It has been nearly four years since the Securities and Exchange Commission finalized its rule mandating that companies annually publish the ratio of chief executive officer compensation to the salary of the company’s median employee. We’ve previously argued that this provision is one of the costliest regulations required by the 2010 Dodd–Frank Act, writing in these pages that “mandating the regular publication of a crude gauge of relative CEO compensation is a costly exercise that fixes precisely nothing” (“The Meaninglessness of the SEC Pay Disclosure Rule,” Spring 2014). Given the information that has been reported to the SEC after this rule’s implementation, we stand by that assessment.

Our chief complaint with the rule when it was first promulgated was that the SEC’s cost estimate for compliance woefully understated reality. To calculate the total compensation of every employee in a company—domestically and internationally—likely would require the use of significant resources and cost firms millions of dollars. However, in its initial proposed rule, the SEC pegged annual costs at just $72 million for all affected firms, with an associated 545,000 paperwork-burden hours. This works out to roughly $18,000 per company and 142 paperwork-burden hours.

Industry objected that these figures significantly underestimated the true compliance costs. For instance, respondents to a U.S. Chamber of Commerce survey of 118 firms—the only such survey on the rule’s cost that we found—estimated it would take 952 paperwork hours per firm at a cost of roughly $185,000. That translated to an industry-wide burden of more than 3 million hours and aggregate costs exceeding $700 million, potentially vaulting it into one of the most expensive Dodd–Frank rules on record.

To its credit, the SEC implicitly acknowledged this error when it released its regulatory impact analysis for the final rule. It increased its annual aggregate compliance cost estimate to $526 million, a seven-fold increase, and paperwork compliance rose to more than 2.3 million hours, a four-fold increase. The total net present value cost jumped eight-fold in the final rule.

The benefits to society in return for these costs are unclear. We see no reason to think that this exercise has shed any useful light on U.S. income inequality or that the rule is ameliorating income inequality’s causes. In addition, there is no evidence the requirement will prompt Congress to take any sort of action on executive pay; such action, if it happens, will be the product of ideology, not SEC reporting. In retrospect, even if compliance costs were minimal, that still would not justify the rule.

The SEC’s final cost estimate incorporated numerous concessions to ease compliance costs. For instance, the final rule allowed multinationals operating in foreign jurisdictions with onerous privacy laws to exclude foreign employees if obtaining total compensation figures would violate those privacy laws. The final rule also limited the burden of collection to just consolidated subsidiaries instead of forcing companies to calculate the median pay of all employees. Finally, the SEC allowed companies to identify the median employee every three years instead of annually, provided there is no reason to believe there would be a significant change in the company’s pay ratio.

Despite these cost-ameliorating measures, the pay ratio disclosure rule still ranks as the eighth most expensive Dodd–Frank rule promulgated, exceeding rules on home mortgage disclosure, standards for swap-dealers, and regulatory capital requirements. Knowing what we know now about the causes of the Great Recession, the pay ratio rule seems especially unjustified. Three years after implementation, the rule continues to impose unnecessary compliance costs without generating any measure of value for investors, the market, and perhaps even the politicians who insisted on its inclusion in Dodd–Frank.

A study by executive compensation consultant Deb Lifshey weighed in on the cost-ineffectiveness of the rule. She observed that it would prove to have disproportionate effects on large multinational firms, something we noted in our 2014 article. More generally, she suggested that the larger the firm, the larger the ratio, with consumer-facing firms especially affected.

A redundant rule / When the SEC began contemplating the pay ratio rule, plenty of data relevant to the rule were already available. For instance, existing law already required the publication of CEO pay for public companies, which meant that the pay ratio rule merely required that companies do the arithmetic necessary to calculate median employee compensation (as opposed to just wages or salary), which turns out to be more complicated than meets the eye for numerous reasons. For starters, assigning a value to fringe benefits of each employee is a complicated and often subjective enterprise. What’s more, aggregating compensation for thousands of employees across numerous countries—a step that is necessary for multinational corporations—also requires numerous decisions to account for exchange-rate fluctuations and purchasing power differentials. Many of these companies may also have to integrate various payroll systems that do not otherwise connect, which can be a costly undertaking. Finally, the treatment of part-time and partial-year employees can easily bias the estimate: retail companies may have the majority of their workforce working part-time, which means the CEO comparison is made to someone working less than 40 hours a week.

It is also not clear that the rule provides any new information. Numerous scholars had already estimated an overall CEO/​median worker ratio before the formal regulation appeared in the Federal Register. A 2014 piece published by the Economic Policy Institute estimated a pay ratio of roughly 200:1, which was consistent with several contemporaneous studies. A few years earlier, the Society for Human Resource Management pegged the ratio at 344:1 and a 2009 study by the Center for American Progress estimated 240:1 in 2005.

There was plenty of research before the rule on CEO pay itself. For instance, one study found that companies with the highest-paid CEOs tend to have below-normal returns—good fodder for corporate board discussion. These days, it takes just a few seconds of sleuthing to uncover a ballpark estimate for the average pay of a particular company that is immune to the SEC’s dictate. Those figures—such as the ones reported on the websites Glassdoor and Payscale—are voluntarily provided from current or prospective employees. For example, Payscale’s data suggest that the typical employee at Honeywell earns $81,000, while at General Electric the typical employee earns $86,400. Both websites also disaggregate salary data within a company by profession, providing more relevant data than the SEC.

When the first set of pay ratios was reported, there were some attention-grabbing revelations. But the numbers were not as extreme as researchers anticipated. For instance, Lifshey found the median pay ratio was 70:1 for Russell 3,000 companies and 166:1 for Equilar 500 companies. Both figures are less than the estimates done prior to the regulation (and the 2008 financial crisis), but that does not mean that the rule resulted in some sort of decline in compensation. More likely, the stock market rise pre-2008 increased CEO compensation and a few firms accelerated compensation into the year prior to the beginning of pay ratio reporting or else arranged for some sort of contingent compensation

As always, the individual firm data provide more perspective than any national average, and extreme outliers can be illuminative. For instance, 10 companies reported a ratio of zero—indicating the CEO did not take a compensation package (a group that included Twitter and RE/MAX). The highest reported ratio was at Weight Watchers, where the CEO earned $35 million compared to median employee compensation of just over $6,000. A ratio of 6,000:1 may very well provoke a modicum of outrage, but it provides investors—who are supposed to benefit from such information—no useful context to understand the ratio. It does implicitly reveal the company has plenty of part-time staffers, which any educated investor would presumably already know. But it provides no relevant insight into the appropriateness of a given CEO’s compensation.

Data for whom? / Several other provisions of Dodd–Frank were inserted with the intent to do little other than shame some firms. For instance, the law’s “Conflict Minerals” rule required businesses to disclose whether its minerals originated in the Democratic Republic of Congo or an adjoining country. That turned out to be quite difficult—and costly—for some to ascertain.

But the pay ratio rule may be even more shabby than the others because it may not provide anything close to an accurate estimate of what Congress intended to be revealed. Deferred pay, accumulated bonuses, or one-time company-wide bonuses of the sort that were provided by numerous corporations in the wake of the 2017 tax reform will distort pay-ratio estimates. Hiring in a growing economy will have a similar distortionary effect. For instance, a company that adds 5,000 new employees mid-year will doubtless increase its pay ratio as many of those workers will be brought in at low “training” wages, but the net result of this development is an unalloyed good for the labor market and the company’s workers.

If we are stuck with the pay ratio rule—and, absent Dodd–Frank reform, we most certainly are—then one way we could improve the statistic so that it measures something useful would be to adjust for part-time workers. Companies like Weight Watchers and McDonalds employ a bevy of part-time workers who may log as few as 10 hours a week. Using their data to determine the “median” informs absolutely no one of the true status of income inequality.

Such a fix would not be all that difficult. Robert Pozen and Kashif Qadeer of MIT’s Sloan School of Management suggested in a 2018 Wall Street Journal op-ed that firms could simply consider full-time equivalents in their calculus, something that is commonly done in other contexts. The SEC could accomplish this by issuing an administrative guidance document allowing for part-time pay to be annualized.

The pay ratio rule is, in fact, incongruous with the body of regulations promulgated by the SEC. By law those regulations must be intended to promote capital formation, increase market efficiency, or facilitate investor protection. It is difficult to argue that the pay ratio rule advances any of those goals, especially given the previous requirement that CEO compensation be publicly disclosed.

U.S. corporations currently keep two different sets of data: one for investors and the other for the government for the purpose of reporting taxes. Keeping those two disparate books may seem redundant, but it’s for a good reason: the information that the IRS requires for tax purposes is not necessarily relevant for discerning how a corporation is actually performing.

The pay ratio represents a datum that is not relevant to either investors or the IRS. The ratio is put forth for those who feel compelled to check the behavior of companies to ensure that they hew to whatever social standards policymakers are embracing at the moment. This sets a troubling precedent: will we soon be asking firms to provide other data that are irrelevant to management, shareholders, or the tax authority?

Readings

  • “Executive Pay: Perception and Reality,” by Robert J. Grossman. HR Magazine, April 1, 2009.
  • “Performance for Pay? The Relation between CEO Incentive Compensation and Future Stock Price Performance,” by Michael J. Cooper, Huseyin Gulen, and P. Raghavendra Rau. Social Science Research Network working paper #1572085, November 1, 2016.
  • “The CEO Pay Ratio: Data and Perspectives from the 2018 Proxy Season,” by Deb Lifshey. Harvard Law School Forum on Corporate Governance and Financial Regulation, Oct. 14, 2018.
  • “The Fix for Misleading ‘CEO Pay Ratios,’ ” by Robert Pozen and Kashif Qadeer. Wall Street Journal, March 20, 2018.