The critics are right; the General Schedule system is broken and needs to be replaced. It no longer meets the needs of government. The introduction of locality pay 26 years ago is the only change in more than half a century. It may be the oldest pay system in the world. Starting salaries at all levels are not competitive. The annual analysis of pay survey data has no credibility. Grade inflation is a problem. The experienced specialists who maintained the system have retired. The best that can be said is that employees are paid on time.
A new system—or maybe systems—should be based on the proven principles adopted by successful large organizations. The planning should begin by defining the goals. Presumably attracting qualified talent is one goal. Another should be managing pay increases as an incentive to reward performance. Many of the proposals floated by lawmakers and discussed in media reports would be contrary to both goals.
A third goal is controlling costs, but that needs to be balanced with the first two goals. Historically, employees were managed as a cost to be minimized, but progressive employers today manage them as assets. Investing in employees can pay off with significantly better performance.
Mr Trump’s nominees include individuals who have spent years working with successful, highly paid individuals. Their companies know the value of highly talented people. The leaders of those organizations believe in rewarding good performance. They, along with many governors and military leaders, understand the importance of effective people management.
Faulty Comparisons: Federal v. Private Sector Salaries
There have been repeated, conflicting claims about federal pay going back years. The latest report from the Federal Salary Council contends General Schedule salaries are 34 percent behind workers doing “comparable work.” That’s at odds with estimates from the conservative think tanks and the Congressional Budget Office showing federal total compensation, including benefits, is well above levels in the private sector. Until more credible market data is assembled, there can be no agreement on establishing fair compensation.
Each of these analyses is based in a faulty premise. They ignore a key, first step: defining the relevant talent markets and employers competing with government for talent. Government is not competing for talent with the millions of mom-and-pop businesses—the restaurants, dry cleaners, hardware stores, etc., that are prevalent in the databases used in their analyses. Those businesses rarely employ people with the knowledge and skills needed in government. The problems addressed in those small organizations are not at all the same as those now confronting federal agencies.
That is particularly relevant for jobs in management and in professional fields. The ability to attract and retain high demand specialists is critical in many fields. The STEM fields (science, technology, engineering and mathematics) are often cited but government needs well qualified professionals in economics, healthcare, procurement, international relations—the list goes on.
The agencies with non-GS pay systems rely on the textbook answer. That is to say, they (or their consultants) identify employers with similar jobs and survey them to determine market rates.
A simple and logical alternative would be to plan salaries to be competitive with government contractors. That is consistent with congressional intent years ago when the agencies regulating the financial industry were authorized to create salary programs to compete within the industry.
That’s important also because those organizations (along with other large companies) maintain more generous benefit plans. Plus, publicly traded companies grant stock ownership opportunities to executives, managers and at least senior professionals. Those grants should be included in “total compensation” analyses. An added issue is that the federal workforce tends to be older, which drives up payroll costs and distorts comparisons.
Developing Credible Data Through Benchmarking
Reaching agreement on the basis for assessing federal salaries, if that is possible, is important because the new Congress appears ready to reduce pay and/or benefits. To avoid a contentious and unproductive debate, it would be valuable to first conduct an independent market pay analysis. The 1990 project that led to the Federal Employee Pay Comparability Act (FEPCA) was based on an independent market analysis. (Full disclosure: I managed the project.)
A key issue that has consistently been ignored is that no other employer, in any sector, relies on the methods used by the Salary Council, CBO or the think tanks. Their analyses are based on statistical techniques that few understand and defy explanation. In contrast, the usual logic and the data used in salary planning are easy to understand.
In every other sector, the planning is based on benchmark job surveys. Those are jobs that tend to be defined in the same way in large numbers of organization. An entry level chemist or a senior accountant are examples of benchmark jobs.
That was the intent when FEPCA was enacted. It was in the mid-1990s that the Bureau of Labor Statistics terminated a benchmark survey and shifted its focus to the Employment Cost Index. Unfortunately, the FEPCA statute requires the use of BLS data. It’s too complex to explain here, but BLS surveys do not report market pay levels for specific jobs. Surprisingly, perhaps, the annual gap analysis does not compare federal salaries with pay levels in other sectors.
To be fair, the think tank and CBO analyses also fail to report job-specific information. Despite their claims, they cannot identify jobs that are overpaid.
Its estimated that over 1,000 benchmark surveys are conducted annually. Most are available for a modest cost. If all the available surveys were purchased, the total would be a small fraction of the cost of the BLS surveys. The project to assemble market data would be similar but more extensive than that which preceded FEPCA in 1990. To ensure credibility, experts should review the analysis at each step.
The market data would then set the stage for planning a replacement salary program. That will be discussed in a second column.