March 10, 2011A reporter for a federal newspaper suggested that Wednesday's hearing titled "Are Federal Workers Underpaid?" was "as close to Thunderdome as a federal pay debate can get." This was only the first round, but it is clear which direction the battle is moving in. The real fight, not surprisingly, was among the members of Congress.
After watching the hearing and reading the written testimonies to the House Oversight and Government Reform Subcommittee on Federal Workforce, U.S. Postal Service and Labor Policy, I was struck by the fact that the individuals who testified took very different routes but reached the same basic conclusion -- the General Schedule salary system is broken and needs to be modified, or replaced with one that is more responsive to labor market trends. After employees have been told for years that they are badly underpaid, it actually might be good news to learn there is agreement that they should be paid competitive market salaries.
From a broader perspective, the debate appears consistent with the public's view that government employees should be paid in line with their private sector counterparts. The new wrinkle is the news media have highlighted the reality that government benefits often are better than the typical package provided by other employers. In their written testimony, the analysts from the American Enterprise Institute and Heritage Foundation focused on that differential. Recent media reports show public support for reducing benefits, but those reports also suggest the public believes "total compensation"-- cash remuneration plus benefits -- for government employees should be in line with market levels.
Rep. Dennis Ross, R-Fla., chairman of the subcommittee, in an interview referred to the hearing as "the first step" toward his goal of establishing a pay-for-performance system for federal employees. The idea of a market-sensitive salary system combined with performance pay is the universal model for white-collar workers in the private sector.
First, Assemble the Market Pay Facts
Another point of agreement is that one of the first steps toward a solution is determining market pay levels. Both sides have invested hours in analyzing survey data. Both can argue their analyses are based on well-established methods. But neither can say which federal jobs are overpaid or underpaid.
There is, however, a key divergence in their conclusions. AEI and Heritage repeated their argument that federal employees are overpaid. Perhaps the only "new" news was that Office of Personnel Management Director John Berry did not reiterate the claim reported in the President's Pay Agent that a significant pay gap exists. He stated, "The wages that the federal government pays its employees are fair and the benefits it offers are competitive." He did not, however, support his point with data.
Berry mentioned two reasons that could explain why he backed away from the Pay Agent gap conclusion. First, he stated the relevant market comparison is "with other large companies for whom we compete for talent." That is a very different market pricing strategy than that dictated by Bureau of Labor Statistics surveys. Second, he acknowledged, "The underlying model and methodology for estimating pay gaps should be re-examined." Significantly, no one defended the model, or the methodology.
Actually, to borrow from the testimony of Max Stier, president of the Partnership for Public Service, the most reasonable conclusion is that of John O'Leary, a fellow at the Ash Center at the Harvard Kennedy School, who cited "the Goldilocks truth": Among any group of public employees, some are underpaid, others overpaid and still others are paid just about right. The problem is the data to sort jobs into those groups have to be collected and compiled.
OPM vs. the Think Tanks
Both sides in Wednesday's debate claimed their conclusions are definitive. There are, however, serious questions about their analyses. Both rely on databases -- OPM has used the BLS' National Compensation Survey, and AEI and Heritage use Census Bureau data -- that were not developed for this purpose. Moreover, no other employer in the country would consider using these databases, or similar analyses.
OPM is in the worst position because it has been forced to "make do" with a database and survey that were developed to support the Employment Cost Index. The ECI is a time series that focuses on the percentage increase in pay. For that purpose, salary levels essentially are irrelevant.
The use of the BLS data triggers a long list of concerns. Both sides rely on statistical methods that defy public scrutiny. Neither side has invited independent analysts to audit their claims. Neither is able to take the next step and develop data showing how much jobs are paid in the labor market. The textbooks make that the first step in salary planning.
There Will Be Winners and Losers
That should not be surprising; it's already started -- though the winners are undetermined at this point. Everyone is watching what's happening at the state and local levels. Recent budget cuts and layoffs are just beginning. At the federal level, the impact of the deficit is still to be decided. The Q&A following the statements on Wednesday made it apparent that House Republicans want new justification for federal salaries.
An important aspect of the problem is the GS salary ranges have never been aligned with market rates. They were not aligned when the GS was enacted in 1949, and are not aligned today. The ranges have been adjusted annually to reflect broad market increases, but a job-by-job market analysis is likely to confirm O'Leary's Goldilocks truth.
Federal pay for some high-demand knowledge jobs is $30,000 or more below the market. Some government jobs are filled by the world's best experts whose credentials could command far higher salaries in the private sector. It is not in government's interest to perpetuate that gap -- or allow it to get worse as the economy recovers.
At the same time, some jobs and individuals might well be overpaid. Both AEI and Heritage discussed some of the reasons. Everyone who knows the GS system knows their criticism was accurate. It's a mixed bag, however, and that's what complicates things. Actually, using the AEI-Heritage logic, a young, outstanding performer would be overpaid. Bill Gates, a college dropout, would have been overpaid by their lights.
The GS program has not been well-managed. That's in part because its design requires unjustifiable staff time. It's also true that agency leaders have not had a reason to make it a priority. And frankly, the reality is many lower-level human resources specialists are not ready to lead the systematic assessment that is needed at every federal work site. Critics such as those at AEI and Heritage are not going to be satisfied, however, until the problems are addressed.
Government Will Be the Winner
The argument for modifying or replacing the GS system is hardly new. It was discussed in the meetings that preceded the Federal Employee Pay Comparability Act, and repeatedly during the subsequent 20 years in reports by the National Academy of Public Administration and by a couple of commissions led by former Federal Reserve Chairman Paul Volcker. OPM issued a solid report in 2002, "A Fresh Start for Federal Pay: The Case for Modernization," that provides an excellent discussion of the problems with the GS system.
The thread running through each of these reports is the need for a pay system that is more flexible, that is market-sensitive and that links increases to performance. Wednesday's hearing confirms that Rep. Ross wants to move in that direction. It's too early to outline the parameters of a new system, but the basic program model was discussed in those reports and has been proven in a variety of private and public organizations.
When the GS system finally is modified or replaced, the winner will be government. A new system will make better use of payroll dollars, respond quickly and support the reorganization of government, and make it more competitive in attracting top talent. Despite the problems with the National Security Personnel System, there is anecdotal evidence it was changing the culture and supporting improved performance before it was discontinued.
Critics often argue that performance pay is unfair and possibly discriminatory. But those concerns can be addressed and the problems minimized through program management. Critics also argue that performance pay does not contribute to improving performance. But that argument refutes the experience of millions of employers across the country and, increasingly, the world. It is also refutes all that we know about the psychology of recognition and rewards. Leaders in the private sector would not be willing to live with a step-based salary system.
Neither the Defense Civilian Intelligence Personnel System/National Intelligence Civilian Compensation Program nor NSPS was truly a failure. Both NAPA in its assessment of DCIPS and the Defense Business Board in its review of NSPS recommended that the pay systems be continued. The reviews concluded that changes were needed, but neither recommended program termination. Mistakes were made, but the problems could have been resolved. With better planning, problems could have been avoided. NAPA's review of DCIPS highlights a number of "lessons learned" that could be useful if there is support to move forward with the Ross' plans.
The problem now is finding a platform on which the parties can work together to develop the market data and eventually rethink the salary system. In 1989-90, Connie Newman, then director of OPM, created a broadly representative committee to discuss the issues. In my role as managing consultant to her, I had a committee of five experts who had authored textbooks on salary management. Newman also asked me to secure input from stakeholders across government. Such a strategy could be successful again. The freeze gives government almost two years to plan a new salary system.
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 is the author or co-author of five books, including Planning Wage and Salary Programs (WorldatWork Press, 2009).
March 10, 2011