By James Kwak
Meg Whitman is what is known as a superstar CEO. She became CEO of eBay in 1998 and took it public; during her reign, eBay became one of the most successful, most valuable Internet companies in existence (and Whitman became a billionaire). She used her celebrity to mount a high-profile, expensive, and ultimately unsuccessful campaign to become governor of California (losing to career politician Jerry Brown) before being named CEO of HP, the iconic Silicon Valley company.
Why did HP, one of the largest information technology companies in existence, hire Whitman, who preceded her stint at eBay (auction house for random stuff from people’s attics) with jobs at Disney, a shoe company, a flower delivery service, and a toy company? Because of the idea of the superstar CEO, with transferable general management skills, who can transform any organization.
Charles Elson and Craig Ferrere have written a new paper about that idea (Harvard Law School Forum blog post; paper at SSRN). The concept of transferable executive skills is the most common justification for outrageous CEO compensation packages: if there really is a single global market for CEO talent, then companies have to pay the going rate for that talent, which is whatever the market will bear. The problem is that the whole theory rests on a myth.
On pages 28–30, Elson and Ferrere cite multiple empirical studies finding no relationship between a CEO’s performance at one company and his performance at the next company that massively overpaid to hire him. They conclude: “the empirical evidence suggests a negative expected benefit from going outside rather than pursuing an internal succession strategy, despite the ability to access an enhanced talent pool. In the aggregate, CEOs appear to be at their most effective only when they have made significant investments in firm-specific human capital.” Nor does the world of CEO hiring actually behave like a market in which stars move progressively from smaller to bigger firms where their marginal product will be higher (since their skills are applied to more people and assets)
Elson and Ferrere’s main argument is that excessive CEO compensation is caused by the ubiquitous practice of benchmarking new CEO packages against those given by “peer group” companies. This practice would only make sense if there really is a liquid market for CEO talent and your CEO could jump ship for a peer at any time. Otherwise, what you really have is a monopoly-monopsony situation where what matters is the CEO’s value to his current firm, not the value of other firms’ CEOs. Peer group benchmarking (and then setting your CEO’s compensation at the 50th, 75th, or 90th percentile of the peer group) is just a mechanism for perpetually ratcheting CEO compensation upward, without any relationship to the value provided to actual firms.
Elson and Ferrere distinguish their position from what they describe as the two main schools of thought on CEO compensation: (a) those who see it as a product of CEO capture of weak boards and (b) those who see it as the natural result of a competitive market for “talent.” I would say their analysis is much more consistent with the former. In a world of weak, captured boards, peer group benchmarking is a convenient fig leaf for outrageous pay packages. They are right, however, that simply having an independent compensation committee will not be enough if that committee is hiring the same old compensation consultant churning out the same old market-based justifications for lavish packages.
So why pick on Meg Whitman? Well, she’s in the news these days, getting the superstar treatment. (“In all likelihood, this is Ms. Whitman’s last great public performance.”) But you can’t blame her for capitalizing on her fame. The people to blame are the board of HP, which somehow thought that a one-hit wonder was the answer to their problems. This after hiring Léo Apotheker, a man who was hugely successful at SAP, where he worked for two decades, yet bombed at HP. And before Apotheker there was Mark Hurd, who spent more than two decades at NCR before doing a middling job at HP (once praised for cutting costs, now criticized for letting the technology world—Google, Apple, Amazon, Facebook—pass HP by).
The lesson is that successful CEOs are often successful because of the people around them, and to the extent that their individual contributions matter, they are often specific to their companies. Meg Whitman may have been a great CEO of eBay (although, remember, eBay also watched the technology world pass it by, with the arguable exception of PayPal). But that doesn’t make her a great CEO of anything else.