Setting the CEO’s salary is one of the most important duties of a public company’s board. So CBS directors decided to give Chief Executive Officer Leslie Moonves a $69.9 million pay package last year only after assessing the competitive market for senior executive talent. The board of directors, however, looked at companies that are, on average, more than twice as large as CBS and included many in businesses far afield from media.
Such lopsided comparisons aren’t unusual at U.S. corporations. To justify how much they offer CEOs, board compensation committees typically measure against the pay of CEOs at other companies. But they routinely pick larger companies, often outside their industry, that pay more and aren’t rivals for the same executive talent. “There is pretty straightforward evidence of cherry-picking, and it’s pervasive,” says Thomas DiPrete, a Columbia University sociologist who has studied the use of board-selected peers to set CEO pay.
In the 1980s and ’90s, when executive compensation rose dramatically and often angered investors, pay consultants sold many companies on the practice of benchmarking against peers. The idea was to compare compensation with that of similar-size companies in similar industries to determine a fair compensation level. The problem is that the choice of peer groups can itself seem arbitrary. Four separate groups of academic researchers have found what they consider to be evidence of bias in the peer groups that U.S. boards use to set pay. Another found no bias. None of the studies identify companies using skewed peer groups by name.
Intentionally picking better-paid peers “has been a big part of the ratcheting up of executive pay,” says Bill George, a Harvard Business School governance professor who’s a director at ExxonMobil and Goldman Sachs and a former CEO of Medtronic. The adjusted average compensation of CEOs in the Standard & Poor’s 500-stock index rose to $12.9 million in 2011, or 380 times the average worker’s pay. That’s up from $625,000, or 42 times average worker pay in 1980, the AFL-CIO said in a report released in April.
Many directors “have incentives to make compensation decisions that are more favorable to executives” than to shareholders, says Lucian Bebchuk, a Harvard Law School professor who has researched CEO pay. One reason: Directors may not want to risk upsetting their board position or their relationship with a CEO, he says. Companies that use peer groups to help set compensation—one study found that 89 percent do—have been required to disclose them by the Securities and Exchange Commission since 2006.
Most companies vary pay based on performance, yet still use competitive peer data to determine the size of a CEO’s target compensation, set at the beginning of a company’s fiscal year. This includes salary, a bonus target whose final amount varies based on annual performance, and an award of stock or options. Directors at CBS measured pay for Moonves against that of CEOs at five other media companies, all of them larger by market value and all but one larger by sales. Directors also look at a bigger group of businesses in a variety of industries, such as IBM and General Electric. The median company on both lists is more than twice the size of CBS by both sales and market value.
GE had 11 times CBS’s sales, based on the latest figures available when CBS determined pay, while IBM had 14 times CBS’s market capitalization. Neither company compares itself with CBS in setting its pay. GE CEO Jeffrey Immelt had total compensation of $21.6 million last year, and former IBM CEO Sam Palmisano, who stepped down in December, had compensation of $31.8 million for 2011. CBS in a statement said benchmarking is “just one data point” it uses when setting CEO pay, which varies based on performance.
CBS’s corporate sibling Viacom, from which it was separated in 2005, uses almost the same groups of outsize peers. The median company on Viacom’s media-industry peer group is 56 percent bigger by market cap and more than twice as big by sales; the median company in its broader peer group also has more than twice Viacom’s sales and market value. Viacom said in a statement that it doesn’t set pay directly based on its chosen peers but “uses this information to get a broader perspective of overall compensation practices.” Sumner Redstone, who is executive chairman of both companies and controls them through a special class of stock, made a combined $41 million in pay last year. Viacom CEO Philippe Dauman collected $43 million.
Some companies measure CEO pay against executives in completely different industries—even as they gauge their corporate financial performance by looking within their own sector. Monsanto, the world’s largest seed company and the maker of Roundup herbicide, compares its financial performance with DuPont, Dow Chemical, and other chemicals and seed-producing rivals. Its board also used to focus on these companies when they set executive compensation.
But since at least 2006, Monsanto’s board stacks up CEO Hugh Grant’s pay mostly against leaders at health-care companies such as Abbott Laboratories and Eli Lilly. They’re among the best-paid on Monsanto’s list, including $24 million for Abbott’s Miles White and $16.4 million for Lilly’s John Lechleiter in 2011. Grant’s pay has risen from about $2.7 million in 2004, his first year as CEO, to $11.6 million in 2011. None of the 14 health-care companies Monsanto cites compares its CEO pay with the seed company. Monsanto in a statement said it’s at least as big or bigger than its median peer, that the group is diversified across nine industries, and most “are science-based and research-focused like Monsanto.”
Boards sometimes justify pay increases for CEOs simply by adjusting their peer groups. Northern Trust’s board added six companies to the list of businesses it deemed CEO-talent rivals in 2009. There weren’t enough asset managers on the old list, the Chicago bank and wealth manager said in explaining the decision to investors. “Only a small number of the current peer companies have a directly comparable business profile,” the bank said.
Left unstated was the impact government-imposed pay restrictions had on banks that accepted bailout funds during the financial crisis: Median pay among Northern Trust’s old peer group dropped from $9.2 million to $6.8 milion in 2008. None of the companies added to the list—five asset-management companies and a brokerage firm—were subject to government pay restrictions. Among them was BlackRock, the largest money manager in the world, which had just had the second most profitable year in its 21-year history. After the shift, CEO Frederick Waddell received a compensation increase to $11.9 million, from about $8 million the year before. The bank declined to comment about whether peer comparisons played a role in its pay decisions.
Two years later, 31 percent of shareholders voted against Northern Trust’s compensation program at its April 2011 annual meeting. The bank responded by increasing Waddell’s salary to $956,250, from $900,000, cutting his cash bonus to $1.6 million from $2 million, and long-term awards to $7 million from $8 million. Directors also dropped the six newcomers and returned to the old peer list, calling it a “more targeted” group.
That’s not necessarily bad news for Waddell. By 2011, most banks, including Northern Trust, had repaid their bailouts. Median CEO pay on the old list of rival banks for 2010 had risen to $12.9 million. Northern Trust said it believes the current peer list is “appropriate.” It declined to comment on the pay differences between the two lists. Since Waddell, a lifer at Northern Trust, became CEO in 2008, the bank’s shares lost 33 percent after dividends through April 25, compared with a 40 percent drop for the KBW Bank Index.
Four groups of researchers, including Columbia’s DiPrete, who looked at several years’ performance at 1,183 companies for a study published in 2011, found evidence that companies were probably cherry-picking better-paid peers intentionally, with three of them concluding that boards had no legitimate justification for doing so.
One group, led by John Bizjak of Texas Christian University, that looked at 707 companies, found that the amount of cherry-picking has tapered off since the companies had to disclose their choices in 2006. But another group, including Michael Faulkender at the University of Maryland and Jun Yang of Indiana University, finds the opposite in a working paper. It says that cherry-picking has gotten more pronounced since the SEC demanded disclosure. “Observing opportunistic peer benchmarking at other firms,” wrote the researchers, who looked at 763 companies, “those firms that did not engage in such practices may have begun to do so.”