Apples, Oranges and Kumquats: Analyzing Federal Pay

This should not be a political issue.

News stories about federal pay and benefits are rarely comical, but every now and then they highlight a claim that is so blatantly political that I have to smile. The latest is about a Cato report focusing on the gap between the average wage for federal civilian employees, $84,153, and the average for the private sector, $56,350 — almost a 50 percent difference.

The report also compares the average total wages and benefits for government workers, $119,934, and for private sector workers, $67,246.

The numbers are based on Bureau of Economic Analysis data. What is not clarified in the data is the impact of part-time employees. The reported compensation levels are for full-time equivalent employees. According to the Bureau of Labor Statistics, there are almost 20 million part-time workers in the private sector — roughly 16 percent of that workforce. They are typically paid less than a full-time employee ($791 versus $240 per week in 2014), and the common benefits are provided to less than half. Federal agencies have comparatively few part-time workers. That has to pull down the private sector averages.

The demographics of the federal workforce are also an issue. Federal workers are older, have longer tenure, and are better educated. All of that is related to higher salaries and benefit costs.

More importantly, high-income occupations are overrepresented in the federal workforce. Physicians, lawyers, engineers, scientists and computer specialists are all well paid and found in significantly larger percentages. Lawyers, for example, account for less than 1 percent of the U.S. workforce, but they seem omnipresent in government. It’s not just the numbers, however. Government has to deal with problems that are often more complex than those of even the largest companies, and that demands well-qualified employees. The labor market pays those individuals better than full-time equivalent salaries.

It would have been only slightly less useful to compare the average salary and benefits of federal physicians with the average for fast food workers in Washington.

A far more useful comparison would be the pay of physicians and nurses working at the D.C. Veterans Medical Center with their counterparts working at other area hospitals. That would be consistent with the universal practice in health care and the private sector. The Federal Wage System is based on that logic.

Pay comparisons are only meaningful when the employers are competing in the same labor market. For pay planning, the comparisons have to be for similar jobs. That is basic to salary management. Many federal jobs are definitely comparable to those in other sectors, and a comparison of salaries is fully appropriate. The Cato report is provocative, but it is not at all helpful to those who may be involved in salary planning.

Despite the questions about the report’s conclusions, it highlights a serious national issue — but it’s not that federal employees are overpaid. The national issue is the depressed wages and reduced retirement incomes affecting a large percentage of the population. There are millions of workers who would benefit from increased job security, higher wages and better benefits. That wins political support, but it should not undercut government’s need to compete for talent.

The General Schedule system and the Office of Personnel Management / BLS methodology for calculating the pay gap at the heart of this problem. There may be no one who truly believes that white-collar federal employees across the country are paid 35 percent less than employees in other sectors.

The annual report of the Federal Salary Council is obviously not accepted as definitive, and that opens the door to competing analyses. There have been several prior studies showing federal employees are paid above national averages. Others will undoubtedly follow. The “premium” is always different but the pattern has been consistent.

As long as this standoff continues, federal employees and their pay can expect to be the focus of the critics. Neither side at this point has the credibility to be convincing. Nor is it realistic to think their analytical strategies will win the day. Neither is easy to understand or suitable for planning a new future.

The argument that federal employees are overpaid will no doubt surface in the 2016 elections, quite possibly more than once and by different candidates. For the critics of government it’s a magnet.

The obvious and only alternative is the universal private sector practice — using salary surveys to compile market pay data. It’s already used for federal medical specialists. It’s the basis for the Federal Wage System. Market data are used by financial regulatory agencies as well as the Federal Reserve. The National Security Agency’s HR office completes a market analysis each year.

The advantage is that the surveys are easy to understand; it’s not a statistical black box. OPM considered ditching the BLS survey not too many years ago, but agency officials decided no single survey provider met their criteria. But by combining surveys — one estimate is that 2,000+ surveys are conducted annually — a credible, defensible database could be compiled. The cost is likely to be less than 5 percent of the BLS budget to conduct compensation surveys.

No, the results are not going to show federal salaries are 35 percent below market levels. But market data for the Washington area will confirm federal employees are not overpaid.

Benefits will also need to be studied. Significantly, the funds in the fiscal 2016 budget for pay and for benefits (for current executive branch employees), 69.5 percent and 30.5 respectively, are split in a pattern that is virtually identical to what BLS reported recently for the private sector — 69.3 percent and 30.7 percent. But it is important to confirm federal benefits are justified. A key is that the comparison is with other large employers.

A solid market analysis should take federal pay out of the crosshairs. This should not be a political issue.