The CEO pay ratio rule is a failed experiment

Back in 2017, a new rule aimed at exposing the excesses of Corporate America went into effect: Every year, most publicly-traded companies would have to disclose a CEO pay ratio, comparing the compensation of the chief executive to that of the median employee.

Having spent the subsequent five years trying to make sense of these disclosures, I think it’s time to call the rule a failed experiment, and scrap it.

I don’t say this lightly. I had high hopes for the requirement: Last year, I suggested that it might yet embarrass companies into addressing pay inequality. But as often happens, the reality hasn’t matched the laudable intentions. Many companies, afraid of reporting too high a ratio, have manipulated it into meaninglessness. It’s not really comparable across different types of businesses. Executive pay keeps growing, seemingly unchecked: A survey by Equilar found that it increased by 30 percent in 2021 at the 100 largest US companies.

Consider the case of delivery platform DoorDash Inc. Earlier this month, the company reported that CEO Tony Xu’s 2021 compensation — a modest $300,000 — was about three times the roughly $90,000 that the median employee received, among the lowest ratios in the US. But the latter figure didn’t include about 6 million workers who make the actual deliveries in the five countries where DoorDash operates — because they’re independent contractors instead of employees, according to a company spokeswoman. That $300,000 isn’t necessarily representative of the CEO’s pay: In 2020, when the company was private for most of the year (and hence didn’t have to report a pay ratio), Xu received more than $400 million in various forms of compensation.

Beyond that, companies have legitimate complaints. For those that operate in multiple countries, calculating the median employee can be fraught. Those that employ a lot of people in relatively low-wage emerging economies, for example, will naturally look worse. Warren Buffett’s Berkshire Hathaway (most recent pay ratio, 6 to 1) has pointed out such problems in its last five proxy statements.

It’s hard to imagine a simple fix. The Securities and Exchange Commission (SEC) could require companies to include all workers, including contractors, or demand more historical data on CEO compensation. But none of this would get to the root of the problem, which is that measuring income inequality at the company level is inherently flawed. A finance company that pays everyone well but preys on its customers can contribute more to economy-wide inequality than a multinational whose CEO earns far more than its tens of thousands of employees. The opportunities to manipulate the ratio are endless.

To be sure, compensation information for a company’s top executives absolutely should remain in the summary compensation table. It could even be expanded to provide additional details on particularly embarrassing numbers, like the $25 million Meta Platforms Inc. spent on personal security for CEO Mark Zuckerberg in 2021, instead of requiring investors to dig through multiple footnotes for limited information. But does the average investor really need to know that Zuckerberg made 92 times what the company’s median employee did?

At a time when critics are circling the SEC, and when some are even challenging its power to make rules at all, it behoves the agency’s leaders to be painstakingly fair and reasonable. The problems with the pay ratio offers a real opportunity: By admitting that the rule hasn’t worked as planned, the SEC could gain a counterpoint to detractors who complain about regulatory overreach, perhaps allowing it to focus on issues that matter more to investors, such as cryptocurrencies and special purpose acquisition companies. Sometimes one must concede the battle to win the war.

Michelle Leder is a Bloomberg Opinion columnist, creator of Footnoted.com, a site devoted to SEC filings, and author of ‘Financial Fingerprint: Uncovering a Company’s True Value’. Views are personal, and do not represent the stand of this publication.