for employees as well as stockholders. A successful corporation provides higher wages and salaries and offers more job security than an unsuccessful corporation.

At Scott Paper, CEO Al Dunlap rescued the firm through radical downsizing and restructuring. As a result he produced $6.5 billion in additional shareholder wealth. His compensation was only $100 million—less that two percent of what he produced. Dunlap asks, “Did I earn that? Damn right I did. I’m a superstar in my field, much like Michael Jordan in basketball and Bruce Springsteen in rock’n'roll.” He says, “You cannot overpay a good CEO and you can’t underpay a bad one. . . . If his compensation is not tied to the shareholders’ returns, then everyone’s playing a fools game.”[4]

Of course, in the face of imperfect knowledge mistakes can be made, and McKenzie and Lee illustrate such failures. However, when mistakes are discovered, they usually are rapidly corrected. Corporations that allow failed executives to keep their high-paying jobs soon become ideal candidates for hostile takeovers. The market for corporate control never rests. When hostile corporate raiders gain control, they show the door to failed executives.

Sometimes failed executives leave their firms with very golden parachutes. Michael Ovitz walked away from a 14-month tenure with Disney with a $90 million severance package, and it took Gilbert Amelio only 17 months on the job at Apple to get $7 million.[5] More recently, David Coulter was dismissed from Bank of America, after two years on the job and large hedge fund losses, with over $40 million as a consolation prize. Golden parachutes like these are likely to seem excessive to everyone except the recipients, but McKenzie and Lee suggest an explanation.

A newly hired top executive is likely to be much more risk averse with respect to the activities of the firm than its stockholders are. This is because stockholders have diversified portfolios; they own shares in many different firms. If one gets into trouble, perhaps others can make up the losses. In contrast, the executive has all his human capital tied up in the firm. This one firm is the executive’s sole employment, and he will be reluctant to undertake risky, but possibly high-yielding, ventures. By consenting to golden parachutes in executive hiring contracts, boards of directors may just be trying to overcome excessive executive risk aversion. On this reading, golden parachutes merely align executive and shareholder attitudes toward risk.

McKenzie and Lee report on several systematic empirical studies of the connection between executive pay and executive productivity. Consider three examples. Sherwin Rosen’s research indicates that on average, when a company’s rate of return increases by 1 percent, top executive pay increases by 1 to 1.25 percent.[6] Inasmuch as executive pay is a small fraction of company income, those executives earn their keep. Michael Jensen and Kevin Murphy say that on average, executives get only two cents more in cash pay and $3.25 in added wealth for every $1,000 they add to stock values. Murphy concludes that “top executives are worth every nickel they get.”[7] James Brickley, Sanjai Bhagat, and Ronald Lease report that on average, stock prices increase by 2.4 percent within two months of the adoption of executive incentive pay over what they would otherwise have been.[8]

The most common form of incentive pay for top executives is stock options, which are rights to purchase shares at a fixed, low price. If the executive is successful, the market price of shares will rise, and he will reap large capital gains. This aligns the interests of the executives and the stockholders. On average, stock options make up two-thirds of top executive compensation. That means that two-thirds is at risk. Therefore, part of the high yields received through stock options may be considered just a risk premium.

Meta-Productivity Arguments

Yet many top executives get compensation that cannot be explained by the productivity argument. They get paid more than they are worth in the narrow productivity sense. McKenzie and Lee offer three arguments to explain such “overpayments.”

Discovery. Knowledge of who has the ability to become a successful top manager is given to no one. Within a firm top executives have to discover who among the ranks of lower-level managers is deserving of significant promotion. When such a person is identified and promoted, his discovery is announced to the general business community. If the promoted manager were paid only what could be justified by his direct productivity, executive raiders from other firms could easily bid for his services. In doing so, these outside bidders would be avoiding their own costs of search for executive talent. Therefore, a firm that promotes an executive is well advised to pay more than his productivity can justify to be assured of keeping the executive and conserving on future search costs for executive talent.

An interesting implication of the discovery argument is that some people who are not promoted may be just as good as those who are. They are simply undiscovered. To be discovered, it may pay for lower-level managers to acquire self-promotion and schmoozing abilities as well as basic skills for the job. It is possible, McKenzie and Lee write, to acquire “too much in the way of basic skills and not enough of, say, political savvy.”[9]

Self-Monitoring. It is very difficult, and often impossible, to monitor the performance of top executives accurately. A good executive doesn’t follow any prescribed routine that a monitor could observe. A good executive is alert to possibilities of turning all sorts of daily encounters, on and off the job, into profit opportunities. One may say a really successful executive is one who knows how to make it up as he goes along. The only effective monitoring of an executive is self-monitoring. An executive can be induced to self-monitor by a compensation package that substantially exceeds his opportunity cost (worth in the market). An executive who knows that shirking, once detected, means losing the difference between current compensation and what other boards of directors are willing to pay is likely to avoid shirking. This implies that the amount of the overpayment will be positively related to the difficulty of detecting shirking. The more subtle the executive’s operations, the higher the necessary overpayment.

Tournaments. One way for top executives to motivate lower-level managers to work as hard and smart as they can is to make clear that rungs on the managerial ladder get increasingly narrow as it is climbed. Firms following this strategy purposefully hire many more low-level managers than they can possibly promote, and they make sure that new hires understand that this is true. Promotions are awarded to the few who can successfully compete for them. At each stage the tournament prize is a promotion and a substantial increase in pay that cannot be justified on narrow productivity grounds. Why does the pay increase have to be so large? Because the contestants will naturally discount the increase by the probability of achieving it. The higher up one is on the managerial ladder, the lower the probability of climbing still farther. Therefore, the gaps in pay must grow as one climbs from one rung to the next.

The institution of a tournament has at least two effects among the contestants. First, those who think they have little chance of winning will self-select out of the contest. Job seekers will not accept jobs with the firm to begin with, and those who are already with the firm will leave and seek more accommodating employment elsewhere. Those with substantial self-confidence will remain. This increases average productivity. Second, competition among the contestants who are willing to play the game will make all of them exert more effort and avoid shirking. All contestants will be more productive than they would have been without the tournament. The “overpayments” that accompany promotion are paid from the added output of the enterprise as a whole rather than the individual productivity of the prize winners.

Gaps in compensation between top executives and average workers are increasing. The tournament model explains this. Deregulation and globalization of competition have flattened firms’ managerial hierarchies. Middle-level management positions are being eliminated and replaced with floor-level team decision-making monitored increasingly by pay-for-performance schemes. These changes in organizational architecture enable firms to respond more quickly to rapidly changing market conditions. The changes also mean the managerial ladder is becoming narrower faster. Thus the probability of lower-level managers being promoted to top executive positions has substantially decreased. To keep a tournament viable under these circumstances, the gaps between top-level and average compensation must grow.

It should not be surprising that the AFL-CIO violates the Tenth Commandment’s proscription against envy. Unions have never been reluctant to exploit misunderstanding of economics for their own gain. The executive pay issue is just the most recent example. The only remedy is for people who do understand economics to expose the spurious reasoning of the enemies of freedom. We should be grateful to McKenzie and Lee for having done so in this case.


  1. Robert L. Rose, “Labor Has Discovered the Perfect Issue For Galvanizing Workers: CEO Pay,” Wall Street Journal Interactive Edition, April 9, 1998.
  2. Richard B. McKenzie and Dwight R. Lee, Managing Through Incentives (New York: Oxford University Press, 1998).
  3. Rose.
  4. McKenzie and Lee, p. 160.
  5. Ibid., p. 171.
  6. Ibid., p. 161.
  7. Ibid., pp. 162–63.
  8. Ibid., p. 164.
  9. Ibid., p. 165.