Federal personnel chief John Berry declared in June that the government's official estimate that federal employees are paid 22 percent less on average than their private-sector counterparts lacks credibility. He was right and he promised to bring in outside experts to study the way the government calculates the pay gap.

But he didn't. Instead he has backed away from his own criticism and has done nothing. In the meantime, a steady stream of reports surfaced, arguing that feds enjoy a whopping pay and benefits advantage over the private sector. Federal advocates counter that those studies are flawed because they lack an apples-to-apples comparison of people doing similar work with similar levels of experience and education.

Still, those studies — including some by Federal Times' sister Gannett publications USA Today and the Asbury Park Press — have helped fan the issue of federal pay into a full-blown political debate. In response, Berry again is promising to study the matter, only this time he says a study won't be done by outside experts, but rather by the same senior administration officials who decide the pay gap.

This is regrettable and dodges an objective look at this important matter.

Berry's back-tracking on this matter will only make federal pay a bigger political football than it already is. And that will fuel even more resentment toward federal employees by Americans than already exists.

A chief reason federal pay is an issue at all is that the process for calculating the gap between government and private-sector compensation — which is used to determine locality pay raises each year — is exceedingly complex and done by senior administration officials who lack independence on the matter. These officials are advised by a panel of labor leaders who have a vested interest in the outcome.

pay-gap results will never be accepted widely unless they are produced by independent experts with no stake in the outcome. And those results need to be based upon methodology that is clear and transparent, rather than on convoluted statistical modeling, as they are now.

There is little excuse in this age, when wage survey data is so abundant, to not use real pay data to determine pay gaps in granular detail: by occupation, by locality, by experience and education. A good analogy is per diem rates. Several years ago, the General Services Administration successfully transitioned from using statistical models to actual survey data in its annual calculation of per diem rates, refining them to be far more appropriate and responsive to the precise needs of federal travelers.

Finally, pay gap calculations today lack timeliness — the pay gap being determined this year will determine federal pay raises two years from now. it is time to conduct an honest — and outside — assessment of the federal pay gap. Anything less will be viewed with understandable suspicion and will be unfair to the public and the federal work force.