It’s Hard to Overstate How Awful the Government’s Pay System Is

It’s that time again. The Federal Salary Council will soon announce its estimate of the salary gap between government employees and their private sector counterparts. Last December, the council reported salaries paid under the General Schedule system were 35 percent lower than private sector levels. But apart from the unions, few people in Washington truly believe the gap is that large.

Outside of government, labor markets are tightening and that’s pushing up pay. For 2016, surveys showed employers budgeted 3 percent for salary increases. Federal employees were granted a 1.3 percent pay boost, including locality adjustments. For 2017, surveys show employers are budgeting 3.1 percent for increases while the president has announced plans to increase federal pay 1.6 percent. (Although it’s rarely mentioned, the step increases should be added to that.) In the 26 years since the Federal Employees Pay Comparability Act was enacted, no president has approved increases that would begin closing the reported gap.

As a point of reference, when the increase budgets reported in those surveys (conducted annually by the professional group WorldatWork) are compounded over the 16 years starting in 2000, the total is 69 percent. For comparison, over the same period the GS ranges in Washington DC were increased 46 percent. That’s not truly apples-to-apples; the step increases should be added and neither number includes promotional dollars. But it highlights what everyone sees in the headlines.

A core issue is that government data does not tell how well employees are paid relative to prevailing market rates. It’s widely believed that some employees are paid above market – private sector surveys confirm that – while the proliferation of “special rates” confirms many jobs are paid below market.  With other employers, top management would want assurance that salaries are not an impediment to recruiting needed talent.

To highlight the obvious, GS salaries are far below market for critically important cybersecurity jobs. But relying on special rates to draw needed talent is not the answer. To attract and retain these specialists and others in the STEM occupations (Science, Technology, Engineering and Mathematics), the only long term solution is a separate pay system to fit these career ladders. 

A related point is that the survey data compiled by the Bureau of Labor Statistics does not provide the information to determine if jobs are overpaid or underpaid. The survey methodology was not planned to develop job-specific market data. The BLS at one time tried to induce other employers to rely on the same data but their “marketing” at conferences was unsuccessful. The methodology is staggeringly complex; it’s a classic black box dependent on multivariate statistics. It’s highly doubtful that anyone involved in administering the GS system would be able to explain or defend it in detail.

By comparison, the local HR association, Human Resource Association of the National Capital Area, conducts a far larger and widely used survey in the metro area. Its annual survey provides data for over 500 benchmark (or common) jobs.  In contrast to the millions spent on BLS surveys, the local survey costs participating employers less than $1,000.  There are over 1,000 similar surveys conducted every year across the country. 

The problems with the GS system were acknowledged as long ago as 1978 when the Civil Service Reform Act allowed agencies to adopt demonstration pay systems. The banking agencies were authorized to adopt their own pay systems in 1989.  The 1990 Federal Employees Pay Comparability Act was expected to provide the flexibility to make the GS system more competitive but the needed funding has never been approved – and there is no reason to think that’s going to change.

A fallback strategy to get a pay increase is getting a job reclassified to a higher grade. Grade inflation has been a recognized but largely ignored problem for years. No one can say how prevalent it is and HR offices no longer have the specialists to “police” job classification. It undermines program integrity.

Realistically, the classification system is no longer viable. When it was conceived, jobs and organizations changed less frequently, and a large percentage of today’s knowledge jobs did not exist. Now with the rapid changes in job knowledge and technology, classification standards should evolve just as rapidly, which is to say continuously. In the private sector, formal approaches to classification (or job evaluation) have been discarded as costly and bureaucratic.

The remaining issue is the continued reliance on step increases. Pay for performance is effectively universal for managers and professionals in other sectors. There are also state and local agencies where it’s been solidly successful for years. It’s a key to retaining Millennials. Despite the prominent failures in the Bush administration, there are federal success stories going back over 30 years. However, it will require a commitment to more effective performance management.

A final point is the misleading question in the Office of Personnel Management’s Federal Employee Viewpoint Survey: “Considering everything, how satisfied are you with your pay?” Feeling “satisfied” is not the same as “satisfaction” as suggested in the discussion of the survey. Research 50 years ago showed pay is not a source of satisfaction. If anything, the high performers are most likely to be dissatisfied. Employees may be satisfied with their salary but no aspect of the GS system can possibly trigger satisfaction.

The best that can be said about the General Schedule (GS) salary system – possibly the only positive thing – is that employees get a regular paycheck. It’s an impediment to talent management.