Most C.E.O.’s used to be able to handle their pay negotiations in private, but the Dodd-Frank reforms, which were passed in 2010, now give shareholders the right to vote on executive compensation. This has helped usher in a so-called “say on pay” revolution, which tries to stop executives from making more money when their companies don’t do that well. In Switzerland, a recent nationwide referendum, passed 2 to 1, gave shareholders the right to restrict the pay for the heads of Swiss companies. The European Union is likely to vote on a similar measure by the end of the year.

C.E.O.’s like Farris have long argued that they should make more money, but what’s surprising is that many business-school professors make the case even more energetically. The standard defense is that a talented marketing director or chief engineer can help a company thrive, but the next-best candidate will probably be successful, too. A great C.E.O., they say, is many times better than an average one, and those great ones need high-powered incentives. C.E.O.-friendly economists also suggest that the salary is moot if the chief executive is creating many multiples of their pay in shareholder value. Some of these enthusiasts point out that Steve Jobs rescued Apple (after his exodus) in 1997, when it was near bankruptcy, and turned it into one of the most valuable companies ever. Jobs was worth an estimated $7 billion at his death, but he made hundreds of billions of dollars for his shareholders. Many now say he was underpaid.

C.E.O.’s might indeed need significant incentives, but the problem is that most of them don’t perform like Steve Jobs even when they get them. The financial research firm Obermatt recently compared the compensation of C.E.O.’s at publicly traded firms and their performance and found no correlation between the two. Like a bottle of wine or a promising college quarterback turning pro, C.E.O.’s are similar to what economists call experience goods: you commit to a price long before you know if they’re worth it. Just ask the shareholders of J.C. Penney, which ousted its C.E.O., Ron Johnson, less than 18 months after hiring him away from Apple. Under his leadership, the company lost more than $500 million in a single quarter.

So far, the “say on pay” revolution feels more like the second season of “Game of Thrones” — there’s a lot of drama, a bit of blood, some cheers, but things end up more or less exactly where they started. While the Dodd-Frank law requires a shareholder vote on executive pay at least every three years, the vote is not binding. Apache’s board eventually lowered Farris’s package by around $3 million, but it is the exception. Shareholders ended up approving pay packages around 97 percent of the time. A vast majority of overpaid C.E.O.’s, it seems, have little to fear from all these new guidelines.

Economically speaking, this is more than a little odd. Shareholders should be motivated to pay their C.E.O.’s according to their success. But doing so involves a tricky dance known to game theorists as the principal-agent problem: how does an employer (the principal) motivate a worker (the agent) to pursue the principal’s interest? This principal-agent problem is everywhere. (Do you pay a contractor per day of work or per project? Do you pay salespeople by the hour or on commission?) It becomes particularly thorny when the agent knows a lot more about his job than the principal. George Costanza was a comic incarnation of the principal-agent problem. He constantly invented schemes to make his employer think he was doing his job well when he wasn’t doing much at all. “When you look annoyed all the time,” he once told Jerry and Elaine, “people think that you’re busy.”

Article source: http://www.nytimes.com/2013/06/02/magazine/ceos-dont-need-to-earn-less-they-need-to-sweat-more.html?partner=rss&emc=rss