Do We Need a Million Fewer Federal Employees?

The Washington Post published an interesting opinion piece on August 29, suggesting the best way to have a better and smaller government is to hire another million Federal employees between now and 2035. The author makes the argument that many Federal programs are operated by state and local government, for-profit contractors and not-for-profit grantees. He also says moving much of that work back to direct performance by Federal employees could reduce cost, generate economies of scale, and result in overall government spending going down. He makes the argument that the number of Federal employees relative to the population of the United States is far lower than it was in 1965. To help make the point that there are too many contractors, the article also says the Department of Homeland Security has more contractors than Federal employees, citing a report from 2010.

DHS Does Not Have 200,000 Contractors

I was Chief Human Capital Officer for DHS in 2010. The Post article was correct when it said DHS reported it had more contractors than Feds. What is missing is that DHS later corrected that number and it actually had less than 100K contractors.

Hire a Million or Fire a Million?

The problem with some of these numbers is that we can make them say whatever we want them to say, based upon our biases and how we analyze and present the data. Using employment data we can argue we should hire a million Feds or that we should fire a million Feds or anything in between. The “hire another million Feds” argument goes like this: We currently have 2,038,000 non-postal Federal employees. The current U.S. population is 316,000,000. That means we have 155 residents per Federal employee. In 1965, we had a population of 193,000,000 and 1,900,000 Federal employees, for a ratio of 102 residents per employee. We have far fewer employees per U.S. resident than we had 49 years ago, so we have shrunk the Federal workforce considerably, at least when in terms of the Fed-to-resident ratio. The Office of Management and Budget cites similar numbers in their Analytical Perspectives on the FY 2013 budget submission, going back to the 1950s, when there were 92 workers for every resident. OMB used the numbers to point out that the Federal workforce is becoming far more highly skilled, but is not growing in real numbers. The Post article argues that the government should hire another million employees by 2035 and transition services back to the government from states, localities, contractors and non-profit grantees. A million new Feds is a nice number and it makes for an interesting headline, but it does not necessarily make good public policy.

Whether we believe the government needs more or fewer people, we should make the case based on the work that needs to be done, the benefits to the taxpayers, and the most efficient way to get it done. The workforce to population comparison does not do that. It takes two data points – population and Federal jobs – links them as top line numbers, then proceeds to argue that the number should be higher today because the ratio has changed. The problem with that is that our nation, our expectations of government, the world, the nature of work, technology, and virtually everything else has changed. Comparing the size of the 1950s or 1960s workforce with the 2014 workforce at a macro level does not tell us anything that we can use to determine if the number is too high, too low, or where it ought to be. When we look at the number of employees, the number of contractors or the number of grantees, we need to dig a lot deeper in the numbers before we make public policy.

The belief that government employment has failed to keep pace with the population is not new. OMB’s Analytical Perspectives addressed the issue directly –

“Fifty years ago, most white-collar Federal employees performed clerical tasks, such as posting Census figures in ledgers and retrieving taxpayer records from file rooms. Today their jobs are vastly different, requiring advanced skills to serve a knowledge-based economy. Professionals such as doctors, engineers, scientists, statisticians, and lawyers now make up a large portion of the Federal workforce. More than half (55 percent) of Federal workers work in the nine highest-paying occupation groups as judges, engineers, scientists, nuclear plant inspectors, etc., compared to about a third (33 percent) of private sector workers in those same nine highest paying occupation groups. In contrast, 45 percent of private sector workers work in the seven lowest-paying occupation groups as cooks, janitors, service workers, clerks, laborers, manufacturing workers, etc. About 26 percent of Federal workers work in those seven lowest-paying occupation groups. Between 1981 and 2011, the proportion of the Federal workforce in clerical occupations fell from 19.4 percent to 5.1 percent of the workforce, and the proportion of blue-collar workers fell from 22.0 percent to 9.7 percent.”

OMB rightly pointed out that the workforce of today is far different from what we had in the 1950s and 1960s. Most readers are probably too young to remember, but in 1965 most offices were full of clerks. In fact, when the Classification Act was passed in 1949, the work force was mostly clerical. A Washington Post article from January 2014 put the number at three quarters of the workforce. (In the 1950s we had about 400,000 blue collar employees, so it is more likely that 3/4 of white collar employees were clerks.)

Where did all of those clerks go?

1950s and 1960s offices were full of clerks doing paperwork. They were pulling paper files from file cabinets. They were processing personnel actions using paper documents and paper official personnel folders. They did bookkeeping (using paper ledgers), they worked at telephone switchboards. And they typed. On typewriters. That work is mostly gone due to automation. People input their own work on the computers sitting on their desks. The idea that we would pay someone to write something with a pen and paper and then hand it off to another person to use a typewriter or word processing software to finish it is a relic of the past. The idea of accounting technicians by the thousands doing double entry bookkeeping in paper ledgers is just as dead. Typewriters and telephone switchboards are museum pieces. The nature of work in the Federal government has changed so radically that comparing top line 1950s and 1960s employment ratios to today tells us nothing. The more accurate comparison of 1950s and 1960s government to 2014 is the white collar nonclerical workforce, because the role of government is defined far more by higher-skilled and higher-paying occupations than by clerical and blue collar jobs.  I say that because most clerical work is done to facilitate other , more substantive, work. Most blue collar work is either maintaining facilities (another example of facilitating other work) or repairing boats, ships and aircraft. If we correct for the number of blue collar and clerical positions in government in the 1950s and 1960s, we get an entirely different view of the ratio of employees to residents. I have included a chart with all the numbers at the end of this post for those folks who want to see details.

In 1955 we had 503 residents per non-clerical white collar employee – today we have 183. The workforce performing substantive government work has grown a lot. As technology moved into government, it displaced clerical workers. The white collar workforce fluctuated over the years, but the number of non-clerical white collar workers has continued to grow. In just the last 5 years, the number of Federal employees started at 2,038,000, rose for a few years, and is now back to 2,038,00, yet the size of the clerical workforce has dropped from 151,000 in 2009 to 123,000 today and the blue collar workforce has dropped from 207,000 to 189,000. Using the comparison of nonclerical white collar work in the past and today, one could make the argument we have a million too many Federal employees. It is an absurd idea, but the idea of laying off a million Feds and cutting the budget more than $108 billion is about as realistic as hiring a million. Neither is going to happen, because neither is based on a realistic analysis of the work that needs to be done, the political environment, our economic conditions, or any of countless other data points that must go into any discussion of the right size of the workforce.


The Details

Federal Employment vs Population 2


Is it All or Nothing on General Schedule Reform?

Last week the Government Accountability Office (GAO) issued a report on the General Schedule. The report, “OPM Needs to Improve the Design, Management, and Oversight of the Federal Classification System” outlines eight key attributes of a modern classification system and assesses the degree to which the General Schedule aligns with those attributes. The eight are:

  • Internal equity. All employees with comparable qualifications and responsibilities for their respective occupations are assigned the same grade level.
  • External equity. All employees with comparable qualifications and responsibilities are assigned grade levels and corresponding pay ranges comparable to the nonfederal sector.
  • Transparency. A comprehensible and predictable system that employees, management, and taxpayers can understand.
  • Flexibility. The ease and ability to modify the system to meet agency-specific needs and mission requirements, including modifying rates of pay for certain occupations to attract a qualified workforce, within the framework of a uniform government-wide system.
  • Adaptability. The ease and ability to conduct a periodic, fundamental review of the entire classification system that enables the system to evolve as the workforce and workplace changes.
  • Simplicity. A system that enables interagency mobility and comparisons with a rational number of occupations and clear career ladders with meaningful differences in skills and performance, as well as a system that can be cost-effectively maintained and managed.
  • Rank-in-position. A classification of positions based on mission needs and then hiring individuals with those qualifications.
  • Rank-in-person. A classification of employees based on their unique skills and abilities.

GAO recognized that some of these attributes appear to be in conflict with one another. For example, classifying positions based upon rank-in-position and based upon rank-in-person appear to be polar opposites. GAO also criticizes the Office of Personnel Management for not having adequate oversight of GS classification, not updating standards frequently enough, and not adequately resourcing the program. OPM partially concurred with most of GAO’s recommendations, but did not concur on the issue of developing a strategy to systematically track and prioritize updates to occupational standards.

Both GAO and OPM positions have merit. GAO is correct that the GS system is outdated and overly complex. They are right that OPM has not adequately resourced the program. They are right that OPM should address the shortcomings of the GS system sooner rather than later. OPM is correct in saying it does not need a new process to get classification standards written and updated more regularly. OPM also said it has only 6 full time classification policy staff. GAO says OPM has to make tradeoffs the same as any other agency.

That last point is one that I think is important. Yes, OPM has to make tradeoffs. I have not hesitated to criticize OPM when it is off track. In this case, I think they are in a no-win situation. The agency’s recent budget requests show OPM’s appropriated dollars are significantly less than they have been in the past. The agency operates more on revolving funds than appropriations. They also have trust fund money that covers management of retirement and insurance programs. Those color-of-money distinctions are important, because they limit the number of people OPM can devote to policy and oversight work. While the agency has over 5,000 employees, less than 20% of them can be assigned policy, oversight and agency management tasks. That puts OPM in the position of playing management whack-a-mole with priorities. If they decide to devote far more resources to classification, they will have to come from other programs, leading to those programs being under resourced.

GAO’s Recommendations

GAO’s report established a good set of criteria for an effective classification system. It also made three recommendations for executive branch action:

  • Working through the CHCO Council, and in conjunction with key stakeholders such as the Office of Management and Budget, unions, and others, should use prior studies and lessons learned from demonstration projects and alternative systems to examine ways to make the GS system’s design and implementation more consistent with the attributes of a modern, effective classification system. To the extent warranted, develop a legislative proposal for congressional consideration.
  • Develop cost-effective mechanisms to oversee agency implementation of the classification system as required by law.
    • Develop a strategy to systematically track and prioritize updates to occupational standards.
    • Develop a strategy that will enable OPM to more effectively and routinely monitor agencies’ implementation of classification standards.

The first recommendation is the most critical, because it could negate the need for at least one of the other two recommendations. In recent years the CHCO Council has become an effective means of driving government human capital policy. A CHCO Council working group, partnering with OMB and with unions (at the national level), could develop a set of policy recommendations that would make the current classification process far less complex, without changing the underlying laws. That last point is critical – the likelihood of significant Civil Service reform that is enacted by the Congress is remote, due to the combination of Congressional dysfunction and a lack of appetite for Civil Service reform.

Reform Does Not Require Congress to Act

How would we dramatically change GS classification without rewriting the law? Most people think the highly complex GS system is entirely a creation of the Classification Act of 1949 (as amended). It is true that the Classification Act created the General Schedule and defines each of the 15 GS grade levels. Much of the complexity (23 occupational families and 420 job series) comes from policy decisions made by OPM and others in the 65 years since the Classification Act passed. There is no legal requirement to have 420 job series. OPM creates new series when it determines they are necessary, is asked by the White House to do so, gets statutory direction, or they are requested by agencies. A great example is Cyber Security. There is a lot of pressure (including in the intro of the GAO report) to create a Cyber Security job series. OPM has not done so, and with good reason. At the Department of Homeland Security we had a need for more Cyber Security professionals. They were not in a single job series and could not be. Cyber Security is a complex field that includes Computer Scientists, Network and Systems Engineers, Security Specialists, Digital Forensics Specialists, Program Managers, Intelligence Specialists, and about 10 more categories. A single job family cannot address so many different positions that have radically different duties and qualifications requirements. In fact, GAO points out that the use of 420 job series adds unnecessary complexity to the GS system.

Where we need to go is in the opposite direction. The number of job series should be reduced by at least half, and more likely by three quarters. One reason OPM cannot maintain current standards for all of the jobs is that there are too many of them. GAO also points out that the stove piping of jobs into narrow series may hamper career growth. So – we have too many series, we cannot maintain the classification standards because of that, and the number limits agency flexibility on reassigning staff. It also makes for an arduous and overly complex hiring process for applicants from outside government.

If the number of job series is reduced to a more manageable number (I suggest no more than 100), we could achieve most of the objectives of GAO’s eight attributes of a modern classification system. Even some of the apparently conflicting attributes can be addressed. For example, on the surface it appears we cannot have a system that includes both rank-in-person and rank-in-position attributes. But we can. Take a look at the Research Grade Evaluation Guide (RGEG) published by OPM. The RGEG includes 4 classification factors:

  1. Research Assignment
  2. Supervisory Controls
  3. Guidelines and Originality, and
  4. Contributions, Impact, and Stature

The RGEG recognizes that “Work commonly expands commensurate with the researcher’s motivation, capability, and creativity.” Evaluation of factor 4 is based upon the researcher’s accomplishments rather than a rigid standard based upon the job itself. The RGEG recognizes that it is difficult to separate the person from the work the person does. OPM could use a similar approach to incorporate both rank-in-person and rank-in-position into new classification standards. The RGEG also covers research in many fields – there is not an RGEG for Physics, one for Chemistry, one for medicine, etc. OPM could use a similar approach with multi-series standards to dramatically reduce the number of classification standards it has to write and maintain. For example, a STEM Grade Evaluation Guide could cover many STEM positions. An Administrative Grade Evaluation Guide could cover financial management, human resources, procurement and other administrative positions.

Benefits of Administrative Reform

Administrative reform is faster, more achievable, and less likely to veer off into Fed bashing than a statutory solution might be. It maintains stability in the legal framework of the Civil Service, yet addresses the Adaptability feature in GAO’s 8 attributes. It achieves Simplicity, adopts both Rank-in-Person and Rank-in-Position attributes, and should also improve Internal Equity. It certainly demonstrates Flexibility as well. While we may not be able to get to a completely modern classification system without Congressional action, we can certainly improve on what we have today. By pursuing an administrative rather than statutory solution, OPM can begin to rapidly address many of the shortcomings of the existing system. The points of view of key stakeholders, such as unions and the Senior Executives Association, can be taken into account, as can those of good government advocates such as the Partnership for Public Service and the National Academy of Public Administration. The CHCO Council can drive the process, ensuring the resulting changes are implementable and consistent with accomplishing agency missions. Rather than waiting for the day when Civil Service reform might be achievable in Congress, we can act now. Why wait??


Tagged , , , , , , , , ,

A Tale of Two Agencies

“It was the best of times, it was the worst of times.” A Tale of Two Cities by Charles Dickens

Dickens’ classic novel may not seem to be related to the Federal government, but the conflicting view in that opening sentence reminds me of the distinctly different views groups of people in the same agency may have. Understanding those differences is crucial to any transformation effort, particularly if what we are trying to change is the culture of the agency. If we do not know where the starting point is, how do we know how to get to our desired end state?

This problem is compounded in culture discussions because agencies are not monoliths with a common set of issues, values and norms. They are most often collections of sub-organizations, teams and groups that have different perspectives on the agency, its mission, and its problems. Sometimes the differences are based upon traditional perspectives such as headquarters and field – others may be based upon project assignments, geographic locations, or type of work. Some of those groups may believe they work in the best place in the government, while others believe they are overworked, under appreciated, and working in one of the worst places in government. Those divisions can be particularly acute when they involve the views of senior leaders vs. rank-and-file employees.

Employee surveys often reveal that employees at the highest levels of agencies (such as GS-15 and SES) have different perceptions of the workplace from those of employees at lower grades. That fact should not come as surprise to anyone who looks at the working conditions for senior employes vs. those of the majority of the work force. Much of my career was in the Department of Defense. Here is a quick look at how my work life changed when I moved up through the ranks from GS-5 to SES.


  • Arrive at work and park reserved parking spot close to door
  • Go to office suite, saying good morning to my Deputy, my Deputy’s Secretary, my Secretary, and my Executive Assistant
  • Enter my private office
  • Participate in meetings I have arranged or am scheduled for, with reminders from staff if I am running behind
  • Based on those meetings, maintain a good understanding of the key issues the agency is facing
  • Make or participate in making decisions about the future of the agency
  • Based on DoD custom, staff members stand when the senior SES, General or Admiral enters the room (I asked my staff to stop doing that)
  • Make assignments and review the work of others

GS-5 through GS-12:

  • Arrive at work, park far from the building and hike to the door
  • Go to my 6×8 foot fabric-covered box, walking past my coworkers in their fabric-covered boxes
  • Participate in meetings I am told to participate in, get chewed out if I am running late
  • Learn about issues and decisions after the fact and when my boss or someone else in the agency communicates with the staff
  • Stand when an SES, General or Admiral enters the room
  • Get work assigned to me and have my work reviewed by several people

Would the two people in these circumstances have different views of the agency? Of course they would. Having a team of people to support you, a great office,    and being fully informed all the time can make for a great workplace. Some of the negatives include having to deal with Congressional staff and members, working directly with political appointees who may have a short term view of the world, and taking the heat when the agency has problems.

My view of the world from that perspective was very different from my view when I was at a lower grade, working in a cubicle, and not having ready access to information. Even with the headaches that come with senior positions, they can offer a rewarding work experience. If an agency wants to address culture issues, it has to recognize that differently situated people have different views of the agency and require different solutions. It is essential for senior leaders to recognize that, even with the challenges their work presents, their view of the organization is likely to be affected by their conditions of employment.

I have to admit I have gotten a frosty reception when I have expressed that view to SES and General and Flag officer colleagues in the past. Some of them did not want to hear that we had two agencies, and that what may be the best of agencies for some was a different experience for others. Because they lived in a different version of the agency, they questioned survey results that told us parts of the agency were hurting.

Some differences are inherent in the types of work done by people who are individual contributors and those who lead an agency. Senior leaders will obviously be better informed about decisions, because they are making them. The breadth of their responsibilities also necessitates a staff to support them. Other distinctions are not so necessary. Reserved parking is not a necessity, nor is standing when someone senior enters a room (unless it is the President). Workplaces can be designed so employees are not working in cubicles. Information can be shared and decision-making processes can be more transparent and have more employee involvement.

Anyone who embarks on a culture change initiative needs to take a hard look at this issue and the other subcultures that exist in most agencies. By doing so, they will get a better understanding of their own work force and their needs, and have a better chance of success with changes they attempt to make.


  1. A Tale of Two Cities and thousands more public domain books are available free from Project Gutenberg.
  2. Thanks to Andrew Krzmarzick for the idea for this post.

To RIF or Not to RIF

Check out my new article on – “To RIF or Not to RIF”

Reduction in force is a term that frightens most federal workers. It means uncertainty, potential loss of a job, disruption, and usually more questions than answers. Politicians (even the ones who want to shrink the federal government) oppose them. So do managers, unions, and most people who write about government issues. Most agencies have been “successful” in recent years in avoiding RIFs. They have used attrition, hiring slowdowns and buyouts to reduce their workforce without resorting to a RIF. Most people will tell you a RIF is something to be avoided at all costs.

During my federal career I developed an early understanding of the RIF process when I faced a reduction in force in my first federal job. I immersed myself in the subject when I had to conduct several RIFs over the next 30 years. The last large scale RIF I conducted abolished 700 occupied jobs in a Navy command of 3,200 employees. With that background, I should join the chorus that says RIFs are always bad and attrition is always better. The problem is that it is just not true. Sometimes the alternatives are worse. In fact, sometimes the alternatives do much more damage and disrupt the organization far longer than a RIF might.

Continue reading on

Tele-Goofing Off?

I read an interesting article on telework in Government Executive last week. The headline asked “Is This the Right Time for the White House to Promote Working from Home?” Reporter Kellie Lunney’s article highlighted OPM Director Katherine Archuleta’s August 22 memorandum to agency heads on expansion of workplace flexibilities and work-life programs. The question regarding timing is certainly legitimate. The US Patent and Trademark Office (USPTO) is in the midst of a scandal regarding its telework program. The USPTO issues are representative in type (but probably not scope) of those other agencies are experiencing.

Those issues (tele-goofing off rather than working, inability to accurately measure productivity) and management unwillingness to deal effectively with performance and conduct issues) put workplace flexibilities at risk. They can damage the reputation of the Federal workforce, and they provide fodder to those who believe Federal employees are doing too little work for too much pay.

Telework and other workplace flexibilities are a good thing for government and the private sector. They allow organizations to dramatically reduce their office footprint and reduce costs. They take traffic off the roads and reduce congestion. They allow employees to reclaim their commuting time and use it for personal pursuits. In a labor market where telework is commonplace, having effective telework programs allows employers to compete for talent. They allow employers to remain open during weather emergencies because their workforce can continue working from their home offices. Telework has its detractors too. They claim it is too hard to monitor performance when people work from home. They say some employees use telework as a substitute for child care and are not productive at home. They believe telework interferes with team dynamics and makes working as a cohesive team much more difficult. There is some truth to all of the complaints and praises of telework, but the fact is telework is expanding and will continue to expand.

Rather than attempting to reverse telework programs, we should be working on fixing the problems. Nothing I have experienced, heard from people currently in government, or read tells me telework and other workplace flexibilities should be killed. That toothpaste is out of the tube. Government should take steps to improve accountability and make the programs work. I know they can. Some of the agencies where I worked handle telework effectively. My current employer, ICF International, uses telework extensively. Our workforce is responsible and incidents of abuse are rare. We are able to assess employee productivity and know when problems are developing. We are not alone. Others in industry and in government manage these programs well.

Some of the problems with telework are also problems when employees are working from their employer’s building. I remember a discussion with a colleague who was an Army Major General. He opposed telework and refused to allow it in his organization. When I asked why, he said “I need to know people are working, and when I want to speak with someone, they need to be here.” So I asked, “Do you know who is working on the other side of your office wall? If you leave your private office and walk back to where the dozens of cubicles are, will you know who is working? When you want to speak with someone, do you summon them? Walk back to their cubicle? Or do you send an email or pick up the phone?” He agreed that he didn’t know if everyone was working at that time and if he wanted to speak with someone he would call them.

Short of walking around all day and looking in all of the cubicles, managers cannot know who is working simply by proximity. They need measures of performance and productivity. Good measures work whether the employee is in a Federal office building or working from home. The most critical key to effective telework programs is good productivity and performance measures. Without them, telework is less effective. However, without them, working in an office – any office – is less effective.

Plans to implement telework often focus on some of the obvious needs – technology, agreements with the union, processes for requesting and approving it, and the adequacy of a home workspace. Performance is usually part of the discussion, but detailed performance and productivity measures are not always part of that discussion. They must be if telework is to be effective. Managers should have a discussion with each teleworker before their first day of working from home to explain what is expected, hours they are expected to work and any flexibility of hours that is allowed, how their productivity will be measured, and what good performance looks like. Notice the difference between telework and working in the office? No? That is because there is no difference. Other than those things that are specifically caused by working remotely, such as the need for access to systems from home, there is no real difference between telework and office work.

The one problem that can be a deal killer for telework is trust. If an employee cannot be trusted to work independently, s/he cannot be trusted to telework. S/he also cannot be trusted to sit in a cubicle in the office and work. The only real difference in those cases is that there are coworkers around and they may report the person for goofing off.

If we address the issues of measures and trust, there is no reason why telework should not be expanded. If the issues are not addressed and support for telework begins to wane, the government may find itself with one more recruiting challenge (on top of Fed-bashing, pay freezes, and furloughs) – the lack of a perceived benefit that is offered by most people who are competing for the same talent. We already have enough challenges in recruiting young people for government.



Get every new post delivered to your Inbox.

Join 695 other followers