CEO pay is clearly out of whack: Here’s how to fix it

November 19, 2012

In 1980, the average American CEO made about 42 times more than the average American worker. In 2011, the CEO made about 380 times more. CEO pay continued to rise significantly every year in the last half of the 2000s, even as corporate profits were falling. And therein lays the problem, say experts: CEO pay increases about 15-20 percent per year, seemingly unrelated to performance.

Fortunately, said Carr Bettis, research professor of finance at the W. P. Carey School of Business, people are finally talking about the issue of executive pay and calling for a relationship between actual pay and performance, as well as actual pay that is “reasonable” relative to comparative companies.

It’s not board capture that’s the problem

But getting to the place where pay is actually tied to performance and is “reasonable” necessitates an understanding of why pay has been increasing so dramatically, often out of sync with performance.

The first and most obvious answer would be “board capture.” That’s where the corporate board is made up of company executives who set pay -- their own pay -- high, explained Charles Elson, director of the John L. Weinberg Center for Corporate Governance at the University of Delaware. Elson and Bettis spoke at a meeting hosted by the W. P. Carey School of Business and the National Association of Corporate Directors.

But the board capture problem has largely been solved with new regulations like Dodd-Frank and Sarbanes-Oxley that (in part) call for independent boards made up of directors who have equity in the company and advised by independent advisors. So Elson and his colleague Craig Ferrere, a research fellow at the Weinberg Center for Corporate Governance, set out to find the real source of rising CEO pay. They report their findings in the paper “Executive Superstars, Peer Groups and Overcompensation: Cause, Effect and Solution.”

The real reason CEO pay has increased out of sync with performance, found Elson and Ferrere, is that it’s set based on a “peer group.” Boards use peer groups -- other companies ostensibly of similar size in the same or similar industries -- to set CEO pay. The use of the peer group leads to out-of-whack compensation for two reasons: First, because it can be easily gamed when boards include companies that are not true peers.

For example, in 2003 the New York Stock Exchange, then a nonprofit organization, established as its peer group highly profitable investment banks and huge financial institutions. That led to a recommended $140 million compensation package for then-chairman Richard Grasso. Including non-peer companies in a peer group can dramatically skew compensation, because CEOs of larger, more profitable, more complex companies, as well as those in certain industries, get paid substantially more than others.

The second reason that the use of the peer group leads to escalating pay is that in order to be competitive with their “peers,” boards aim to pay at or above the median. But when every board targets the 50th, 75th, or 90th percentile, that leads to the “ratcheting effect” where median executive pay escalates substantially (about 17 percent) every year.

Why the peer group doesn’t work: CEO skills are not transferable

The rationale for paying executives based on peer groups is the idea that if the CEO at Acme Co. is offered only $5 million in compensation, then he will leave and go to Widget Co., which pays $7 million. The crux of Elson and Ferrere’s research is that Acme’s CEO is actually not likely to be able -- or inclined -- to go work at Widget Co. CEO talents are not transferable, Elson and Ferrere found, so CEOs don’t easily move from company to company.

In a study of all of the S&P 500 companies over a period of five years, researchers found just six instances where a CEO jumped from one company to another. And in five of those six cases, “the CEOs went to work for dramatically larger and more complex companies -- ones that could not be defined as peers of the original company,” explained Ferrere.

Out of the finding that CEO talents are actually not transferable, Elson draws three conclusions: First, the “Superstar” theory (that great CEOs can be great anywhere) doesn’t hold. Second, the theory that CEOs will simply move to Widget Co. if they don’t like what Acme Co. has offered does not actually play out. And third, the concept of non-transferability of CEO talent -- the fact that a “great” CEO is one who knows the company intimately -- suggests that it’s far better to hire a CEO from within the company.

Use of the peer group tasks boards with a “messy job”

Paying CEOs according to a peer group when talents are not actually transferable, and CEOs not likely to jump ship for a better offer, can cause two problems for a company. First, when the CEO’s compensation increases 15-20 percent annually no matter how the company has performed, at the same time that the pay of other employees rises much less quickly (or not at all), this leads to declines in employee morale. And that can lead to poor employee performance -- which ultimately affects the customers.

The second problem created by the use of peer groups is that ultimately, paying executives out of sync with company performance will affect profitability. Said Elson, “If the CEO is not being paid what he or she earned, then it’s a gift, not compensation. And that’s a waste of company assets.”

That sticks corporate boards between a rock and a hard place, explained Bettis. On one hand, boards have long used compensation consultants to help them set executive compensation: ostensibly, the consultants have more insight into compensation trends, what works, and what doesn’t, given that it is their business. Yet, said Bettis, even among compensation consultants there is “a lack of science and a lack of results measurement.”

On the other hand, boards must remain compliant with new rules for executive compensation at public companies -- and that requires navigating a complex web of regulations. Proxy advisors, designed to help shareholders, also need better science and results measurement.

“All the way around, there really needs to be better science that shows what works, and what doesn’t,” Bettis said. Absent that, “it’s the boards that end up doing the heavy lifting to try and sort through the mess.”

What is pay?

The tricky question of how much to pay CEOs is only half the decision battle, said Bettis. The other half is determining how to pay them. “There is now a significant debate over the question: What is pay? There are concerns about how to align incentives correctly, and that’s an issue that needs discussion and evolution. What is needed is science and better data.”

Most of the focus to date, Bettis explained, has been on realizable pay, which measures not how much in stock and stock options the CEO has actually cashed in, but rather the potential value of such awards. The problem with realizable pay as a measure of executive compensation, said Bettis, is that every company calculates it a bit differently, so it’s not comparable across companies. And it’s complex, which makes it very difficult for board members and shareholders to understand.

The conversation around what form CEO compensation takes needs to answer this question: Are we getting what we want out of incentive alignment? Bettis said. Today, 56 percent of CEO compensation is in the form of long-term incentives. “But sometimes there is no relationship between executive compensation and the performance of the company.”

A better way to pay

However pay is defined, compensation must incentivize the kind of behavior that the board wants to see from the CEO. Said Bettis, “The executive needs to really know the value of achieving his or her targets. That’s hard when the targets are complex and there are a lot of them. So the concept of performance vesting is good but you have to demonstrate the value of the rewards.”

According to Bettis, the focus must be on changing CEO pay to align incentives with compensation and ensuring that the CEO “has a clear line of sight between actions and rewards.” Though they disagree on the details about what pay should look like, both Elson and Bettis agree to aim for simplicity -- to ensure that all stakeholders (board members, shareholders, and the CEO) understand what the incentives are, and what the rewards will be.

In order to pay better, they say, something will have to be done about the level of influence wielded by compensation consultants, whose own incentives are not aligned with those of the board or the company. Where boards should be seeking simplicity, compensation consultants are incentivized to introduce complexity -- to ensure that they’re needed to help boards and executives navigate increasingly complex compensation plans. “Boards have to work with the consultant to ensure that he or she is an effective representative of the company,” Elson said.

At the end of the day, boards need “better science all around to show what works,” said Bettis. “Are we delivering a compensation plan that the executive understands? Can the board articulate the value of the compensation plan to the shareholders?”

Bottom line

  • CEO pay increases about 15-20 percent per year, seemingly unrelated to performance. The real reason CEO pay has increased out of sync with performance is that it’s set based on a “peer group” -- other companies ostensibly of similar size in the same or similar industries used to set CEO pay.
  • The use of peer groups is problematic because they can be easily gamed when boards include companies that are not true peers; and because of the “ratcheting effect” -- when every board targets the 50th, 75th, or 90th percentile, median executive pay escalates substantially (about 17 percent) every year.
  • The rationale for paying CEOs based on the compensation of the CEOs of peer group companies is that if Acme Co. doesn’t pay competitively, the CEO will simply jump to Widget Co. In reality, CEO skills are not transferable -- they don’t jump from peer company to peer company.
  • Proxy advisors, compensations consultants, and boards themselves need better science and better data -- they need an understanding of what works, and what doesn’t, to align CEO incentives with corporate performance goals.
  • According to Carr Bettis, research professor of finance at the W. P. Carey School of Business, CEOs need to really know the value of achieving their targets -- they need a “clear line of sight between actions and rewards.”