By Howard Risher
December 3, 2010Two years? Three years? Complete freeze? Partial freeze? It seems certain that pay for federal employees will be frozen in some form.
It also appears that their retirement benefits will be reduced. The final recommendations of President Obama's fiscal commission released on Dec. 1 would bring federal benefit plans more in line with common private sector practices. The savings clearly are important, but the impact on government operations in 2011 could be far greater than analysts expect.
Almost 500,000 federal employees were eligible to retire in 2009. If the commission's recommendations are adopted for 2012, that could be an incentive for many to retire next year. A switch to the high five will reduce vested benefits as well, so that many more could opt to leave as soon as they can.
The freeze also will inhibit federal hiring and trigger attrition. Resignations among young new hires, already a concern, are likely to increase as the economy improves and private sector alternatives rise. Perhaps more important is the image of government as a career path will be damaged. Indeed, the federal pay debate already has denigrated the value of federal workers.
There are questions about the comparisons of federal and nonfederal jobs and salaries. It is indisputable that federal salaries for new college graduates already are low, however. The national average this year is $48,351 as compared with the General Schedule Grade 7 pay of $42,209 in Washington. For high-tech grads -- engineers and computer specialists -- the averages are around $60,000. That gap risks damaging future government operations -- and it's going to widen with the freeze.
Time to Assess Facts
Because a freeze will not alter the comparison between federal and nonfederal salaries, the debate will continue. The Office of Personnel Management claims federal employees are badly underpaid, while the Cato Institute and the Heritage Foundation say they are overpaid. So far, both sides have based their arguments on arcane statistical analyses.
The debators on both sides bear a striking resemblance to the ministers in the tale The Emperor's New Clothes. Up to this point, neither side has generated information showing how much federal employees should be paid.
The duration of the two-year freeze coincides with the run-up to the 2012 election. The period would be an ideal time to develop credible market pay data and give leaders the opportunity to evaluate the facts. It would be advantageous to agree on the facts so that federal pay does not become a contentious issue in the election campaign.
Neither side currently relies on a data source, or on analytical methods suited to salary planning. The analyses each used are based on government data that was collected for other purposes. Cato relied on Bureau of Economic Analysis data while Heritage used the Current Population Survey database. OPM and the Federal Salary Council rely on Bureau of Labor Statistics data. So regardless of how the data are analyzed, results are not comparable.
Actually, no employer determining salaries uses a methodology remotely similar to the Salary Council's or that used by either Cato or Heritage. Other employers would not consider using those data sources.
The universal approach used by corporations, hospitals, universities, and even many mom-and-pop businesses is to obtain information on what other employers are paying in the relevant labor market(s) for similar jobs. Their logic is simple.
There are an estimated 2,000-plus surveys conducted annually that report pay for benchmark jobs. There are also a growing number of Internet data sources. Data are available showing pay levels for a long list of occupations. The surveys report salary levels, job by job, using simple, descriptive statistics -- means, medians and percentiles. Data from several surveys are commonly combined to determine market pay levels, but anyone opening a survey report would be able to interpret them.
To be sure, salary surveys can be poorly executed and misused. Some are garbage in, garbage out. But there are good ones. In the Washington area, for example, the survey a local human resources association conducted has a broad cross-section of 340 local employers and data for more than 80,000 job incumbents -- a database far larger database than the BLS National Compensation Survey. OPM some years ago matched federal jobs to the benchmarks in the local survey. There are similar surveys done around the country.
A few weeks ago The Washington Post reported a poll showing a solid 75 percent of respondents agreeing that "federal agencies should offer pay and benefits that are about the same as those provided in comparable private sector jobs." That could be the basis for a policy that would be broadly accepted.
The Problem Is the General Schedule
All the parties are on record agreeing that the General Schedule should be replaced with one more market-sensitive. But market concerns are only the tip of the iceberg. The roots of the job classification system can be traced back a century. It is difficult to imagine any practice more bureaucratic and less conducive to smart utilization of employee capabilities.
Step increases are another concern. Pay for performance is effectively universal for white-collar jobs in virtually every other sector. In the academic world, even the critics of performance pay get rewarded in some way for their performance.
OPM in 2002 published a white paper that remains the most extensive discussion of the reasons a new salary system is needed. OPM Director John Berry voiced basic agreement with the white paper in a speech late last year. Heritage analyst James Sherk said, "Uncle Sam should move to performance-based pay based on market rates -- the same system most private employers use." The same recommendation is made in the OPM white paper, at least three reports by the National Academy of Public Administration, and two of the many reports generated over the years under the leadership of former Federal Reserve chairman Paul Volcker.
A few weeks ago the Partnership for Public Service published a report, Closing the Gap: Seven Obstacles to a First-Class Federal Workforce, summarizing the views of agency chief human capital officers. The officers were asked to say what it would take to bring about change, further strengthen the civil service and improve government operations. Most replies focused on hiring reform, but the second-most cited obstacle was "an antiquated and overly rigid pay and classification system."
Proponents of pay for performance were dealt a setback by the Defense Department's experience with the National Security Personnel System and the Defense Civilian Intelligence Personnel System, along with the subsequent decision to revert to the GS system. The trend to move away from rigid, tenure-based systems continues, however, both here and in other countries.
Significantly, the teams that evaluated NSPS and DCIPS found problems, but concluded both systems could and should be salvaged. There is a long list of successful alternative demonstration projects and Title 5-exempt agencies with separate salary systems.
This winter would be an ideal time to assess that experience to understand what worked and what didn't. The evaluation of DCIPS that NAPA conducted is a good prototype. Federal leaders should use it to avoid mistakes and to build on proven practices.
Their conclusions then should be used to plan a replacement for the GS system. The basic program model -- a framework based on salary bands -- has been used in some places for 30 years. The lessons learned from that experience will facilitate the drafting of new policies and procedures to manage salaries.
Such a project is very doable during the next two years. The GS system must be replaced. Agencies need a reward system that supports mission accomplishment. The country would not be well-served by allowing this debate to continue into 2013.
Howard Risher is an independent compensation and performance management consultant. He was the managing consultant for the studies leading to the 1990 Federal Employees Pay Comparability Act. He has a doctorate in labor and economics from the Wharton School at the University of Pennsylvania.
By Howard Risher
December 3, 2010