Benefits of CEO Pay Ratio Guidance

After hearing that the CEO pay ratio rules were still cumbersome and difficult to maneuver, the Securities and Exchange Commission (SEC) recently issued three pieces of guidance that will markedly improve the process, especially for global companies. We believe that the SEC is now much closer to its goal of providing flexibility in a manner that would “reduce costs and burdens for registrants” that should prompt companies to revisit their approaches.

The SEC guidance was in three parts that included an Interpretive Release, new Division of Corporation Finance Guidance on Calculation of Pay Ratio Disclosure and amended and restated Compliance & Disclosure Interpretations (C&DIs).

Key aspects of the changes include:

Defining Employees: Unfortunately, many companies already have spent a great deal of time and effort applying the guidelines from the former C&DI Question 128C.05, which required them to review whether the company or its consolidated subsidiaries determine the pay levels of those who work for a third party. That C&DI has now been withdrawn, with a new “safe harbor” rule inserted into the Interpretive Guidance.

Companies that already have done the work under the prior rule have choices:

• Continue to use the existing conclusions reached because the regulatory guidance is still in place. It states:

“The definition of employee or employee of the registrant does not include those workers who are employed, and whose compensation is determined, by an unaffiliated third party but who provide services to the registrant or its consolidated subsidiaries as independent contractors or ‘leased’ workers.”

So any application of this rule, even one that used the guidelines from the former C&DI that focused on whether a company “determines” the pay of a third party worker, would likely be fine.

• Follow the Interpretive Release’s guidance, which says in reference to the regulatory language cited above: “The provision was not intended to serve as an exclusive basis for determining whether a worker is an employee of the registrant. Accordingly, we believe it would be consistent with Item 402(u) for a registrant to apply a widely recognized test under another area of law that the registrant otherwise uses to determine whether its workers are employees.” The regulatory guidance referenced the definition of employee for U.S. tax purposes, but that is simply an example. It is equally plausible this rule would permit companies to follow the rules that are set in a global company’s home country, perhaps just for tax purposes or more expansively for local employment law purposes.

We read the above rules to suggest that any approach a company takes should be consistent. This raises the question about what might be consistent when a company uses “another area of law” in many different countries. Would that mean the company should just use tax law concepts in all countries, or could it use tax law in some and employment law concepts in others? We read the rule expansively so that a company could mix and match, based on whichever approach is commonly applied in a home country.

A more vexing issue would be for a company that wanted to mix and match the regulatory rule with the “safe harbor” rule. If those different approaches were used in good faith, that approach would likely also be fine. However, a company might take pause if selective mixing and matching drove to a result that would artificially increase (or decrease) median pay.

• Savings and Benefits: Companies need not create an entirely separate process to determine who is an “employee” solely for purposes of calculating the CEO pay ratio.

Easing the risks for potential liability: Even though the regulations were crafted to promote flexible application, companies and their counsel had expressed concerns to us over continued uncertainty that using statistical sampling and reasonable assumptions were permissible for a “filed” disclosure included on a company proxy. We think the Interpretive Release alleviates such pressure:

“In our view, if a registrant uses reasonable estimates, assumptions or methodologies, the pay ratio and related disclosure that results from such use would not provide the basis for Commission enforcement action unless the disclosure was made or reaffirmed without a reasonable basis or was provided other than in good faith.”

Our read is that this rule now shifts the burden of proof from companies to the SEC itself to demonstrate that the approach taken was without a reasonable basis or was in bad faith (the seeming meaning of “other than in good faith”). This does not mean companies can apply the rules without regard to their existing pay demographics, particularly in selecting an appropriate, consistently applied compensation measure (“CACM”), but as we will see, the rules will seem a lot more liberal for those that have resisted using reasonable estimates, assumptions or methodologies.

• Savings and Benefits: Companies and their counsel, previously wary of using reasonable estimates and statistical sampling, should be comforted by an acknowledgment that the burden of proof in SEC enforcement actions is extremely high.  We think this opportunity to use these alternatives can ease data gathering efforts and shorten the process.

Permitting reliance on internal records:  Working in tandem with the above rule, the Interpretive Guidance includes another rule that acknowledges the inherent challenges in getting accurate pay and demographic data for global companies and eases concerns about data accuracy. This portion of the Interpretive Guidance permits companies to use existing internal records, such as tax and payroll records, to determine application of the 5% de minimis exception and to determine the median employee. The former use is relatively straightforward because it only requires determination of headcount on the determination date. The latter is more helpful because it endorses that use of internal records, which we believe can include human resources information systems (HRIS), to determine annual compensation using the CACM chosen by the company.

The SEC adopted this interpretation because companies were concerned that the data in their HRIS systems might not be accurate or up-to-date. This rule eases that concern quite a bit when read together with the portion of the Interpretive Guidance that places the onus on the SEC to demonstrate a company did not act in good faith articulated above. That is, a company that has a reasonable basis or acts in good faith to rely on the data in its systems for its calculation would be able to withstand SEC scrutiny.

• Savings and benefits: By not having to “audit” the accuracy of pay data when they believe in good faith it is an accurate representation of their pay demographics, companies will reduce compliance costs and concern that their systems might not have precise and accurate pay data for all employees.

Can disclose that your pay ratio is an estimate: We are grateful that the SEC Staff determined that every pay ratio calculation that uses a CACM other than summary compensation table (SCT) total compensation is truly an estimate (see our blog post and subsequent comment letter). The salient language from C&DI 128C.06 states:

“Therefore, the staff would not object if a registrant states in any required disclosure that the pay ratio is a reasonable estimate calculated in a manner consistent with Item 402(u). [September 21, 2017]”

• Savings and benefits: This further alleviates concerns about getting a pristine result that works in tandem with some of the other guidance regarding the SEC’s burden to show that a company did not act reasonably and in good faith in its use of statistical sampling and reasonable estimates.

Statistical sampling and reasonable estimates or methodologies: The SEC’s Corporate Finance Staff guidance endorses the approaches we have advocated for our clients and will help eliminate concerns about whether they are doing sampling right.

• Statistical Sampling: The guidance reiterates the notion that any statistical sampling approach selected must be based on observations about compensation distributions.  This may mean different things for different companies, but it reflects that companies have enough information about pay distributions from their HRIS systems or other sources to make observations sufficient for sampling. The SEC provides some examples of the sampling methods that could be appropriate to use (alone or in combination), depending on the company’s circumstances:
  • Simple random sampling (drawing at random a certain number or proportion of employees from the entire employee population).
  • Stratified sampling (dividing the employee population into strata, e.g., based on location, business unit, type of employee, collective bargaining agreement, or functional role and sampling within each strata). This is the approach we have taken with most companies we work with. It helps companies focus only on gathering pay data for those employees paid at or near the median.
  • Cluster sampling (dividing the employee population into clusters based on some criteria, drawing a subset of clusters, and sampling observations within appropriately selected clusters; cluster sampling may be conducted in one stage or multiple stages). This approach is often used to help companies that cannot get pay data from some locations, but are able to craft assumptions about those locations’ pay demographics by observing results from other, similarly paid locations.
  • Systematic sampling (the sample is drawn according to a random starting point and a fixed sampling interval, every nth employee is drawn from a listing of employees sorted on the basis of some criteria).

In our experience, most organizations that use statistical sampling have relied upon stratified sampling and clustered sampling as methods that have been helpful for simplifying the data collection process and reducing the necessary cost and effort.

Reasonable Estimates or Methodologies: The guidance on reasonable estimates and methodologies is equally helpful, with the following items that describe permissible approaches worth highlighting:

• Use of the mid-point of a compensation range: Compensation can be estimated analyzing the composition of the company’s workforce (i.e., by geographic unit, business unit, and employee type). This is helpful for companies that only have direct access to pay grades, job levels or salary guidelines rather than actual pay data. Making a safe assumption that people at those levels are paid a certain amount in different locales can significantly reduce time and effort.
• Identification of multiple employees around the middle of the compensation spectrum:This approach helps companies broaden the group of potential median employees who might be considered the true median employee for disclosure purposes. Companies should not be limited to choose an employee at the precise median using the CACM they elected, especially if that employee is not likely to have consistent SCT pay from year to year or relative to others who are near the median value of the CACM. This guidance seems to permit a choice among those equally likely to be the actual median employee.
• Calculation of a consistent measure of compensation for determining the CACM: We read this guidance to permit companies to make reasonable assumptions about their pay demographics when determining whether the CACM they have selected “reasonably reflects the annual compensation of employees,” as per the regulations. Recall, C&DI 28C.01 provides that “The appropriateness of any measure will depend on the registrant’s particular facts and circumstances.” Being able to use reasonable estimates to make this determination is helpful in that it does not require a detailed analysis up-front to determine if the CACM is appropriate.
  • Examples of Reasonable Estimates or Methodologies: We encourage companies to read the examples themselves to get more comfortable with using these techniques to suit their circumstances. Our observation would be that the breadth of these rules may reach far beyond what companies previously had believed was permissible.  Assumptions and estimates can be made about business unit pay, country pay, pay by job levels, pay for full vs. part-timers, pay of locations similar to those where you have pay – the potential options are expansive.
• Savings and Benefits: The guidance on statistical sampling and reasonable estimates may open the eyes of companies that have been reluctant to use these approaches before the guidance was issued. We believe this will open opportunities to make decisions on data gathering from a cost perspective so that companies can choose the most efficient method of determining a reasonable estimate of the pay of their median employee.