Pay for performance haunted by NSPS failure
The imminent demise of the National Security Personnel System leaves many wondering about the prospects for pay for performance in government
- By John Stein Monroe
- Oct 29, 2009
From the day the Federal Civil Service was created in 1872, the government has made it a goal to get the federal workforce to operate more like a private enterprise, with pay and advancement based on merit rather than patronage or spoils.
The General Schedule system, which formalized federal employee pay scales in 1949, was once the solution, but now it is the problem because it puts a premium on longevity rather than performance.
The next logical step — linking pay directly to performance — has been the tantalizing goal for more than 30 years, and in recent years, numerous agencies have conducted experiments with so-called pay-for-performance systems.
But that goal seems more elusive than ever now that Congress is about to jettison the Defense Department’s National Security Personnel System, a three-year-old project that would have put more than 700,000 employees under a pay-for-performance system.
The demise of NSPS has left federal officials and reform advocates grappling with a fundamental question: Did NSPS fail because of poor planning and execution? Or, more worrisome, did it fail because the concept of linking pay to performance — however sensible it sounds — is simply not possible in the federal government?
The answer to the latter question will shape future efforts to update or revamp the government’s antiquated system for managing its workforce and attracting and retaining a new generation of workers, particularly in the highly competitive fields of acquisition and technology.
Conventional wisdom says a properly designed and managed pay-for-performance system will provide added motivation to talented, hard-working employees and, just as important, pressure dead-weight employees to step up or get out.
By all accounts, such reform is needed. In its current form, the GS system is largely viewed as a hindrance to the government’s ability to manage its workforce because of its rigid approach to compensation and career development.
“As much as the unions have been throwing rocks at NSPS, they are not advocates for the General Schedule system,” said Howard Risher, a consultant who specializes in pay and performance. “Everyone seems to understand the damn thing is broken.”
But union officials do not discuss the issue in such black-and-white terms. They agree that the GS system has flaws, but they view it as a viable starting point for any initiative aimed at reforming the pay system.
If anything, NSPS has hardened resistance to any wholesale change.
Federal officials would be wise to steer clear of talking explicitly about pay for performance, said John Palguta, vice president of policy at the Partnership for Public Service, a not-for-profit organization focused on improving government performance.
“That is a toxic term. People associate it with something bad,” he said. Instead, federal leaders should frame the discussion in terms of how to “update the federal approach to compensation so that it is more relevant to today’s environment.”
Where NSPS went wrong
Critics say NSPS failed because DOD leaders developed a system that was unfair and unmanageable, which alienated employees and managers alike.
That perception is borne out by the comments that federal employees have posted on NSPS-related stories on FCW.com.
One of the most common complaints from employees is that their job objectives are too broadly defined to be measured in a meaningful way, which gives managers a lot of subjective latitude when it comes to rating performance.
In other cases, the objectives are too far-fetched, having been handed down from the top levels of the organization with little regard for the day-to-day responsibilities of frontline employees.
The result is that many employees feel as though there is little connection between their performance on the job and the assessments they receive. In such an environment, a negative review can seem like a product of managerial caprice, while a co-worker’s positive review could be chalked up to favoritism.
“The measurement criteria against objectives are like throwing a dart at a dartboard,” wrote a reader named Johnny. “You aim for the highest, hope to get there, but if you don't hit your target, you don't get any points [and] you fail.”
The situation is exacerbated by assessments that are reviewed by members of a pay pool panel who might have little familiarity with the particulars of an employee’s job. In such situations, employees’ ratings and salaries ultimately hinge on the ability of supervisors to write a clear and compelling assessment.
But even that might be a moot point. Many employees say they suspect that their pay pool panels use rating quotas to keep a lid on salary increases.
“NSPS is constrained by funds available to award for performance, just like the demonstration projects of previous years,” one reader observed. “This forces supervisors to fit the actual performance and...grades of their employees into a ‘normal distribution model’ to not give ‘too many’ high marks. Once again, the system puts the supervisor and employee in an unresolvable conflict.”
Any given organization has a fixed amount of money to hand out. In an ideal scenario, an organization should be able to heap rewards on its top performers using money saved by giving smaller increases — or none at all — to employees on the bottom rating rungs.
But what happens if an organization is filled with overachievers? According to many employees, the pay pool panels resort to rating everyone on a curve in those situations, with only a small handful of employees receiving a top or bottom rating and everyone else falling in the middle.
The problem, whether real or perceived, stems from the disconnect between employee assessments and their ratings, said Brenda Farrell, director of defense capabilities and management at the Government Accountability Office.
“Unless NSPS is implemented in a manner that encourages meaningful distinctions in employee ratings in accordance with employees’ performance, there will be unspoken forced distribution of ratings, and employees’ confidence in the system may be undermined,” she wrote in a statement for a June hearing of the Defense Business Board’s NSPS task group.
Fear, uncertainty and doubt
Most observers agree that DOD mismanaged NSPS. But some also say many of NSPS’ weaknesses are likely to show up in any large-scale effort to implement a pay-for-performance system in the federal government. In their eyes, NSPS is not a bad example of pay for performance but a telling one.
Such systems depend on having clear and measurable job objectives. The problem is that the nature of government work often defies simple metrics.
That is especially true with large federal programs in which teamwork, rather than the individual’s contribution, is the principal measure of performance, said Laura Langbein, a professor in the Department of Public Administration and Policy at American University in Washington, D.C.
“When I have an assembly-line job, the link between what I do and what is desired is very clear,” she said. But in a team-oriented environment, which involves endless meetings, “the contribution of my skill to the total output is not clear.”
That reality makes it even more difficult to counter the distrust that comes from having an employee’s salary depend in part on a manager’s subjective judgment.
As long as subjectivity is an element, some employees will find evidence of favoritism and the good ol’ boy network at work, said Darryl Perkinson, national president of the Federal Managers Association. “That is going to exist in whatever system we create.”
Perkinson also questioned the business model for pay for performance. He said DOD officials apparently assumed that they had a lot of underperforming employees whose salary increases could be commandeered to raise the pay of others.
But that did not turn out to be the case. In 2008, for example, fewer than 2 percent of DOD employees received fair (Level 2) or unacceptable (Level 1) ratings, according to data posted on the Civilian Personnel Management Service’s Web site.
Perkinson said future reform efforts cannot be based on similarly mistaken assumptions. “The worst thing we could do is put a pay-for-performance system in place that somehow gets capped or limited by budget constraints.”
Where to go from here
Risher, who has been involved with pay reform efforts since 1990, rejects the notion that NSPS should scare people away from pay for performance.
NSPS was unprecedented in its scale, but it was hardly the first such system in government. In fact, dozens of organizations have conducted projects to test new approaches to pay and personnel management, as authorized by the Civil Service Reform Act of 1978.
Many of the projects have tested the concept of pay banding, in which the GS system’s 15 pay grades are collapsed into a handful of pay bands, each of which covers a specific salary range and accommodates various career groups. An employee does not advance by steps based on longevity but instead receives performance-based raises.
Risher and other experts often point to the success of that approach in demonstration projects at the Naval Weapons Center in China Lake, Calif., and at the National Institute of Standards and Technology. Both projects are now permanent.
The fate of NSPS notwithstanding, “we do know how to do this,” Risher said.
Indeed, DOD might have benefited from the experience of agencies whose projects did not go well. For better or worse, “there are a lot of lessons to be learned,” Risher said.
Other experts are more skeptical about the prospects for developing a pay-for-performance system that could work at DOD or governmentwide.
Pay banding has worked best in relatively small organizations, such as China Lake and NIST, said James Thompson, an associate professor in the Graduate Program in Public Administration at the University of Illinois at Chicago.
At those places, employees “almost know the top management personally, and the top managers can convey to the employees, ‘Here is what this is about. Here is what the implications are. And here is how you can appeal,’ ” Thompson said.
Union officials are wary of any reform initiative that involves building a pay system from scratch, as DOD tried with NSPS. They say the best strategy is to work within the framework of the GS system.
Indeed, they argue that the system offers numerous flexibilities that many federal managers never take advantage of, such as quality step increases, which make it possible to give outstanding employees a bigger-than-usual jump within a pay grade. Other options include recruitment and retention bonuses and paid-time-off awards.
“A lot of the things that have been put into the General Schedule system have never been implemented,” said Randy Erwin, legislative director at the National Federation of Federal Employees.
Palguta agreed that the GS system is the most likely starting point for any reform initiative. But that doesn’t mean the federal government should give up on much-needed improvements.
“The way going forward is not to simply go back to 1949 but to take everything we have learned from what did not work in NSPS, what has worked with other performance pay systems and what does not work for the General Schedule system and let’s try to build a compensation approach that is better than what we have,” Palguta said.
The Obama administration will likely be unable to garner support for converting the GS system wholesale into a pay-for-performance system. But officials should look for ways to “make it more performance sensitive,” Palguta said.
One option is to incorporate pay bands into the system to offer managers some discretionary authority for rewarding top performers or attracting new talent while still providing annual increases for other employees based on longevity.
That approach would be more flexible than the GS system but less risky than NSPS. Such a compromise has a sound precedent, Palguta said.
“The history of federal [human resources] systems is [about] evolution rather than revolution,” he said.