Last fall, with headlines of million-dollar Wall Street bonuses appearing amid the worst economic crisis in a generation, I attended a lecture on corporate ethics by the CEO of one of America's most venerable corporations. This captain of American industry was critical of the compensation committees and formulas that generated these outsize payouts and felt that in his own case, he didn't deserve more than $10 million, regardless of what the committee came up with.
It is truly a statement of the times we live in that a self-imposed $10 million pay cap is a sign of modesty and virtue. Half a century ago, the median pay of top executives in U.S. companies was 30 times an average worker's salary; by 2005, the ratio was nearly 110. How did we get here?
The popular perception is of America's CEOs greedily rubbing their hands as they approve their own paychecks, and there certainly has been some of that. Others argue that in most cases CEOs are richly compensated because they're good at what they do.
Several recent studies stake out a middle ground, assuming that CEOs are neither villains nor business masterminds. These studies argue that the seemingly innocuous practice of benchmarking pay against other companies' CEOs may be to blame, because the list of comparable executives is often formed selectively to include highly paid peers and to omit lower-paid ones.
It makes sense to see what competing firms are spending to reward and retain their leaders. That's where the peer-group comparison comes in. But who is the "right" peer? You probably want to pick someone running a company in the same industry, of similar size, of comparable profitability, and with similar experience.
Compensation committees use their discretion, and that has led to claims of abuse by committees with too-cozy (or just plain incestuous) relations with the CEO. (Compensation consultants, often brought in to help figure out the "right" pay, have also been accused of currying favor with executives in the hope of securing other, more lucrative business from the company.) To shed more light on what had been a less-than-transparent process, the Securities and Exchange Commission mandated in 2006 that companies had to make their peer lists public.
In addition to arming shareholders and corporate watchdogs with better information on how compensation gets decided, the ruling provided financial economists Michael Faulkender and Jun Yang with an opportunity to compare the chosen peers with the peer might-have-beens. The researchers combed through the SEC filings of more than 600 companies, recording the set of peer CEOs that informed executive compensation decisions. For each CEO, they also assembled their own group of peers based on their own calculations of which other executives were most similar.
A thoroughly compromised compensation committee would pick peer firms that are larger, more profitable, and otherwise more likely to have high-earning CEOs than their own. Yet Faulkender and Yang found that by and large, there was enormous overlap between their list of peers and the ones actually chosen, so that compensation committees were for the most part being "kept honest" in choosing pay benchmarks.
Discretion did have its benefits, at least for CEOs. The peer CEOs selected by compensation committees had total earnings that averaged nearly $850,000 more than those of ignored potential peers. One executive's raise feeds into the pay hikes of others, as yet other companies will use the now 5 percent higher paycheck as a peer comparison next year.
Add to the mix the fact that every year a few corporate leaders are awarded outsize compensations that allow them to leapfrog their competitors' pays (perhaps because of an exceptionally weak board, threats of retirement, or other unusual circumstances), becoming the new standard for pay comparisons in the process, and it's easy to see how a few decades of favorable peer lists could snowball into the enormous incomes we're seeing today. In fact, according to the calculations of sociologists Tom DiPrete, Greg Eirich, and Matthew Pittinsky, it is possible to account for much of the recent rise in executive pay based on the positive feedback loop and interrelatedness among CEOs' pay.
Why do compensation committees err on the side of generosity in spending shareholders' dollars on CEO pay? We all want to be a little better than average, and for board members who like to think they have an above-average CEO, this may translate into the choice of a relatively favorable pay package and a set of peers to match.
This explanation for runaway salaries in the corner office isn't going to sell a lot of papers — there aren't any insidious backroom conspiracies to spin into a story of intrigue. Yet for the very same reasons, it comes across as a more likely account for the rise in CEO pay — compensation committee members are normal people, not conscienceless scoundrels, who are for the most part doing their best to attract and retain leaders. Moreover, the lesson isn't that we should dump the baby of peer comparison out with the bathwater. If CEOs and others should earn "what the market will bear," how better to figure this out than to look at how the market is treating other CEOs? But this CEO labor market will work only if all companies also keep an eye on the more basic market principle that higher CEO pay must first and foremost be tied to the success of the companies they lead.
Ray Fisman is the author of Economic Gangsters: Corruption, Violence, and the Poverty of Nations (with Edward Miguel). He is the Lambert Family professor of social enterprise and research director of the Social Enterprise Program at the Columbia Business School.