Skip to main content
Article | Executive Pay Memo North America

New challenges arise for 2020 CEO pay ratio calculations 

Governance Advisory Services |Executive Compensation
COVID 19 Coronavirus

By Jamie Teo , Rich Luss and Steve Seelig | June 2, 2020

Employing a dynamic model to anticipate headcount and pay changes later in 2020 is strongly recommended.

For many organizations, 2020 is likely to involve the most work on identifying the pay for the median employee since their initial calculations in 2017-18. The unprecedented headcount and pay changes many companies experienced thus far due to COVID-19 have dramatically shifted pay demographics at those organizations, particularly for global companies. While some of these effects may be reversed as economic conditions change, it is unclear just what the timing of these actions will be and how they will impact identifying who the median employee is and what that individual is paid.

Many companies have also seen CEO pay reduced or adjusted, with some companies not yet knowing precisely where pay levels will end up at the end of this fiscal year. Now is the time to consider the challenges the CEO pay ratio calculation for 2020 is going to present, challenges that have not existed until now.

This article is not intended to be a compendium of all the issues companies may confront in the 2020 calculation — those details are in a prior article (link provided below). Rather, it highlights key issues to start considering now to help devise a resilient game plan.

Companies are changing their pay programs and organizational demographics

According to a recent study by the Federal Reserve 19% of all adults reported either losing a job or experiencing a reduction in work hours in March, compared to only 7% who increased their work hours or worked overtime. In the same survey, three out of four individuals said their employers expect them to return, but have not been given return date. These actions were more frequently targeted at lower paid positions and would have tended to reduce the population disproportionately from individuals below the 2019 median at many firms.

In the Willis Towers Watson 2020 Returning to the Workplace Survey, almost half the respondents said they are somewhat or very likely to reduce their workforce if it becomes necessary. The typical company considering a reduction in force (RIF) expects it to impact 10% of the active population in any targeted employee group and up to 25% of those in the group who had been furloughed. Among those that have yet to undertake a RIF, most anticipate doing so in Q2 or Q3 of 2020 if they do make one. It is hard to predict how these actions will affect median pay levels, but companies need to be ready to calculate the impact if they take any of these actions.

In another survey we conducted on cost management actions, one out of 10 companies indicated having already adopted targeted pay actions, while another four in 10 are planning or considering doing so. Reductions indicated are typically between 10% and 20% and expected to last four months. On the flip side, some companies are raising pay for “essential” employees (a term many use colloquially to cover more than those legally defined as such). Many of these employees were in lower paid positions prior to receiving these premiums. While these pay premiums are temporary, most organizations have not set a time line for when they will discontinue them, indicating they may last a long time and have the impact of moving those individuals above the median pay level.

As one can imagine, the combination of workforce actions, pay reductions and pay increases — either already taken or on the horizon — can add considerable complexity to this year’s pay ratio calculation.

Start now to avoid surprises

For companies that relied on the three-year rule permitting them to use the same median employee for the 2017, 2018 and 2019 calculations, 2020 is going to be a year of recalculation. Those that did a recalculation of the median employee during one of these years will have the same challenges in addition to those we detailed in an earlier article for those making Year 2 (2018) pay ratio decisions.

Given all the changes that have occurred and may occur during 2020, companies should develop a dynamic model now, to avoid competing against other HR priorities that will fall in Q4. Changes in compensation programs that could take place during Q4 could include bringing back furloughed employees or taking further actions to reduce pay or headcount before year-end. If a company is performing a recalculation in 2020, this analysis may influence the Determination Date it selects. For those companies still hoping to use the three-year rule using the median employee identified in 2018 or 2019, these changes during 2020 could mean they must start anew to determine a median employee for 2020.

Assuming companies intend to use the 2020 median employee in ensuing years, global pay demographic changes as a result of companies restoring stability will present complications in the Year 2 and Year 3 calculations. While companies may aspire to use the three-year rule, for many 2020 may simply not be an appropriate baseline going forward as demographics continue to change from 2020 to 2021.

Complications around furloughs

The question for 2020 calculations that consistently arises is how to count employees and their compensation levels in the wake of furloughs, layoffs and RIFs. The same rules continue to apply for determining employee headcounts as of a Determination Date established during the last three months of the fiscal year.

More vexing is the question on the impact of furloughs for a portion of 2020. This can impact both:

  • Headcount: Headcount levels for those on furlough during the last three months of the year
  • Annualization: Pay levels for those determined to be employed on the Determination Date who were furloughed for a portion of time during 2020

The SEC addressed this in its Compliance & Disclosure Interpretations (C&DIs), Question 128C.04, essentially leaving it to the employer “to determine whether furloughed workers should be included as employees based on the facts and circumstances.” The CD&I does not instruct companies on how they would determine employee status; only that employees would need to be placed into one of the four categories once status is established: full-time, part-time, temporary and seasonal employees.

Once headcount is determined, compensation for full- or part-time employees can be annualized, consistent with Instruction 5 of Item 402(u). That instruction states that a registrant may annualize the total compensation for all permanent employees who were employed for less than the full fiscal year, or who were on an unpaid leave of absence during the period. Annualization is not allowed for temporary or seasonal employees.

Figuring out what the regulations might mean for COVID-19-related furloughs is not straightforward. The guidance states that companies can use the tax or employment law designations of their workers as a guidepost for determining employee status. The facts as to how companies treat employees in the U.S. during COVID-19 related furloughs are unique to that company, with the goal often being to continue to fund healthcare and other benefits. Some examples include:

  1. Continuous leave without pay (e.g., three months) where workers remain on payroll as employees
  2. Intermittent leave without pay (e.g., one week on, one week off),
  3. Hours per week reduction (e.g., workweek reduced from 40 hours to 32)
  4. Straight layoffs where the employer may or may not bring back the workers when able.

The preamble to the regulations provides some clues as to how pay status may be determined under each of the above circumstances, stating that:

Annualization is allowed under the rule for full-time and part-time employees who did not work for the registrant’s full fiscal year for some reason, such as they were employees who were newly hired, on leave under the Family and Medical Leave Act of 1993, called for active military duty, or took an unpaid leave of absence during the period.

Companies will need to resolve whether reduced hours are more akin to circumstance 1 above (this would seem to be within the preamble’s notion of an unpaid leave of absence where annualization is permitted), compared to 2 or 3 (this could mean the employee is considered part-time for their furlough time, thus annualization is not permitted).

International companies will have even more factual circumstances to parse through and should take great care not to apply U.S.-centric resolutions to the entire workforce. Note also that the impact of furloughs, layoffs or RIFs on headcounts could significantly change the countries that may be excluded under the 5% de minimis exception for non-U.S. employees in 2020 relative to prior or subsequent years.

Payroll volatility and determining the median employee

While resolving the furlough question will be daunting, the impact of annualization only matters to the extent it causes an individual who was above the median in pay to be below the median in pay (or vice versa). Pay reductions among lower paid employees already well below the median or modest pay reductions for those well above the median will have no effect on the calculation. On the other hand, pay reductions among employees at or near the median, and even those above the median, may cause the median pay level to shift downward if the company is unable to annualize their pay.

Not only will companies need to contend with headcount and pay fluctuations that have occurred during 2020, any decisions about selecting the median employee should be made with an eye toward future pay ratio calculations if the company wants to continue using that person for 2021 and 2022. We have consistently advocated for companies to select a median employee whose Summary Compensation Table (SCT) total compensation levels are likely to be consistent from year-to-year to avoid fluctuations in the pay ratio denominator. Doing so for this calculation cycle could be more challenging. The pay level of the median employee for 2020, and those similarly situated, may be significantly different than that individual’s pay levels for 2019 or in 2021 and beyond, for any number of reasons including workforce actions taken in 2020, or in future years.

The priority should be to try to find someone whose pay is likely to be representative both this year and next, once all the actions that are likely to be taken occur. In some cases that will involve selecting an individual whose pay has changed substantially this year and is likely to change substantially again next year, if that person’s pay changes are representative of the changes occurring for other employees paid at or near the median.

Using a dynamic model that employs reasonable estimates and statistical sampling can help avoid last-minute scrambling

Each of the discussed scenarios above can present challenges to companies in their data gathering and calculations, each of which can be eased by using statistical sampling methodologies. We have previously noted the many different forms of statistical sampling endorsed by the SEC in its guidance and for which we previously provided a detailed description of methods available.

Developing a dynamic model based on the use of reasonable estimates and statistical sampling is the approach most companies we work with have taken in prior years. For 2020, we think this approach is even more important to consider for identifying the median employee amid global pay demographic volatility. It leverages the regulations’ flexibility to use reasonable estimates to quickly model changes to a company’s global pay demographics as workforce and pay actions are made. In addition, through statistical sampling, companies will be able to identify multiple potential employees within a range around the median (“medianable employees”) who might be considered the true median employee for disclosure purposes.

Companies that limit themselves to the employee at the precise median, may find this individual’s SCT pay varying significantly year over year, relative to others in the “medianable” group, and thus may not be representative of the median in future years. Instead, identifying multiple “medianable” employees and then selecting one individual from that group enables the company to select a median employee at a unit where workforce actions have not been and may not be taken, rather than picking an employee who fell to the median as a result of a temporary pay reduction.

This approach can also help companies pick a wide enough group of “medianable” employees so they can choose someone unlikely to move out of that group in 2021 or 2022. The dynamic model will also allow the company to identify other individuals who might be selected as the median employee if other pay or workforce actions are taken later in the year, eliminating the need to search for a new median employee during these changes.


Director, Talent and Rewards (New York)

Rich Luss
Senior Director in Research

Senior Director, Executive Compensation (Arlington)

Related Insights

Contact Us