Executive Compensation

Cracking the CEO Pay Puzzle
By Laura Vanderkam
From the March/April 2008 Issue

Filed under: Big Ideas

Why has executive compensation exploded in recent years? New York University’s Xavier Gabaix gives one answer.
Americans may be used to the rich getting richer, but the issue of CEO compensation never fails to set them fuming, especially in the post-Enron era. Since 1980, executive pay at big corporations has increased more than sixfold. According to the Corporate Library, an independent research firm, the average CEO of an S&P 500 company made $15.06 million in 2006. Late last year, the House Committee on Oversight and Government Reform held hearings on the issue. “Reports of astronomical payouts to corporate CEOs have led many to question the fairness and effectiveness of the system for setting executive pay,” said Chairman Henry Waxman (D-CA).

The furor over CEO compensation intrigued Xavier Gabaix, 36, a French-born and Harvard-trained economist who now teaches at New York University’s Stern School of Business. He and fellow Stern economist Augustin Landier decided to investigate whether corporate bosses were, in fact, “overpaid.” They tested a variety of possible explanations for the recent spike in executive rewards, and came to a conclusion that may surprise the populist scourges on Capitol Hill.

As Gabaix and Landier write in a new Quarterly Journal of Economics article, the sixfold increase in American CEO pay from 1980 to 2003 is almost wholly explained by the roughly sixfold increase in market capitalization of big U.S. companies over the same period. (Asset values have increased sixfold because both corporate earnings and the price-to-earnings ratio investors are willing to tolerate have increased by factors of 2.5.) The trend lines of market capitalization and executive payouts rose and dipped in near-perfect tandem.

According to Gabaix and Landier’s model, the talent differences among CEOs are generally minor. For example, if a given firm substituted the most talented CEO for the 250th most talented CEO, its market capitalization would only increase by 0.016 percent. But for a $500 billion company like ExxonMobil, 0.016 percent is equivalent to some $80 million. In other words, as companies get bigger, a talented CEO can have a greater impact. Therefore, large companies bid up prices across the board for the small number of men and women deemed capable of managing them. The reason CEO pay in other countries (such as Germany) tends to be lower is that the “big” companies abroad are generally smaller than the big companies in America. We do not yet have a global market for CEO talent.

Given the controversy surrounding this issue, Gabaix’s explanation for ballooning CEO pay has come under fire. He freely admits it is not a universal rule. For example, the model doesn’t work well in Japan. Gabaix reckons this is because the Japanese CEO market is relatively static, at least when compared to the more fluid and competitive U.S. market. “In Japan, you groom your CEO in-house,” he says. This means that firms need not bid up CEO salaries by competing for outside talent.

Averaged over the largest 500 companies, CEOs have not managed to systematically boost their pay during down markets. CEO pay has fallen as the market has tanked.Nor does the Gabaix model work well for the United States before 1970. Union power—which has declined steadily over the past four decades—may help explain that. Or it could be that new managerial talent glutted the U.S. market in the immediate postwar era. Or perhaps “the U.S. in the 1950s may have been like Japan is now,” Gabaix speculates, with Organization Men refusing to dart from one company to another to advance their careers. Or it could be a combination of these factors. “We don’t completely know.”

There’s also the question of folks like former Home Depot CEO Robert Nardelli, who received a whopping $210 million severance package even though Home Depot’s stock price declined under his tenure. Gabaix notes that there are always outliers. And as Enron showed, there will always be genuine corporate swindlers. Averaged over the largest 500 companies though, chief executives have not managed to systematically boost their pay during down markets. Indeed, CEO compensation fell from 2000 to 2003, as the market tanked.

Of course, it’s not just the sixfold increase in CEO pay that bothers people: it’s also that executive compensation has risen relative to that of the median worker. Critics want to know why CEOs seem to be the only ones who are benefiting from asset-price gains.

Gabaix has two answers. First, “the median worker, sadly for him or her, just manages a very small part of the economy,” he says. “The economy as a whole has not grown by that much—just 2.5 percent a year or so.” In most jobs, the stakes are very small. The head of procurement in a medium-sized business may be supremely talented—but since even an abundance of talent in that job will save the business only a small amount of money on supplies, superior talent isn’t terribly valuable.

‘People don’t really begrudge the compensation to sports stars,’ Gabaix observes. But with individual CEOs, ‘it’s harder to measure talent.’Second, it’s not just CEOs who are seeing outsized returns these days. In virtually every field where a small difference in talent translates into a big difference in profits, superstar performers have seen their compensation skyrocket. Tom Cruise may only be marginally better at his craft than a thousand actors you have never heard of—but putting Cruise in a movie will make it a blockbuster, so he’s paid accordingly. Likewise, New York Yankees third baseman Alex Rodriguez has used his impressive statistics to negotiate a mind-boggling salary.

“People don’t really begrudge the compensation to sports stars,” Gabaix observes. But with CEOs, “it’s harder to measure talent.” Their individual performance cannot be quantified with pure statistics. “Perhaps there’s more room for suspicion that they’ve put their friends on the board—that sort of thing,” adds Gabaix. All of this translates into a fierce resentment of executive pay—even though, as the Gabaix-Landier model shows, America’s CEO market is working pretty well.

Gabaix is not a free-market absolutist. “I’m completely open-minded,” he says. “I have no ideological commitment at all.” He has also done research on “shrouding”: when businesses charge annoying add-ons and fees after luring you in with a low price. While the CEO market works well, this one does not. Rather than stimulating productive competition, shrouded markets lead to a game of cat and mouse. For example, consumers bring their cell phones to hotels in order to avoid telephone charges—only to find that they’re paying a “resort fee” for use of the pool. And yet few businesses seek to compete by advertising that they have no fees or add-ons—a situation Gabaix finds fascinating, (if frustrating). “There’s always this extra x percent of the economic transaction that could be improved upon,” Gabaix says. “I’m not advocating that you should regulate that, because there’s a lot of innovation and often the regulation is too far behind.” But forcing full disclosure is “probably a good idea.”

Whatever project Gabaix tackles, there’s always a heavy dose of math involved. Indeed, he warns that he’s just relying on “weak sociology” and “liquid, fuzzy notions” whenever he’s asked to explain why his mathematical models work the way they do in real-world situations. This may explain why one of his ongoing research projects—bits of which have been published in Nature magazine—includes teaming up with physicists to crunch numbers on the origins of large stock market fluctuations. At the bottom of it all, Gabaix says, he’s just “trying to better understand the simple economic structures that govern the apparently messy world.”