America's boards of directors are transfixed on a myth, says Michael Dorff, a professor at Southwestern Law School in Los Angeles.

They think they are paying their chief executives for performance. They're not.

"If oil goes to $150 a barrel, is the CEO of Chevron a genius?" he asks. "When oil then falls from $150 to $100, does that make the CEO of Chevron an idiot?"

How about when a raging bull market sends stocks of the nation's biggest companies soaring--or when an economic crash sends them spiraling--should CEOs be rewarded or punished for that?

How about when a CEO borrows billions--because the Federal Reserve is flooding the world with cheap dollars--and uses those funds for stock repurchase plans, or to pay higher dividends? Is that why CEOs are supposedly worth the millions that their shareholders end up paying them?

Mr. Dorff, who has been studying CEO pay for more than 10 years, has come to a simple conclusion: This thing we've come to think of as "performance pay" should be abolished. He calls it the "lynchpin" to all that is wrong with corporate governance and a society that is increasingly divided between the super rich and everyone else.

At its worse, performance pay leads to scandals such as backdated stock options and manipulated financial statements. At its best, it can still distract corporate leaders into thinking more about their own personal rewards than the long-term sustainability of the companies they are supposed to manage.

Mr. Dorff is working on a book, "Indispensable and Other Myths: The True Story of CEO Pay," slated to be published next spring by the University of California Press. In it, he argues that contrary to popular belief, CEOs don't really matter.

"There are always outliers," he concedes. "There are very salient examples, like Steve Jobs or Warren Buffett. You'd say those guys certainly matter, but on average CEOs don't matter."

If 250 qualified candidates applied for a CEO job, the differences between the first and last candidate could be quantified in fractions of a percent, Mr. Dorff said.

Boards of directors--mostly stacked with other CEOs--don't see it that way. If they have five solid candidates, but the top candidate wants considerably more money than the fifth candidate, they're apparently hardwired to think it's because he's worth more.

"It doesn't make much difference if you go with your first choice or your fifth choice," Mr. Dorff said. "I'm saying that if your fifth choice will work for less, then go with your fifth choice."

A CEO isn't a god, and shouldn't be paid as such. A CEO doesn't control the economy, the market, the industry, the customers, or even, to a large extent, the employee base. CEOs get paid more, mostly because they are typically hired by other CEOs with a grossly distorted sense of the value of their contributions.

"Studies show CEOs have very little influence over corporate performance, and even stock performance," Mr. Dorff said. "A lot of other things have a much bigger influence than whatever the CEO is doing."

CEO pay at the 300 biggest U.S. corporations rose by a median 3.6% to $10.1 million in 2012, according to a study by The Wall Street Journal and the Hay Group, a management-consulting firm, published last week.

Most of that increase was due to rising stock values, the report said. Corporate boards of directors have been holding steady on CEO pay--pegging it with salary of $1.1 million, bonus of about $2 million, and equity grants valued at around $7 million, the report said. That ol" rising tide is floating their boats again. And if the tide pulls out again it will be another perfect time to pick up some newly priced stock options from the ocean bottom.

"Some corporations are paying 10% to 15% of corporate profits to their top five executives, and what are we getting for that?" Mr. Dorff asks.

From the 1940s through the 1970s, CEO pay at large companies was relatively stable at about $1 million a year in 2000 dollars, he said. Now, it can be a hundred times that amount.

This was an idea that emerged in the late 1970s--just load the CEO's plate with stock options--and it has resulted in higher CEO pay in every decade since.

"Boards did it because economists told them to," Mr. Dorff said. "But it's a very crude measure of corporate performance. If the market as a whole is going up 10% a year, then those options can be really valuable, even if your company isn't doing so well relative to other companies."

In pursuing his research, Mr. Dorff initially thought performance pay sounded perfectly reasonable and even intuitive. He thought it just needed a more nuanced approach that would filter out externalities such as market performance, industry performance, or whether the whole stinking performance was due to financial restructurings rather than organic improvements.

Now he has concluded it should just be ditched entirely. Performance pay, he says, works well for manual laborers, but not for people who are supposed to be geniuses for a living.

"If you pay somebody by the brick, they're going to lay more bricks," he said. "But for creative tasks, analytical tasks, laboratory studies show it does not work well at all."

When people see millions of dollars hanging in the balance--dependent to a large extent upon forces they don't really control--they get nervous, and may actually perform worse, Mr. Dorff said. It would be better to pay them, straight up, rather than hand them a rigged lottery ticket every year, he said.

"These are high-powered people," he said. "They deserve a good salary. But they don't deserve $10 million [a year] and they certainly don't deserve $100 million or $300 million."

(Al's Emporium, written by Dow Jones Newswires columnist Al Lewis, offers commentary and analysis on a wide range of business subjects through an unconventional perspective. The column is published each Tuesday and Thursday at 9 a.m. ET. Contact Al at al.lewis@dowjones.com or tellittoal.com)