Another Misleading Claim About Federal Pay

One of the rituals of Washington occurred last week. A new report from the libertarian Cato Institute argued that federal employees are paid above market rates: “Federal Pay Outpaces Private Pay in 2016.” Similar claims have been made annually by one or more think tanks for more than a decade. Normally the reports are forgotten a day or two later.  

This report, however, could be different. It highlights the ease with which both sides to this argument can find competing government data that support their claims. The Cato report opens with a comparison of the average federal wage ($88, 809) in 2016, and the average of private workers ($59,458). The data are from the Bureau of Economic Analyses.

That ignores the vast differences in pay across industry sectors: Employees working in the securities and investments sector averaged $219, 414. In oil and gas extraction the average was $164, 811. At the other extreme, the average pay of the 9 million workers in food services was $26,294. For the 13.8 million workers in retail it was $35,880; in farming, the average was $36, 893.  

Consider that the average pay for those working in computers and system design was $115,393—that helps explain the government’s problems hiring cybersecurity specialists. And that’s the average pay—if you want the best and brightest, expect to pay more.

The Cato claim contrasts with the argument that General Schedule salaries are 35 percent below the pay of similar jobs. That claim is based on Bureau of Labor Statistics data as reported in Federal Salary Council reports. Those reports have about as much credibility and impact as the Cato reports.

The focus on this issue comes at a good time. If there is to be civil service reform, replacing the General Schedule system will be a hotly debated, core issue. Gaining agreement on a new pay system is central to successful reform and agreeing on the underlying facts—the comparisons with market pay levels—will be essential. Currently, the government does not compile the data to make those comparisons.

In every other sector, the universal practice is to assemble pay survey data for benchmark (or commonly defined) jobs. BLS years ago stopped conducting surveys based on benchmark jobs because job matching is an imprecise process, but those old reports were easily understood and never triggered the controversy that exists today.  

Credible market pay data are readily available. There are surveys covering virtually every occupation and geographic area. It’s been estimated that 2,500 surveys are conducted annually across the country. They will need to be evaluated for quality but it will not be difficult to assemble credible data across the spectrum of job categories and job levels. It might be useful to create a task force of survey experts to assess possible data sources.

And while there is likely to be a fee to use many surveys, the aggregate cost will be a fraction of what BLS spends to conduct its pay surveys.

The data reported by Cato and BLS are useless for this purpose. Neither source is meaningful for determining how much a job is paid in relevant labor markets.

Effective compensation planning starts with agreement on a compensation philosophy and program goals. That should involve both a review of near-term essential skills (i.e., workforce planning, ending with a staffing plan) and an understanding of the best strategies for recruiting, hiring and retaining fully qualified job candidates. All of that is basic to civil service reform.

In well managed companies, this is standard operating procedure.

This is not to ignore the implementation problems. One is that in switching to a new salary system there will be winners and losers. While there will be higher pay for high-demand jobs, it’s also understood that some employees and some jobs are now paid above market rates for their skill set. Second, a new system will undoubtedly be based on pay for performance. A successful transition will necessitate the investment in new performance systems, adequate training for managers and supervisors, plus a solid commitment from agency leaders to make it a priority. Finally, HR offices will need to gear up to support managers and supervisors and commit to addressing the inevitable problems.  The management of performance is not an HR problem, but employees will want continued assurance that its managed fairly.

The need for a new salary system has been discussed for over three decades. One of the early recommendations emerged from the National Commission on the Public Service in 1989. Before that, the 1978 Civil Service Reform Act opened the door to experimenting with new pay models. The National Academy of Public Administration has discussed the problems and made recommendations in at least three reports. The Partnership for Public Service has addressed the problems as well. OPM issued a solid report outlining a new salary system in 2002. Significantly they all recommend essentially the same program model. It’s time to shake the dust off those reports and get started.