Federal Pay is Way Off Track

The General Schedule system is unresponsive to labor market dynamics and agency staffing needs.

A good contact sent me an old 2002 report produced by the Office of Personnel Management, “A Fresh Start for Federal Pay: The Case for Modernization.” It could have been written last week. It was based on analyses and conversations with a long list of people over a three-year period.  

The authors gave six reasons the pay system was failing government:

  • “The Government asks its agency leaders to face new and unprecedented management challenges using an antiquated pay system. Work level descriptions in law that date back more than 50 years are not meaningful for today’s knowledge-driven organizations.” [A large but unknown number of class standards have not been updated since the report was released.]
  • “The current pay system does not reflect market pay levels. Instead, pay increases and locality adjustments result from a cumbersome and costly measurement system that may be trying to answer the wrong questions.” [They could have added the system has lost credibility.]
  • “It has minimal ability to encourage and reward achievement and results. Over 75 percent of the increase in Federal pay bears no relationship to individual achievement or competence. [The authors could have added that ratings are badly inflated.]
  • “Its structure suits the workforce of 1950, not today’s knowledge workers. In 1950, over 75 percent of Federal workers – mostly clerical – were in grade GS-7 or below; today that percentage has dropped to less than 30 percent.” [Cybersecurity problems were not a public concern in 2002.]
  • “Its prescribed procedures and practices effectively preclude agencies from tailoring pay programs to their specific missions and labor markets. It is unlikely that a common and highly structured system is appropriate for positions as diverse as those found in agencies such as the National Weather Service, the Social Security Administration, the Defense, Finance and Accounting Service, and the Centers for Disease Control.” [COVID 19 has added a new dynamic to staffing.]
  • “It is disintegrating. Through special authorities, a number of agencies already have begun to move toward more modern systems, and our ability to promote common policies across the government where appropriate is diminishing.” [Added to this is the mushrooming number of jobs and locations paid under the special rate authority.]

In the three years the report was in production, the country elected a new president from a different party, experienced the 2001 recession and the 9/11 terrorist attacks. Since then, the Defense Department’s National Security Personnel System failed for reasons that were never fully explored.

An added reason the system has failed was captured in a recent story, “House Lawmakers Call for Pay Parity Between Military and Civilian Federal Workers.” Absent from this discussion was any reference to how federal salaries compare with the salaries for similar jobs in other sectors. The point is that the logic for estimating the pay gap is disconnected from the goal in analyzing pay data in business, healthcare, and higher education — to determine competitive salaries. Neither the Bureau of Labor Statistics or OPM considers competitive pay levels. Pay increases have been political decisions for decades. 

Paying competitive salaries is not the GS program’s goal. When the system was created in 1949, that was not a consideration. There were no salary surveys in the post war years.  

The focus on increases reflects the interests of employees in a system where salaries follow a rigid schedule. Responses to the annual employee survey show 63 percent are satisfied with their pay. (The percentage in private sector surveys is generally lower.) Their satisfaction contributes to low federal turnover rates; employees generally do not leave for higher paying jobs. Further, the Salary Council’s union members emphasize finding a “single number” — the increase needed to close the pay gap. 

Salary Council Reports

Council reports now regularly confirm its union representatives and the three “political appointees” disagree on the logic for conducting data analyses. The union members emphasize closing the pay gap as required under the Federal Employee Pay Comparability Act. They also support adding new locality areas, which increases salaries. Of course, no president since the 1990 Federal Employees Pay Comparability Act was signed into law has adopted the Council’s recommended increase.

At the same time, the three appointees — Council Chair Ron Sanders and members Jill Nelson and Katja Bullock — want to expand the analyses to focus on hiring problems. Government has had an ongoing problem hiring talent with essential skills. Starting salary levels of course are an issue in hiring. BLS does not produce pay data relevant to understanding starting salaries.  

OPM’s 2002 report added an issue not generally addressed in reports on pay. Compensation strategies in other sectors consider how reward opportunities are expected to affect the caliber of applicants. As often stated, you get what you pay for. Some companies align salaries, for example, with 75th percentile pay levels to attract better qualified applicants. That’s solidly consistent with economic theory.

The problems of staffing technology positions illustrate how important this is. The discussion in the new report, “Future of the Federal IT Workforce Update,” explains clearly why a market competitive pay program is essential. Raising starting salaries will increase the number and caliber of applicants.

But it's ignored in administering federal pay. As OPM’s report states, “the pay system takes into account only the fact that the employee meets basic qualifications requirements . . . The statute then specifies that the pay for a newly hired employee must be set at the minimum rate of the grade of the employee’s position, unless the employee has unusually high or unique qualifications or the agency has a special need.  An appointee with better than average qualifications still gets the minimum rate.” 

The Salary Council is now effectively split. The report from last May concludes with a three-page statement from the union members arguing to retain the GS system but administer it as expected when the Federal Employee Pay Comparability Act was enacted. That was followed by an eight-page argument from the chair and appointed members discussing the need for a “major redo.” The report itself confirms the two groups are not on the same page. 

To repeat a phrase from the report, “at the risk of generating some controversy,” the three appointed members argued for changes that would amount to replacing FEPCA. They contend that government should shift to aligning with market pay levels “the way every other major employer does.” In other words, government should adopt private sector practices.

They are correct that when employers in other sectors are unable to attract essential talent, they look beyond comparative pay levels to understand why. Recent Council reports have cited the need to understand the impact of local housing costs, commuting problems, and the local balance of supply and demand for specific skills. An unmentioned factor is the employer’s “brand” or reputation. The only issue recognized in FEPCA is the pay gap.

Their discussion, however, misses two virtually universal practices that do not fit easily in the civil service paradigm. First, the focus is always on specific talent competitors. In competing for talent, hospitals plan their pay programs to compete with other area hospitals. Colleges limit their analyses to pay data for other area colleges. That’s true in every industry group. That explains the industry focus common to pay surveys across the country. Second, in developing comparative data, those employers all have greater flexibility to adjust starting salaries as needed to stay competitive. They pay what they have to pay. Rigid salary systems like the GS system were eliminated years ago.  

The appointed Council members also argue for adopting a total compensation approach. All employers have to be concerned with payroll costs. Starting in the 1980s the pressure of global competition prompted employers to start reducing what they spend for benefits. The most prominent change has been replacing defined benefit pension plans with defined contribution plans. It makes business sense for employers in the same industry to offer similar benefit packages to new hires.

They are solidly correct in their argument but six additional points need to be considered.  

  1. Federal benefit costs are naturally higher simply because of the average age and longer service of federal employees. Costs also vary with prior funding patterns and actuarial assumptions.
  2. It's rare in the surveys conducted across the country to find cash compensation and benefit costs combined. The two are typically summarized and reported separately. Further, the analyses are often not based on the same group of employers.
  3. BLS surveys currently ignore the added compensation from incentives as well as the income from stock ownership. Both are integral and significant components of the total compensation of many white-collar workers. Ignoring those components understates planned compensation.
  4. The majority of employers in BLS surveys are small, locally managed businesses that provide little beyond legally required benefits. Including the smallest businesses understates the cost of benefits among those employers competing with government for talent. 
  5. Benefit levels and costs vary by industry, staffing and turnover patterns, age of the organization (recently created companies provide fewer benefits), and the mix of local employers.  
  6. Benefits are more important to older workers. Government needs to transition to benefits that help in competing for young talent.  

The argument for focusing on total compensation is completely reasonable but difficult for government to operationalize.  

The bottom line is that government has a workforce problem. The federal workforce is aging and heavy retirements are inevitable. Vacancies and skill shortages in key fields are reported to be causing problems. Newer, high demand skills are the future. The country’s problems make it clear — Agencies need highly qualified talent to address issues that are more complex and sensitive than those confronting private employers. The seniority-based increases contribute to the common culture of compliance. Now the pandemic underscores the importance of recruiting and retaining the best and the brightest.  

BLS surveys are not in any way helpful. Simply stated, BLS surveys do not generate data relevant to managing pay. Further, the GS system is unresponsive to labor market dynamics and agency staffing concerns.

A New Commission on White Collar Pay

The answer proposed by the Council’s chair and appointed members is to create a bilateral commission similar to the Quadrennial Review of Military Compensation to undertake a similar review for civilian pay and benefits.  

The proposal is fully warranted. It’s clear the Federal Salary Council has lost its relevance. As stated in Attachment 8 of the May report,the Council is “in danger of becoming largely irrelevant.”  

However, it’s also clear there have been a number of prior studies over roughly half a century concluding the GS system needs to be replaced. The National Academy of Public Administration has published at least three. The OPM report is the best at defining the problems along with a framework for a replacement system. The many reports have spelled out clearly a picture of a new pay system.

If there is to be a new commission, its mandate should be broad enough to understand how the GS system affects workforce management. Elected leaders should be the intended audience and make it clear that a new pay system is needed along with the impact on government performance. A commission could also develop a timetable so everyone would know what to expect.

Possibly the most important step would be a bilateral market pay analysis by job series and location. Data identifying jobs that are overpaid or underpaid would help to resolve what has become a contentious debate. Professional groups should be asked for input on relevant surveys and job matching.  

In addition, a commission should undertake the following additional studies:

  • Summaries of the experience in off-budget federal and state governments with pay reform. The Tennessee success story is possibly the most recent.
  • Agency by agency workforce analyses to understand anticipated retirements and turnover as well as current vacancies. Data are needed to understand the parameters of the staffing problem.
  • Surveys of student career plans to understand the job and organizational factors that influence their preferences. A 2019 survey by the National Association of Colleges and Employers found only 2.1 percent planned a federal career. That is a problem.
  • Focus group discussions and exit interviews to understand the work experience factors contributing to early turnover of new hires. That is also a problem. Studies from other sectors would be informative.
  • Analysis of emerging skill requirements, emphasizing job series where federal agencies should plan for shortages.  Professional associations would no doubt agree to help.
  • A focused analysis of the experience of government organizations that have attempted the shift to pay for performance. That should include NSPS. 

Here the focus should be broader than the reviews of military pay. A new commission should address both staffing and the role of financial rewards in creating a high performance work environment.

The March report from the National Commission on Military, National, and Public Service presented an important argument for once again making government an employer of choice. Their recommendations recognize the federal personnel system needs to be replaced to enable agencies “to replenish the workforce . . .  with younger Americans and workers with critical skills . . . in the face of a looming wave of retirements.” That should be a priority starting next January.