Is Your Chief Executive Really Necessary?
from
September/October 2002
by Judy Ward
Like orchestra conductors, CEOs tend to be expensive and despotic.
So might it be possible, then, just as the Grammy-winning Orpheus
Chamber Orchestra has built an international reputation by working
without a conductor, that certain companies could learn to get along,
even flourish, without someone to wave a baton?
A controversial Harvard Business School study—"When Does Leadership Matter? The Contingent Opportunities View of CEO Leadership"—raises that intriguing possibility. "In some situations, the CEO matters a lot. In others, the CEO is not that important," says Nitin Nohria, a professor of organizational behavior and one of the study's three co-authors. "It is fair to say that certainly one sometimes wonders if the obsession with great CEOs is worth it or not. In other cases, finding the right CEO can make or break the firm."
As far as Nohria knows, no company has ever put the no-CEO theory to the test. But the notion is tantalizing and has evidently roused the curiosity of at least some major corporations. Drugmaker Novartis, for example, called Orpheus in to demonstrate how a conductorless orchestra works. The secret seems to be committees. Different sections of the orchestra (strings, winds, and so on) elect leaders for each piece Orpheus will play. The leaders work together to determine how to perform the composition, and then run through it as a mini-orchestra. When they're happy with how it sounds, the committee members brief their individual sections, and then the orchestra rehearses the piece as a whole. Each member can go sit out front to listen to the performance and make suggestions on how to improve it.
The Harvard Business School study looked at four factors that affect how well a company does: macroeconomic conditions, the company's particular industry, its competitive position, and its chief executive. Using regression analysis, the researchers then calculated how each factor influenced the bottom lines of 531 companies in 42 industries.
Industry by industry, CEOs weren't that effective. A company's competitive position was shown to be most important in 28 industries, macroeconomic factors in the other 14. In more than half the industries, the CEOs had a "near-zero" effect on how their companies performed.
The study also concluded that CEOs can be most influential at companies in boring, slow-growth industries that are highly concentrated and cash-rich, and where there is little opportunity for acquisitions. In other words, an effective CEO will move fast when a rare deal shows itself. The manufacture of measuring and controlling devices topped this list. Conversely, according to the study, CEOs wield little influence over company performance in high-growth, high-debt, highly competitive industries such as telecommunications and electric power. These are "the sexy industries," says Bharat Anand, another co-author and an associate professor at the B-school. "Everyone is going along for the ride," he adds. "Unless somebody really screws up, everybody will do well." Of course, since the report was written, telecoms aren't as sexy as they used to be—mainly because people did screw up.
The real surprise of the study may be the industries it places in the middle, between boring and sexy. These include computer programming, data processing, and software, an industry dominated by the hardly boring likes of Microsoft's successive CEOs, Bill Gates and Steve Ballmer, and Oracle's Larry Ellison. "These are larger-than-life leaders," Nohria admits. But apparently the exception.
Should boards consider eliminating the CEO position in order to set up an Orpheus-like series of committees—for example, one to make a final decision on an impending merger and another on a stock buyback? Probably not. But the study might give ammunition to the compensation committee at Qwest, say, or other telecoms. "If you're in a low-CEO-impact industry, there is no point in creating a large variable-compensation package," Nohria says.
In what the authors call high-impact industries, compensation committees can justifiably be more generous to CEOs who do well by their companies—and more tightfisted with those who fail to deliver. At the same time, "those boards need to be more vigilant and to keep their CEO on a shorter leash," says Nohria, "because the damage he can do to the company is huge. In a low-impact industry, they can be more tolerant and give him a longer rope."
Companies can also possibly economize on CEO pay by adjusting the scope of their executive searches to correspond to the likely importance of the top job. "If you are in a high-CEO-impact industry, it makes sense to invest a lot more in finding the right person, and to be exhaustive," Nohria says. "If you are in a low-CEO-impact industry, you may want to have a constrained search. On the low end of the spectrum, the CEOs you choose matter, but they don't matter that much."
The study, not surprisingly, has critics. Among them: shareholder-activist Nell Minow, who edits the Corporate Library (www.thecorporatelibrary.com), a website on corporate governance. "I don't think that a board should ever say, 'We're in an industry where the CEO doesn't make much difference, so it doesn't matter who the CEO is,'" she says.
Others, like Jeffrey Sonnenfeld, an adjunct professor at the Yale School of Management and the founder and president of the school's Chief Executive Leadership Institute, fault the study's methodology. Sonnenfeld says its conclusions are skewed by the researchers' assumption that companies with ample cash on hand provide their CEOs with more opportunity to exercise leadership. As a result, the CEOs of companies that make acquisitions are credited with being more influential. But evaluating leadership on this basis, though once widely embraced, has been "overwhelmingly discredited," says Sonnenfeld. He adds that the study also fails to give credit to CEOs in cash-strapped industries who have to use different strategies to prosper, such as forming partnerships. Sonnenfeld ticks off the names of chief executives who have made their mark in what the study deems a very low-impact industry, hotels and motels. Marriott International's Bill Marriott, he says, "completely reshaped the industry. . . . Where do you get off saying that CEOs don't have much influence in the hotel industry? That's shocking."
Even the authors concede that their study has holes. Nohria says the biggest may be that it doesn't explain what type of CEO is best for each industry—something that is of prime interest to board members trying to fill the top job.
Still, the study could do some good by stiffening director resolve come
CEO contract-renewal time. And it might not hurt to pipe some music by
Orpheus—perhaps their big hit Shadow Dances: Stravinsky Miniatures—into
the boardroom.
