I have written three posts on replacing the General Schedule. The first outlined some fallacies about GS, the second addressed facts, and the third outlined my reasons for opposing efforts to implement pay for performance as the primary means of setting Federal employee pay. What we know about the GS now is that it was designed for a workforce that we no longer have. When the GS was created by the Classification Act of 1949, it was viewed as a leap forward. It consolidated numerous pay systems, created the GS-16, 17 and 18 “supergrades” that were the predecessors of the Senior Executive Service, and decentralized much of the routine pay-setting authority of the Civil Service Commission (now OPM). The Classification Act grew out of the recommendations of the Commission on Organization of the Executive Branch of the Government – commonly known as the Hoover Commission. The commission was remarkably prescient, making recommendations that are still relevant today. Its recommendations included category rating, simplified and more effective performance ratings, and selection processes for supervisory jobs that focused more on ability to be a supervisor than on technical experience. The Hoover Commission also recommended pay include locality or industry differentials. The Classification Act did not include all of the Hoover Commission’s recommendations, but it made much progress.
Since the the Classification Act was passed, the state of the federal workforce has changed dramatically. It has shifted from a balance where more than half of employees were in the bottom 5 grades to one where more than half are in the top 4 grades. The complexity of Federal jobs has increased, lower-graded jobs were automated or outsourced out of existence, and the labor market has moved from one with more trade and craft jobs to one where knowledge work is the norm. The GS has not kept pace, agencies raised grade levels in an attempt to remain competitive in the talent market, and it has now gotten to the point where tweaking the General Schedule is not enough – it needs to be replaced.
Characteristics of an Effective Classification and Pay System
Any measure of effectiveness of the civil service system is dependent on how we define the goals of that system. Outlining those goals is more complex than it might seem. Most people (but not all) would agree Federal employees deserve fair pay for their work. Some people argue that Federal pay should not let government compete with the private sector, while others would argue the Federal government should be a model employer. Some argue the Federal government, even at its worst, is a remarkably stable employer and that stability of employment should be a factor in compensation decisions (i.e., pay should be lower). Almost everyone would agree that selections should be based on merit, but merit, like beauty, is in the eye of the beholder. Some would argue we should pay all Federal employees in a given grade at the same rate. Others argue it should be based on pay in the labor market where the employees are hired. Some argue Federal pay should rise when labor markets demand it, but they should not decrease when the opposite is true. Some would argue that pay should be subject to collective bargaining, others argue there is no point in that because the size of the pie doesn’t change – bargaining would just rearrange the slices. As always, the devil is in the details. In order to lay out a set of reasonable recommendations, I have to start with what I believe are the classification and pay characteristics we should be striving for.
- Less Complexity. The GS has become more complex in the past 65 years, partly as a result of the growing complexity of many Federal jobs, partly as a result of previous attempts at reform, and partly because of the natural tendency of any bureaucracy to grow processes, rules and checkpoints. For example, the GS Handbook of Occupational Groups and Families today has 407 job series in 22 Occupational Groups. Add to that the classification guidance for supervisory positions, the various parenthetical designations in many position (11 in the Information Technology series alone), and you have the potential for more than a thousand ways to classify a white collar job. Add individual grade levels (8 for the typical 2-grade interval series), and the number possible job classifications becomes many thousands. That fact presents an illusion of precision that simply does not exist. The reality is that one characteristic of seasoned position classifiers is their ability to finesse the classification process to get the type of classification the organization needs or wants. One experienced classifier used to tell me she could “classify a Dixie Cup as a GS-15.” False precision comes at a cost. More than 150 series have individual position classification standards. Many others have classification standards that cover multiple series, while others have “flysheets” that provide basic classification guidance. The complexity of the multitude of job classifications places a tremendous burden on the Office of Personnel Management, whose small group of standards writers have the herculean task of keeping them current, and agencies that have to administer the system. It places a burden on job applicants who are bewildered by the multitude of job classifications in the government. And it puts a tremendous burden on agencies to manage the entire process. Faced with a highly complex system and pressure to reduce overhead jobs, many agencies have turned to automated classification systems and reduced the number of highly experienced classifiers. The scale and breadth of Federal responsibilities means it will never be a simple system, but it can certainly be far less complex than it is now.
- Fairness. This is a tough one, because everyone seems to have their own definition of fairness. Here is mine – pay should be based upon the work we do. Age, gender, race, ethnicity, sexual orientation, religion and other non-merit factors should have no place in classification and pay decisions, nor should discrimination or unintended systemic adverse effects on people for those reasons be allowed to exist.
- Public service as a public trust. Federal employees work for the people of the United States – that makes the government unlike any other employer. As Thomas Jefferson stated in the Declaration of Independence, governments derive “their just powers from the consent of the governed.” That means Federal employment is a public trust. Public servants do not go to work to make their fortunes. Government is not and never will be a place to get rich. Trust, however, is a two-way street. When we the people ask Federal employees to do the people’s work, we must treat them as public servants rather than as undesirable bureaucrats.
- Based upon sound management principles. One factor that makes civil service reform hard to do is our highly political environment. From the beginning of the republic, decisions regarding government and its workers have been based as much on politics as on merit and sound management. Many of the recommendations of the Hoover Commission died in the political process. Whatever we do now in the name of reform, we should as much as possible base it on effective management rather than politics.
- Reward that which we want to encourage. Recent efforts at pay for performance have attempted to address one thing we want to encourage – performance. They have often failed to encourage retention as well. Rather than treating longevity-based pay increases as evil, we should recognize it is in the interest of good government to have a stable workforce that does not suffer from unnecessary turnover. The costs associated with turnover are significant. Pay for performance efforts also generally fail to recognize and reward teamwork. If we want to encourage the kind of teamwork that is necessary to deliver consistently excellent results, we have to encourage, recognize and incentivize it.
- Where discretion on pay matters exists, it should be executed by people who are trained and equipped to do it. One problem with pay for performance systems is that they often place pay decisions in the hands of managers who lack the expertise to make informed pay decisions. We often hear that managers struggle because they were selected for their technical skill rather than their leadership abilities. We also hear that agencies do not invest enough time and money in training supervisors and helping them make the transition from doers to leaders. Why do we assume these same managers now have the ability to successfully execute a pay system when they have no training in compensation, do not have access to data on pay comparability, and often do not have access to data that shows how the rest of their agencies pay similarly situated employees. Placing that burden on managers is not fair to them or to the people they lead.
When I started writing this series, I assumed I would get to this point and then lay out specific ideas for replacing the GS, with a lot of detail about each. As I researched and wrote this and the preceding 3 posts, I concluded the issue is far too complex to be solved with one person’s ideas, or even those of a small group. Some of the problems are long-standing (and were discussed 65 years ago by the Hoover Commission) with no resolution for many decades. Because of the complexity of the issues, the data needed to make sound recommendations, and the absolutely essential requirement that any substantial Civil Service reform follow an open and robust discussion of the public policy issues that are involved, I am going to limit my recommendations to the means by which I believe we can address replacing or reforming the General Schedule and the areas I believe should be included in that public policy discussion.
Civil Service Reform in Our Politicized Government. With the current state of our political system being what it is, using the regular legislative process to design a replacement for the General Schedule is unrealistic. The parties’ views of the civil service and how it should work are too far apart. We need some mechanism for allowing a policy discussion to occur outside the walls of the Capitol. There is a model that might work, and that is the one created by the Base Realignment and Closure (BRAC) process. The BRAC process was designed because the Department of Defense needed a means of shedding excess infrastructure, but it was apparent that no sane member of Congress wanted to go on record voting to close bases where their constituents work. There was also a recognition that DoD’s infrastructure was too big and too costly. In order to reduce the political problems, Congress passed the Base Closure and Realignment Act of 1990. The Act (amended and used again in 2005) created a Presidentially appointed BRAC Commission to review and make recommendations for closures and realignments. The process began with recommendations from the Department of Defense, followed by Commission reviews and recommendations. The Commission’s recommendations went to the President, who could make recommendations back to the Commission. Following that, the President had to approve or disapprove the entire list. If the President approved and sent it to Congress, the recommendations were implemented unless Congress passed a joint resolution stopping them. There was no provision for Congress to modify the list.
Although there are debates about the quality of some decisions the BRAC Commissions made, there was widespread agreement that the process allowed a far less political, more thoughtful and deliberate approach to a highly sensitive issue. A similar approach could be used to make recommendations for Civil Service reform, including replacement of the General Schedule. A Commission on the Public Service, authorized by BRAC-like legislation, appointed by the President and confirmed by the Senate, should be created. It should include representatives from government, academia, the private sector, and organized labor, and have a professional staff that includes experts in human resources, labor economics, industrial/organizational psychology, and other necessary professional disciplines. The Hoover Commission provides an excellent example of the work such a Commission can do. Presidents Truman and Hoover did not have a history of being friends. In fact, President Truman often cited President Hoover during his campaign (while Hoover was chairing the Commission), saying of President Hoover “…the great engineer we elected backed the train all the way into the waiting room and brought us to panic, depression and despair….” Following the 1948 election, when President Truman’s reelection surprised everyone (including Mr. Hoover), the two Presidents found a way to work together. The result was a personal friendship and a remarkable set of recommendations that are still felt in government (including the creation of GSA and the Joint Chiefs of Staff). Such a Commission today could create a new model for the Civil Service, with little risk that either party could drive it to extremes. Absent such a model, there is little likelihood our current political climate can produce significant reform.
The Commission’s Tasking. Following are some of the areas a Commission should address, with some rationale for including them.
- Simplify the numbers and types of job classifications. We should not have thousands of ways to classify a Federal job (and that doesn’t even consider the Federal Wage System). Rather than classifying jobs based on occupation-specific characteristics, the Commission should explore alternative approaches, such as classifying them based on cross-occupational characteristics that make up the requirements for a job. Using the biological concept of a Phenotype (a composite of the observable characteristics of an organism), the government could create a model that identifies characteristics of positions such as education, types of skills, physical characteristics, work environments, interactions with others, competency requirements, and others. Rather than a thousand or more ways to classify jobs, we could drive the number down to as few as a hundred.
- Reduce the number of grades to a smaller number of more inclusive grades. Broad banding systems that have been implemented in the past have had some success, but some have fallen victim to pay for performance issues and systemic bias. For a typical two-grade interval job series that currently has 8 grades (5/7/9/11/12/13/14/15), reducing to 3 or 4 grades would provide ample room for pay differences based on individual position requirements.
- Incentivize retention through longevity-based pay increases. The current GS system has steps within grades. Employees get step increases based on time in grade, with steps 2, 3 and 4 taking one year, steps 5, 6 and 7 taking two and steps 8, 9 and 10 taking three years. Each step represents about 3% of the pay for step 4 (the representative rate), so the difference in step 1 to step 10 in the same grade is about 30%. Going from step 1 to step 10 takes 18 years. The logic behind extending waiting periods is that employees are more likely to change jobs in the first few years in a job and their abilities grow quickly early in their tenure in a job. As the years pass, added experience becomes less of a differentiator. Twenty years doing the same job is generally no more valuable than 15 years. The Commission could explore options for making the incentives simpler to administer in combination with broader grades/bands using an approach that incentivizes retention in the early years when an employee is more likely to change careers and while competencies are rapidly increasing. Developing the specifics requires significant data and analysis, but it would help reduce turnover early in careers and make it clear that experience in government has value.
- Pay employees based upon the realities of the labor market. The Federal government competes for talent in a labor market that values some occupations more than others. For example, a Physician makes more than an HR Specialist. We also know that the cost of talent varies from one location to another, and the value of a given occupation can fluctuate due to supply and demand. The GS system is unresponsive to market fluctuations and the current locality pay process does not work effectively and results in pay localities that are far too large. Studies that compare Federal pay to the private sector often find pay is too high for some lower-graded jobs and too low for some higher-graded jobs. My experience was that the GS system is also particularly ill-suited to recruiting entry-level staff in occupations with high demand and high entry salaries. Gathering enough data to make market-based pay decisions is not simple, but it is not so difficult that we should avoid trying to do it.
- Include some accomplishment-based pay for truly extraordinary work. Although pay for performance has many flaws, most people recognize there are some employees whose accomplishments are far outside the norm. Pay increases based on such documentable accomplishments (not effort) can be done in a way that ensures they are not based on flawed performance rating processes.
- Annual comparability increases for employees whose performance is is fully successful. The pay raises the Federal workforce receives annually (when there is no pay freeze) should not be denied to employees whose performance meets requirements. One of the weaknesses in DoD’s National Security Personnel System was the policy of granting only 50% of the increase to employees with a “Valued Performer” (fully successful) rating. That undermined the objective of convincing employees a level 3 rating was a good rating.
Any effort to reform or replace the General Schedule will uncover many more considerations that must be accounted for in order to create an effective classification and pay system. If we have a robust public policy debate, informed by data and facts rather than politics, a Commission such as I am recommending could position us to face the challenges of this century with a system designed for it rather than one designed for the last century. If we fail to act and grades continue to creep up to the top of the grade structure, I believe the stresses on the General Schedule will reach a point where it becomes a crisis. It is critical that we address the problems in a thoughtful and informed way, rather than waiting for a future where the government is unable to recruit the talent it needs.