The century-old GS system is 'disintegrating' and government can't agree on how to fix it

Both Republicans and Democrats see it as a barrier to better government but for totally different reasons. It is central to the “compliance culture” that impedes better government. It still reflects the way work and workers were managed a century ago. It has not been modified, except for separating the Senior Executive Service and adding locality pay, since it was created in 1923.

It’s more than ironic that in that era the head of the Bureau of Efficiency was prominent in the administration of the civil service system. The buzzword then was “scientific management.” Workers were expected to do what they were told, and managed as a cost. That is still true in smaller, owner-managed companies. 

Not surprisingly, Elon Musk is known for top-down control. Quotes attributed to him confirm a very negative view of the federal government: “Regulations are immortal.” “The bureaucracy is the problem.” Musk and his chainsaw created an us-vs-them distrust of management. The workforce was “traumatized,” to quote Max Stier, of the Partnership for Public Service.

Now, the government is moving full-steam ahead with AI. It also threatens job security. What has not been addressed is that jobs will be changing rapidly and that is at odds with job classification. A year or two from now the GS system is likely to be unsupportable. AI is exacerbating the already poor employee morale. 

Employee costs are not the problem 

By any standard, the costs attributed to building a productive workforce are a small percentage – less than 5% – of what the government spends. The cost to raise GS salaries to market rates would be less than 1%.

A related point highlights the political problem. No administration has been concerned with how much federal contractors pay their people. Added to that are the organizations receiving federal grants. Those organizations are not expected to defend how much they pay their staff. That’s true of the Federal Reserve System and the many independent agencies. 

There is an alternative management philosophy.  That is employees should be managed not as costs but as valued assets. That emerged in the 1990s. It started a few years earlier with W. Edwards Deming’s book, Out of the Crisis, where he argued the problem was not workers, it was the system they work in. Over the decade Gallup first cited research showing engaged workers are more productive. Fortune's “Great Places to Work” lists first appeared. And research confirmed employees are managed differently in “High Performance Organizations.” Those companies have been high on lists of the best  performers.

It’s not coincidental that in the 1990s the Clinton-Gore National Performance Review (NPR) confirmed “empowered” federal workers are fully capable of significantly better performance. That initiative resulted in the deletion of over 350,000 jobs and savings in excess of a billion. In that context employees were committed to improving operations.

But just before President Bush took office, the Heritage Foundation released the report, “Taking Charge of Federal Personnel.” It influenced the new administration to centralize management with OMB, returning day-to-day operational control to appointees. That effectively ended the brief recognition that employees are ready to play a role in improving performance.  

The Government Accountability Office, then led by David Walker, initiated reform in the late 1990s when budget cuts and staffing reductions forced a strategic restructuring. Walker is known to be a conservative but his reforms emphasized employee empowerment and deep employee involvement in the planning. GAO has been at or close to the top of the Partnerships’ Best Places to Work in the Federal Government Rankings for mid-size agencies since the list was created.

Strategic human capital management has been consistently on GAO’s High Risk list since 2001. 

The Pay Agent’s Recommendations for the GS Framework

In December, the Pay Agent’s “Annual Report” stated it would not “approve of further additions to existing locality pay area boundaries or the creation of new locality pay areas at this time.” It was not like prior federal reports in that it was highly critical of the “antiquated” GS system. From the report:

• The locality areas “do not align with geographic realities or labor market conditions.”
• “It makes no sense to continue expanding locality pay boundaries – or to tinker with the arcane methodology ... given the need for a better pay system.” 
• “ ... a methodology that produces implausible results while ignoring occupational realties, mission needs, and performance considerations.”

It went on to state, “It has long been clear what the locality pay system’s flaws are” and referred to the 2002 Office of Personnel Management (OPM) white paper, “A Fresh Start for Federal Pay: The Case for Modernization.”  The authors contended the GS system “hinders the performance of the Federal Government.” 

As the director of the project in 1990 that led to the Federal Employee Pay Comparability Act, I agree with their critique. It was passed as a rider, and there were too many compromises. The Bureau of Labor Statistics radically changed its survey methodology soon after passage. Over the years BLS has made additional changes. Today it’s so complex it’s doubtful anyone at OPM or involved with the Federal Salary Council meetings could describe the analytic methodology. It is completely different from anything used in other sectors and far more costly. 

The Pay Agent’s report briefly discussed three recommendations “for improving the General Schedule:”

• “Assess the total compensation gap.” This is not a new idea. It has been the focus of several Congressional Budget Office reports arguing the cost of government provided benefits offsets the lower salaries. The flaw in that argument is clear in the BLS statistics. Smaller companies provide fewer benefits and their numbers distort any comparison with government’s true talent competitors. Moreover, BLS does not include cash incentive income or stock related income. It would have to be an “apples-to-apples” comparison to be meaningful. 
• “Provide different pay ranges for different occupations.” This could be a practical answer for high demand occupations. Obviously, the Federal Wage System is based on this argument. It’s also the rationale for the Law Enforcement Officer (LEO) pay system.  And from the recommendations in our 1990 locality pay report, the Department of Veterans Affairs was successful three years later in establishing a separate pay policy under Title 38 for physicians, dentists, nurses and several other medical care specialists. It is the answer for high demand occupations.
• “Eliminate GS steps and create an open range” to shift to pay for performance. While this has been highly controversial, there is a list of pay demos and independent pay systems with successful pay for performance systems. The state of Tennessee transitioned successfully in the years before COVID to a performance based pay policy. The state is a model for how to make the transition. A core point is that this is culture change.  Tennessee invested three years in manager training and practice before actually changing the pay policy. Federal managers and employees are far from ready today. 

The Pay Agent report is accurate – the GS system is “a legacy framework from the 1950s.” “It is disintegrating.” The government needs to provide better services. The public’s support continues to decline.

The 2020 Rand report, “Federal Civilian Workforce Hiring, Recruitment, and Related Compensation Practices for the Twenty-First Century,” evaluated the many successful initiatives to develop “modern” pay systems. A common thread is that employees were involved in the planning and implementation. The most notable failure was the roll out of the National Security Personnel System (NSPS) covering 226,000 Defense Department employees. It initially had employee support but they were not involved and after three years the system was terminated.

Employees want their organization to be seen as a success, and as the Clinton-Gore NPR made clear, they understand the problems better than outside experts and want to be involved in improving performance.

There have been repeated recommendations to replace the GS system. It will be a substantial undertaking, far more complicated than the study leading to FEPCA. However, the history here and in other countries suggests it is necessary to improve government performance.