Defense to probe fairness of pay for performance system

The Defense Department is reviewing pay data for employees under its new performance-based pay system to determine whether decisions were fair and credible.

Defense is examining 2007 data from the National Security Personnel System to see if there were significant pay disparities based on an employee's age, race or gender, said Brad Bunn, NSPS program executive officer, in a recent interview.

"Certain elements of the raw data are out there now," Bunn said. "Like any system, you have to look and see what the results were and how it operated and analyze those outcomes."

Government Executive obtained data on employees' performance ratings, pay increases, bonuses and total payouts in July through a Freedom of Information Act request. Defense identified employees by their age range, gender, race, ethnicity, location and component agency.

Bunn said it's still too early in the review process to draw any specific conclusions about disparities in the pay system, but noted that some inconsistencies have generated concern. As a result, he said, the program office has launched its own analysis of the information and asked individual Defense components to examine their internal evaluation and pay pool processes to identify any discrepancies.

"We've given each organization some guidance along the way on what they ought to be focused on, including looking at the rating results across different demographic categories," Bunn said. "They will make adjustments, address any issues and put more focus on training their managers. They also will look at how their pay pools are constructed and how they diversify the pay pool panel."

Defense has added more than 181,500 nonbargaining unit employees to NSPS since 2006 and conducted two performance cycles for workers in the system. The first of those cycles affected only the original 11,000 employees to convert to the system. The second cycle marked the first large-scale run of NSPS, affecting 102,000 employees.

Bunn said some disparities could be a result of the fact that NSPS is still in the implementation phase, meaning pay data does not always reflect that some components enter the system at different times during the year.

For example, he said, lower payouts identified at the Navy largely stem from a midyear transfer of more than 7,000 employees from the Space and Naval Warfare Systems Command into NSPS. Those employees received a portion of their payout midyear under their previous personnel system and the other portion of the payout under NSPS in January, he said, making NSPS increases account for only about half of their total payout.

Richard Brown, president of the National Federation of Federal Employees, said Thursday that while the union might reserve judgment for a couple of years until trends are clearer, it might be a high price to pay to "wait and see if the system is unfair to different people."

"The agency is going to try to make light of these figures, saying one year of data does not indicate a trend," Brown said. "We should not let them off the hook that easily. This is a big problem, and the agency should recognize that."

Brown and Matt Biggs, legislative director for the International Federation of Professional and Technical Engineers, said the data already confirms what many labor unions have said about pay for performance. "All you have to do is look at a couple of these pages, and if you're white and male, you're doing well," Biggs said. "It's not surprising; we've seen these pay-for-performance systems in the federal government, and they are discriminatory against people of color and women."

Meanwhile, Bunn said the program office and the Government Accountability Office also are examining the effectiveness of current safeguards built into NSPS to ensure they provide appropriate recourse for employees.

One of the most important safeguards, Bunn said, is a reconsideration process in which employees can challenge their ratings after the pay pool process. For the 2007 payout, about 2,000 employees challenged their ratings; of those, one-third had their ratings overturned, he said.

"This should say to employees that the process is effective," Bunn said. "That indicates the credibility of the process."

But Biggs said many employees expressed concern that while their ratings might be increased through the reconsideration process, they might not be entitled to a higher pay raise as a result. Since the pay pools already have divvied up the amounts available for pay raises, he said, there's no funding left for employees who are entitled to a higher payout after they've appealed and had their rating increased.

But Bunn said a rating change does guarantee an employee a higher payout. "If there are changes to ratings in the reconsideration process, organizations will pay out the higher amount regardless of the fact that the pay pool money has been spent," he said.

Biggs urged a return to the General Schedule, saying the system is fair and credible and gives agencies an opportunity to reward employees. "We don't need pay for performance," he said. "If it ain't broke, don't fix it."

Bunn pledged to examine the data further and communicate to managers and employees that standards and benchmarks for rating employees should remain objective. Still, he acknowledged, no pay system that involves human judgment could ever be purely objective.

"What we've attempted to do with NSPS is essentially make the proportion of objectivity greater than the subjectivity by emphasizing good, solid job objectives that are measurable and observable so employees can decide what it takes to get a 3, 4 or 5 [rating]," Bunn said. "But obviously, someone has to apply professional judgment."