Why We Need to Focus on Pay Equity in Government

A recent headline on an HR website alerted private employers to renewed commitment to dealing with a long recognized problem in the United States: “Biden admin likely to act on pay equity, systemic bias, says former EEOC official.” That is consistent with the new administration’s public policy goals and with statements President Biden has made in the past. In a 2018 speech at Brookings, Biden focused on “the hollowing of the middle, income inequality and lack of opportunity.” 

Support for equity and the elimination of bias or discrimination is virtually universal, especially in the public sector. But it's also clear the debate remains an issue in every sector including the federal government.  Last year the Government Accountability Office reported that men and women now have, on average, the same years of federal experience, but women are still paid less.

Statistics do not tell the whole story. Equity often is in the eye of the beholder. Further, U.S. culture makes it important to recognize and encourage star performers. That starts in our earliest years—the pressure to achieve A’s in school contributes to a shared belief that “good” performance should be encouraged. And we all live near Lake Wobegon. It’s not a simple problem.

For women, it started as the demand for “comparable worth” in the 1970s when the focus was on the difference in pay for “men’s jobs” and “women’s jobs.” It was then that the courts confirmed that male and female jobs requiring “equal skill, effort, and responsibility, and which are performed under similar working conditions” have to be paid the same. 

Here it’s important to highlight a nuance central to the debate. The “equal job” issue is governed by job evaluation or classification. That was a hot topic in the HR community through the 1980s, with pay experts introducing variations on the use of statistical models to assure gender was excluded as a factor in those decisions. The models were never really accepted, however, because few people are comfortable with multivariate statistics. 

A lot has changed since then. The interest in pay equity began to wane in 1990 when the recession pushed it to a back burner. By then women had made considerable progress in moving into what had been male dominated jobs. That decade also saw the beginning of the revolution in the management of work, with a shift from seniority to individual performance and capabilities. In 1990, women earned 73% of what men earned. By 2000, the gap had narrowed to 80%—an improvement, but still, men earned in four days what women did in five. Today, the gap is 85%, according to Pew Research. 

The Black-white pay gap actually widened over the four decades by roughly 10% for both Black men and Black women. The ratio for Black men fell from 80% to 70%; for women the decline was from 95% to 82%. 

The change in the Black-white ratio highlights an important issue—the lack of meaningful data. The Bureau of Labor Statistics refers to the puzzling trend as “unexplainable.” Simple ratios can be damning but tell us little. 

The BLS refers to a study by the Federal Reserve Bank of San Francisco where economists tried to explain the trend. They conducted a multiple regression analysis (when I taught statistics, this is where eyes begin to glaze over) using variables for occupation, industry, educational attainment, age, state of residence and part-time status. Still, accounting for these variables, one-third of the gap remained unexplained. 

The unexplained percentage could be attributed to discrimination. However, there are additional factors that over the course of careers affect individual salaries but cannot be quantified. Quality of education is one—not all schools are equal. The labor market also recognizes that some college majors command higher pay than others. Using state of residence is also suspect since in every state significant differences in pay exist from urban to exurban locations. That is why Congress enacted the Federal Employee Pay Comparability Act. 

As the economists at the Federal Reserve Bank of San Francisco noted, “theory tells us that workers’ earnings should largely be a reflection of their productivity” and that “this component can be measured indirectly using data on characteristics such as age, education, and industry/occupation.” However, while those variables are widely used as proxies to test or explain pay differentials, age and education are correlates, not measures of productivity. Moreover, the so-called dummy variables used for occupation and industry do not begin to explain pay differentials for specific specialties (e.g., cybersecurity) or industries like Wall Street.

In December, the Equal Employment Opportunity Commission announced a new interactive data search and mapping tool, named “EEOC Explore,” that “enables stakeholders to explore and compare data trends across a number of categories, including location, sex, race and ethnicity, and industry sector without the need for experience in computer programming or statistical analysis.” Despite the technical concerns, it looks like statistical analyses will get wider attention in the future.

Implications for the Federal Workforce

Federal agencies obviously have the same obligation for equitable practices. How would federal statistics compare? Was the GAO methodology the “right” approach? What variables are relevant? When the civil service system was created, workforce management was simple—women and minorities had limited job opportunities and realistically did not compete for the jobs reserved for white males. There were no employment laws. That was true in every sector. 

The problem is further complicated now by the importance of competing for and retaining the better qualified talent. That has surfaced repeatedly since technology has become so important. The recent SolarWinds hack highlights the essential importance of specialized talent. 

The problem is further complicated by all the separate salary systems and locality schedules created over the past three decades. The methodology developed by economists and statisticians, now promoted by the EEOC, was not developed for employers with as many locations and occupations as government. It also fails to consider individual capabilities or performance adequately.

That does not refute the importance of EEOC’s goal. Federal agencies need to reconfirm the commitment to fair and equitable employment opportunities. Biden’s management philosophy should be reflected in workforce management across government. His promise to “pursue a comprehensive approach to advancing equity for all” defines the goal. It is important for leaders in every organization and at every level to reinforce the importance of equitable employment decisions.

A first step for agencies would be simply to develop dashboards to highlight the commonly used metrics widely used to summarize key workforce issues like pay equity. The metrics should be developed and displayed for every federal facility. Creating the dashboards calls attention to the need to create a more diverse workforce as well as those facilities where metrics suggest problems.

There is an old cliché: “What gets measured gets done.” That is not always true in government but it is true that if a problem is not documented and the information shared, it's not going to be solved. Reporting employment metrics only calls attention to the issues, but it is necessary. 

Google sets an example for aggressively working to achieve diversity goals, with progress reported each year in its Diversity Annual Report, something the company started in 2014. Google “people managers” work closely with diversity experts “to identify opportunities on their teams that are helping us meet our company-wide 2020” diversity objectives. Their efforts have produced results. Many of Google’s jobs are in the highly competitive technology specialties. Each agency should make similar progress reports available to the public.

As important as pay equity is, discrimination should be mitigated at each stage of an employee’s career. Equitable treatment should be a concern in all decisions affecting employees. Google’s approach is a good model. Employees need to understand the extent of their employer’s commitment. Instead of canceling diversity training, it should be expanded along with adding diversity experts.

The start of a new administration makes this an ideal time for agencies to renew their commitment to fairness and equity. Statistical analyses are valuable for identifying possible problems. Task forces would be well suited to the follow up as needed to assess situations and develop solutions. Government should take the lead, not emulate the practices of companies like Google.

The planning for pay equity analyses needs to be coordinated by the Office of Personnel Management. However, a search on OPM’s website for “pay equity” produced only a couple of relevant reports. A 2019 survey reported by the Society for Human Resources found 60% of the respondents were “working to resolve pay inequities.” Agencies should make equity a higher priority.