The Heat Is Turned Up on Federal Pay

The author offers some analysis on the debate over whether federal employees are paid too much or too little and says that neither side is going to concede to the other. He says that salary surveys are a waste and must be done differently for there to be any real conclusion on the debate.

No doubt one of the busiest websites late last week was FedsDataCenter.com, the source of everything anyone could ask about how much federal employees are paid.  The load was so heavy last year that upgrades were done to the servers to avoid a crash this year.

All salaries are relative – they cannot be evaluated until they are compared with others in similar jobs.  All those ‘high pay’ VA physicians are actually paid less than their counterparts in other large hospitals.  The SEC has a long list of lawyers all paid the same for 2013 — $230,700 according to the website — which sounds high until their salaries are compared with their counterparts on Wall Street.

Yes that’s more than the salaries of members of Congress and the same as the Vice President’s salary.   Are SEC lawyers overpaid?  Not if the agency wants to keep lawyers capable of dealing with the complex problems in the nation’s financial sector.

Somehow I’m reminded of the old story about Babe Ruth who was asked why he was paid more than President Hoover.  His response supposedly was, “What the hell has Hoover got to do with it?  Besides, I had a better year than he did.”

Years ago in the Reagan administration proposals were made to privatize NIH, all because of the high salaries of those world class physicians.  But reason prevailed possibly because the proposal would have actually cost more!  The physicians would have moved to medical schools where they would be paid higher salaries, conduct the same research, hire a support staff, purchase costly equipment, and in some cases build new labs – and it would all be funded with NIH grants.  Were they overpaid at NIH?  Of course not; they were underpaid compared to the compensation of their peers.

Today’s reality is no different.  The reluctance to pay competitive salaries means agencies fall back on costly contract personnel.  Rumor has it that the same people who resigned recently come back on a contractor’s payroll to perform the same job.

Would SEC be able to find lawyers willing to work for a lower salary?  As critics are quick to point out, the typical lawyer works for a lot less.  But then the Wall Street firms are not typical.  The major firms start law school graduates at $160,000.  SEC is competing for talent with those firms.

Are federal employees paid too well? 

The fact is no one can point out the jobs – if any — that are paid above competitive levels.  The standard practice would be to define the needed job skills and identify other employers competing for the same talent.  Then surveys are used to determine competitive pay levels, job by job.  It’s a simple logic.

That basic approach is used to adjust Federal Wage System (FWS) pay levels.  It’s also used to maintain competitive salaries for VA nurses.  They use survey data to track the pay of nurses in local hospitals.  The FIRREA agencies use surveys of employers in the financial industry.  The Federal Reserve Board does as well.  It’s universal outside of government.

For reasons that are buried in history the Bureau of Labor Statistics (BLS) has been hung up on maintaining the statistical validity of the survey used to adjust the General Schedule (GS) salary ranges.  Now with budget cuts BLS has been scrambling to patch together data from a couple of surveys.

The key to salary planning is gaining agreement on the survey data used to assess current pay rates.  It’s not a science.  Decision makers need to accept the job by job analyses as credible.  Surveys report simple descriptive statistics — means, percentiles etc. – not multivariate black boxes.  A disturbingly high percentage of the BLS data is statistical estimates, not actual pay for actual workers.  For several years BLS tried to market its data at HR conferences but today no employer, except the Pay Agent, uses BLS data as its primary survey source.

The problem is compounded by the cost of the BLS surveys.  The FY 2014 budget to compile and report “a diverse set of measures” related to “compensation and working conditions” is $81.5 million.  In fairness that covers several surveys.

There are hundreds of salary surveys conducted each year.  The cost to acquire the complete inventory and compile summary data would not exceed $250,000.  The primary survey for the DC area, for example, costs roughly $1,200 (the price is lower for participating organizations) and includes a significantly larger database than that compiled by BLS. There are similar surveys in virtually every city.

More importantly, however, the BLS data are simply wrong for this purpose.  The recommendations submitted by the Pay Agent to the President have been rejected annually for two decades.  It’s clear the reported conclusions are not credible.

I doubt if even the union members of the Federal Salary Council believe employees on average are paid 34.6% below market rates.

At the same time, the studies showing federal employees are overpaid are not convincing for the same reasons.  They are also statistical black boxes.  The problem is analogous to using Consumer Price Index data and a macro statistical analysis to predict the price of a jar of mustard.

Government Needs to Move Forward

The stalemate has already gone on too long.  Neither side is going to concede it’s wrong.  The law requiring use of BLS data will at some point have to change.  The current BLS methodology will never be accepted.  Government and employees are the losers as long as salaries are not competitive.

The recommendation is to develop an independent database of market pay data that is accepted by all parties.  If necessary, the country’s experts could be asked for advice.

About the Author

Howard Risher is a private consultant who focuses on pay and performance. His career extends over 40 years and includes years managing consulting practices for two national firms. He recently became the editor of the journal Compensation and Benefits Review. He has written four books, including Aligning Pay and Results. He has an MBA and Ph.D from the Wharton School of the University of Pennsylvania.