Pay Analyses: Which Type is Right for You?

Share
"Equal Pay" is everywhere. In recent months, several states have passed legislation around pay equity, creating a patchwork of legal obligations for multi-state employers. Government agencies are increasingly focused on unearthing pay discrimination. Employers of all sizes and in all industries are grappling with new obligations, rising numbers of pay discrimination claims, and increased media attention in the area of pay equity.

Given this climate and complex legal landscape, many employers are choosing to proactively conduct pay analyses in an effort to understand and reduce their compensation-related legal exposure. Robust, self-critical, EEO pay analyses are an effective way for employers to identify and remedy problematic race- and/or sex-based pay disparities before a claim is received. Pay gap analyses allow employers to show overall pay parity between protected groups. While most employers take steps to cloak these analyses in privilege, some employers choose to voluntarily publish certain results to showcase their transparency and commitment to equal pay.

Facing mounting social pressures around pay equity, employers are asking whether and which types of pay analyses to conduct, as well as whether to publicly disclose this information or not.
 
Types of Analyses
EEO Pay Analysis
The goal of an EEO pay analysis is to determine whether there are systemic, unexplained pay disparities between similarly situated men and women or based on race. Given that even similar employees rarely have the exact same pay, an EEO pay analysis will help employers understand whether there are legitimate, non-discriminatory explanations for any observed differences. Said another way, an EEO pay analysis will help determine whether employees are paid fairly based on factors that are relevant in the employer’s pay system. This type of analysis will identify whether a protected group (race/sex) is negatively impacted. It also serves to identify outliers, or individuals who are significantly above or below their predicted pay; determine which individuals are driving pay differences; and ascertain those factors that tend to influence pay in the organization. EEO pay analyses also give employers a clear sense of their potential legal exposure and point out specifically where that exposure lies. If statistically significant disparities based on race or sex are identified, employers are encouraged to make pay adjustments to mitigate the identified disparities. Employers who conduct EEO pay analyses are encouraged to secure organizational "buy-in" to ensure that there is commitment to make changes, including possible salary adjustments, where appropriate.

In terms of methodology, an EEO pay analysis is designed to mirror an employer’s pay system and analyze employee pay in a meaningful, diagnostic way. An employer’s workforce is first broken into groupings comprised of comparable or "similarly situated" employees. These employees may have similar roles in terms of their required skills, job duties, and responsibilities. While differing standards exist under federal and state law, the best practice is to conduct a "Title VII-style" EEO pay analysis using similarly situated employee groups. Additional analyses can be conducted to ensure compliance with the Equal Pay Act or state laws. Notably, under certain state laws, employers may benefit from safe-harbor provisions that are meant to mitigate liability for employers who conduct proactive analyses. After thoughtfully creating similarly situated employee groups, all available variables that might influence pay – such as tenure, performance, function, and education – are tested to determine if each is independently correlated with pay. Those factors that are practically or statistically correlated with pay are then "controlled for" using regression analysis. Pay comparisons are only made within these similarly situated employee groups, allowing for a more meaningful application of job-related factors. For example, education level may significantly influence pay in a group, whereas education may not play a role in another group. An EEO pay analysis allows for group-specific explanations in contrast to a one-size fits all approach. Employers may then drill down into particular problem areas within each group, exploring causes, explanations, and potential remedies for pay differences.
 
Pay Gap Analysis
Employers may choose to perform a pay gap analysis, also referred to as a pay parity analysis, log-linear analysis, or pay ratio analysis. The objective of a pay gap analysis is to provide high-level insights on pay trends across the workforce without categorizing employees into similarly situated employee groups, or considering factors that influence pay. Many employers opt to conduct this type of analysis, which is more limited in scope, in advance of making a public statement about its pay practices. This type of analysis allows employers to state, for example, that women in their workforce are paid 99 cents on the dollar as compared men.
 This [regression analysis] prevents outliers from driving results and excludes premiums associated with particular job titles that may skew results.
A pay gap analysis analyzes an entire workforce in a single analysis. There are two types of pay gap analyses: "controlled" and "raw." A controlled pay gap analysis involves analyzing an employer’s workforce in a single statistical model. However, a robust statistical regression analysis is performed to identify dollar values associated with certain job descriptions, as salaries are standardized into a condensed range. This prevents outliers from driving results and excludes premiums associated with particular job titles that may skew results. Put differently, the regression analysis shows how much an employee is predicted to earn based on his or her job description. Once the dollar value of each job is established, those dollar amounts are removed from the statistical model. The controlled pay gap analysis subtracts the portion of the salary that is due to a particular job description. The analysis then evaluates what remains and compares pay based on gender or race.

To illustrate, assume a regression analysis identified that being an Accountant at a company guarantees an employee will earn $60,000 in pay, while being an Analyst guarantees an employee will earn $40,000. If we have three male Accountants earning $60,000, $65,000, and $70,000, the analysis will exclude the $60,000 guaranteed salary and view these employees as earning $0, $5,000, and $10,000. If these were the only three males in the workforce, the statistical model would determine that the value of being male is, on average, $5,000. Similarly, if we have three female Analysts earning, respectively, $40,000, $45,000, and $50,000, the analysis would exclude the $40,000 and would see the value of being female as, on average, $5,000. As a result, the controlled pay gap would be 0%.

A raw pay gap analysis is, in essence, average male versus average female pay – or, average minority versus average non-minority pay – across an organization. This type of analysis may provide general insights on pay trends across the workforce at a high level, but does not account for factors tied to actual pay practices; fails to consider job-related explanations for pay differences; and does not conform with the groupings that employers are required to analyze under state or federal law. In the same scenario described above, a raw pay gap analysis would take the average pay of males and the average pay of females without excluding the value of the job title. Therefore, the average male pay would be $65,000, and the average female pay would be $45,000. Comparing the average male to average female pay would yield a 31% pay gap.

As is illustrated here, there are significant differences between the different types of pay analyses available to employers. While both an EEO pay analysis and a pay gap analysis may be helpful, it is imperative that employers explore the "why" or purpose of conducting the analysis before deciding which analyses to conduct.
 
To Disclose, or Not to Disclose?
Years ago, employers generally did not publicize information regarding their pay practices or the results of pay analyses, as doing so might generate increased scrutiny and lead to potential legal exposure. Today, pay is a topic that is out in the open. Some governments outside the United States mandate the disclosure of pay gap information. An example of this is the U.K. Equality Act. Since 2018, employers with at least 250 U.K.-based employees are required to publish their gender pay gap on an annual basis. Additionally, over the past few years, a number of large companies have voluntarily disclosed pay gap information touting, for example, their achievement of pay parity or their closing the gender pay gap within their workforce.

Given the reputational and legal risks associated with such disclosures, one might wonder why employers would choose to open this door. As it turns out, shareholder activism, and not necessarily an employer’s willingness to be transparent, has been the driving force behind many of these public disclosures. Activist shareholder groups have pressured large public companies through shareholder proposals to audit their payrolls to identify, disclose, and remedy pay gaps. Even those companies not experiencing shareholder pressures may face pressure nonetheless, as industry peers choose to reveal their own pay gap-related information. Moreover, once companies reveal information regarding the gender pay gap, questions inevitably arise around a race-based pay gap.
 Along those same lines, employers who disclose gender pay gap information may thereafter be pressured to publish similar results based on race.
So, what is an employer to do? The decision around public disclosures of an employer’s pay practices is largely a strategic one and depends on the pressures and public relations considerations the organization is facing. Employers may be hesitant to disclose for any number of reasons. For instance, disclosure may cause discord in the workplace or perhaps trigger targeting by plaintiffs’ attorneys. Employers also may be reluctant to disclose because, though their results may be positive in one year’s disclosure, they may deteriorate in subsequent years. In addition, disclosing results one year will likely create an expectation of disclosure in subsequent years. Along those same lines, employers who disclose gender pay gap information may thereafter be pressured to publish similar results based on race. Critics may pressure employers to release not only summary results, but also the back-up data and information used in the analysis. Employers may wonder whether releasing their results will ever really be "enough."

Weighing these considerations, employers may lean towards disclosing pay gap information due to internal or external pressures; to create transparency in the workplace; or to generate positive publicity. Some progressive employers may opt to disclose due to the effect disclosure has on the workplace: i.e., when employers commit to reporting and make public disclosures, the pay gap tends to improve. Whatever the reason to make this information publicly available, employers who decide to go public should be prudent and strategic in their approach. Of utmost importance is defining the purpose of the project at the outset to ensure the appropriate analysis is done.

Prior to disclosure, employers should consider what to disclose and to whom, as well as be particularly thoughtful about messaging. Importantly, there is no one-size-fits-all approach to public disclosure. Employers who conduct analyses in advance of public disclosure should cloak their analyses in privilege from the outset of the project, ensuring that proper privilege protocols have been established. Employers can, of course, subsequently waive privilege if positive results are achieved. However, an assertion of privilege will likely fail if an employer does not lay the appropriate groundwork up-front

If you have any pay analysis questions or concerns, including whether and what to publicly disclose, please reach out to Michelle or Suzanne.