How the new CEO-to-worker pay ratios can mislead

Public companies are required for the first time this year to compare the compensation of their chief executive officer to that of the median employee in an annual Securities and Exchange Commission filing. As expected, the pay ratios so far show that CEOs in many cases are making hundreds of times more than the Average Joe and Jane.

But those numbers may be misleading -- on both sides of the ratio equation, analysts said. The SEC allows a lot of flexibility concerning the wage calculations, and the differences can provide deceptively high or low wage information for both median worker pay and the CEO compensation used to create the ratios, they said.

"You have a lot of variation in how these numbers are calculated," said Seth Duppstadt, of Proxy Insight, which provides corporate governance information for shareholders. "Methodologies are not consistently applied across all companies."

For example, regarding worker pay the SEC allows for estimates or statistical sampling instead of actual pay in some cases. Employers also have flexibility about the counting of overseas and part-time workers, he said.

"So an employee may see the average pay listed for their company and think they are being under- or overpaid, or one company in an industry may appear to pay more than another," Duppstadt said. "Instead, the differences may be based on how the calculation was done."

Data on median worker pay are also affected by a company's labor model, which is determined by the type of jobs, the distribution of full-time and part-time employees, and geographic location, said Jamie Teo, an associate director for Willis Towers Watson, a compensation consultancy. Even two companies within the same industry can have very different labor models, she said.

"We encourage anyone looking at these numbers to read them with an open and curious mind and to investigate the fundamentals that went into the calculations," she said. "That's important rather than assuming the numbers are an indication of whether a company may be overpaying or underpaying their employees."

The executive compensation side of the equation may also be skewed, said Stephen Quinlivan, a corporate finance attorney and partner with the law firm of Stinson Leonard Street. The executive calculations often use compensation, such as unrealized options, that executives haven't received and may never receive, he said.

All these inconsistencies make it difficult to use the new data for comparison purposes. At the same time, workers within an organization shouldn't compare their wages to the median because of the variety of factors involved, he said.

"There was a lot of concern about how you actually calculate these numbers with companies that have complex workforces throughout the world and even the U.S.," Quinlivan said. "In some senses, these numbers are like comparing apples to oranges."

The differences in calculation methods may result in ambiguity in wage information across the economy, Duppstadt said. For example, the U.S. Census Bureau determined that the median household income in 2016 was $59,039. And yet an analysis of 2017 SEC filings from companies included in the Russell 3000 Index shows the median worker makes nearly $79,000, Duppstadt said. That's about 25 percent higher than Census date just a year earlier.

"Either workers in the Russell 3000 Index make a lot more than the average household, or something else is wrong," he said.

The pay ratio disclosures were mandated as part of the Dodd-Frank Act, which went into effect after the financial crisis in 2010 to reform Wall Street and better protect consumers. This provision's goal was to provide a concise way to identify the pay differences between executives and their workers.

"Everyone was somewhat mystified by why this burden was thrust upon public companies because of the ambiguity of the results and the burden they can be to calculate," Quinlivan said. "But nevertheless it's here."

If the numbers don't provide clear information about the pay of workers or executives, and provide deceptive information when compared to other companies, their value comes into question, said Brian Blackwood, a compensation consultant also with Willis Towers Watson.

"I don't know the value of the numbers at this point," he said. "The jury is still out. We'll just have to wait and see."