Rating employee performance appraisal systems
How would you rate your agency's employee performance appraisal system?
In his October 1999 Public Service column in Government Executive, Brookings Institution scholar Paul Light argued that "Congress and the President have known for years that the performance appraisal system is broken and have done nothing to fix it." (See "Behold Lake Wobegon East.")
But in a letter responding to Light's column, Office of Personnel Management Director Janice Lachance defended the Clinton Administration's efforts to improve federal employees' performance evaluation systems. Her letter is published below, followed by a response from Light.
Janice Lachance, director of the Office of Personnel Management, writes:
Upon reading the recent commentary by Paul Light 'Behold Lake Wobegon East' in your magazine, I felt compelled to respond.
His criticism of the federal performance appraisal system reveals a misunderstanding of the process, the significant improvements which have been made over the past several years, and the role of the Office of Personnel Management. The fact is that the decentralized performance appraisal systems of today, for all of their faults, are much more effective than the 'top-down' mandated system of the past. In fact, the important changes to the performance appraisal system of recent years renders the generalizations made in the commentary all the more untenable. Therefore, I would like to offer your readers a more complete and realistic appraisal of the situation.
First of all, I am troubled by the charge that OPM attempts to hide data on government-wide averages. We opted to sacrifice the calculation of average ratings in exchange for custom tailored performance appraisal systems that could work for individual agencies.
OPM no longer calculates government-wide appraisal rating averages because the system has been decentralized. It is impossible to average ratings that are based on different scales. Regulations dictate that '3' represents a passing rating, however, a '3' based on a five-point scale is not equivalent to a '3' based on a two-point scale. To average those numbers would be inappropriate and utterly meaningless to anyone studying the data.
Next, Mr. Light seems to base his opinions on two mistaken assumptions. He assumes that the performance rating of record is the measure by which agencies evaluate all performance. The fact is that government agencies (as well as private sector industries) are well aware of dangers of placing too much importance on the individual summary rating, which can represent just one form of evaluating employee performance. Increasingly, agencies like the Department of Veterans Affairs and the U.S. Mint are also using balanced measures of results to make judgments about employee performance, particularly group performance.
Mr. Light's assumption that all agencies follow one system of performance appraisal is demonstrably false and unfortunately misinforms the rest of his piece. The fact is, contrary to his assertion of indifference on the part of the President and Vice President, the Clinton-Gore administration significantly decentralized the performance appraisal process in 1995. As a result, agencies were able to craft performance appraisal systems that addressed their individual mission better than a one-size-fits-all, government-wide approach.
The results have been very encouraging both for agencies and for OPM. Agencies are experimenting with bold new approaches to performance appraisal and reward systems, including linking cash awards to group level achievement rather than individual ratings.
Rather than resort to so-called solutions like artificially capping outstanding ratings (which would only lead to the different problem of rotating the top ratings), agencies are taking prudent steps to develop performance appraisal systems that effectively evaluate the productivity and quality of their workforces.
In fact, many agencies have taken steps (including the use of pass/fail ratings which Mr. Light bemoans) to deal with the very problem of rating inflation that is addressed in the commentary.
I realize that performance ratings are a subject that many managers approach with discomfort. This is true not only in federal departments and agencies, but in private companies and industries as well. Few managers, if any, relish the opportunity to rate a fellow employee, especially when those ratings have important consequences. In recognition of this reality, OPM has a responsibility to assist managers in this task by allowing them to evaluate performance in ways that will best serve their mission and be fairest to those who are rated. I believe that the development of our new decentralized performance appraisal system was a move toward this objective. It was a move to meet our obligation to both managers and employees across our diverse federal community.
Paul Light responds:
Janice Lachance makes a number of useful points in her response to my recent column on Lake Wobegone East, and I'm delighted to hear of all the progress being made on performance appraisal and getting rid of dead wood.
If I'm reading Ms. Lachance correctly and she is actually suggesting that we just get rid of the old system entirely, I wholeheartedly agree. Better to have nothing at all than an annual charade. (Am I the only one who thinks Ms. Lachance undermines her endorsement of all the good work out there when she suggests that capping the number of outstanding rankings would "only lead to the different problem of rotating the top rankings"? Can she be suggesting that federal managers might actually play games with the annual ratings system to compensate for the lack of competitive pay?)
The only problem is that the administration is unwilling to drop the old pay for performance rhetoric. Taxpayers won't drop it either. There's a real politics about this, after all. We keep talking about rescuing pay for performance by linking it to customer service or balanced scorecards, all the while refusing to let managers cap the number of outstanding ratings or put real money into play. Hence, we get the rise of pass/fail systems that surrender all hope of using the performance appraisal system as a disciplining or rewarding tool. If the action is all elsewhere, and pass/fail is just a device to dodge a trip to the grievance office, we ought to just admit it.
As for the data, if OPM has the agency-by-agency and system-by-system ratings, let's see them on the Web site. The Fact Book creates the illusion that there is still such a thing as a single civil service. Why not get rid of the Fact Book, then? And in such a wonderful, wildly decentralized system in which agencies can pretty much do what they wish, and managers are fully liberated from the rules, one wonders why we still have that ridiculous classification system and Title 5 CFR. If Ms. Lachance wants to get rid of that, I'm in her corner, too. Perhaps we're not so far apart after all.
Finally, as for the notion that a pass/fail system is the answer to grade inflation, tell that to all the schools around the country that have been pushed to limit grade inflation by using tougher standarized tests. Better yet, try getting the president or vice president to make that recommendation in a campaign speech. Americans won't tolerate it for their children anymore than they will tolerate it in their government. I'm all for better ways of evaluating employees as long as they give managers the ability to pull the plug on poor performance and shine the light on excellence. Pass/fail systems do neither very well.
Add your opinion:
Send your comments on this debate to firstname.lastname@example.org, and we will add them to this page. Please include name, title and agency. Responses may be condensed.
As a Federal employee, I have to agree with Mr. Light . . . the system is broken. The only way I can see to fix it is to a) make it easier to fire non-performing individuals and individuals with behavioral issues that are in the civil service; b) tie appraisals to actual, non-subjective, criteria so that the "boss's pet," who no one else can work with, is rated objectively, not based on their mutual admiration fest with the boss; and c) tie performance awards (no matter how piddly they are) to some hard criteria, rather than the boss's discretion.
Finally, I do not think that basing salary on performance is the answer. The same individuals that get the highest ratings and most awards because they are the "boss's buddy" will be receiving the pay raises. Think about it. We need to remove any possibility of favoritism from the system. Personally, I vote for the 360 degree performance rating system. As a past federal manager, I employed this method one year, and received excellent feedback (i.e., areas that I needed to work on were identified) and what I felt was a fair rating (and no, it wasn't a superior).
Thanks for the opportunity to state my point of view!
EPA, Region 4
I have experienced both the five-level system and the pass/fail system. The five-level system is hobbled by anxiety on the part of the supervisor and employee because it is judgemental. The pass/fail system is much easier on the nerves because it is much less judgemental, though it is noncommittal.
However, neither can compensate for the problems encountered due to favoritism. The list of favored persons, persons for whom there is no opinion, and the persons out of favor changes depending on who is in charge. I have been both a favored person and an unfavored person and have felt the elation and consternation of both positions.
I have seen good people leave due to their out of favor position. People who have nowhere to go are left behind.
I work for the Department of the Navy and we have been using a pass/fail system for the past four years.
I agree with the statement that "the performance appraisal system is broken." I have been rated and have rated employees under the five-tier system and the pass/fail system. Neither system accomplishes much of anything. I have noticed consistently with the government that reinventing the wheel is the motto, which most of the time costs extra money and never works.
My suggestion is to adopt a system from the private sector where competition is real and rating employees means something. What the present system of pass/fail does, which not every government employee is rated on, is put the pass/fail-rated employee at a disadvantage during a [reduction-in-force].
The present RIF policies are based on a five-tier rating system. With pass/fail, the most points an employee who gets a pass rating gets is that of successful. In a time where RIFs are running rampant, the decision to decentralize the rating systems of federal employees is sheer madness and only hurts the already disgruntled federal employee.
Department of the Navy
Ms. Lachance puts a different spin on being a "Team Player." She also misses the dangers of a pass/fail appraisal system.
What motivation is there for an employee to excel if he/she will receive the same rating as a less productive co-worker? We need to scrap the present system and replace it with a system that will financially reward an employee based solely on performance and allow supervision to remove "dead wood." The average civil servant is an honest, hard working individual who deserves better.
Veterans Affairs Hospital
Shame, shame, shame. [Paul Light's]article, "Behold Lake Wobegon East," complains about the so called hyperinflation in ratings-the average employee being rated above fully satisfactory. In her response, Ms. Janice Lachance was more than polite in pointing out that Mr. Light misunderstands the Federal appraisal process.
Unfortunately, Mr. Light is not the only one who does not understand the basis of the system. Consider the following jaundiced, but accurate view of the appraisal process:
1. The appraisal process is first of all based on written performance standards. All subsequent ratings are based on whether each standard is met or unmet. Thus, the standards are key to the process. These standards are developed by the immediate supervisor (and approved by higher levels). Some agencies require at least two standards (one critical, the other non-critical), but there is no limit to the number of standards that can be established.
We must remember that a fundamental requirement is that standards must be established that can be exceeded. The result of this requirement is that the level of performance for any standard is what can be tolerated. Any less for a critical element standard is the basis for removal from the federal service. Any more can serve as the basis for rewarding the employee. Of course, supervisors can demand harder or easier standards, but the end result is only a shifting of what can be tolerated.
2. Given this foundation, the rating process is largely mechanical. To be fully successful, an employee must meet all elements of the performance standard. This translates to employee performance that can only be tolerated. How would you like to be rated fully successful on this basis? Most supervisors understand that a rating of fully successful is the kiss of death for an employee when being considered for promotion-they will just not be competitive.
3. Consider now performance below a fully satisfactory level. Minimally successful means only that one or more non-critical elements have not been met. This rating has nothing to do with ratings on critical performance standards. Failure to meet any critical element results in an unacceptable level of performance. This is what is meant by fail in a pass/fail system. It provides a means to take action when an employee can't or won't perform the critical requirements of his or her position.
A supervisor is required by law to take action (removal, downgrading, or reassignment) for the failure to meet any single standard. But, employees must be given an opportunity to improve their performance. It is this process, and concern for subsequent employee grievances or appeals that result in making supervisors hesitant in rating an element as unmet. Most supervisors do not know that the burden of proof for their actions in performance ratings was substantially reduced by the Civil Service Reform Act of 1978. The earlier requirement had been "the preponderance of the evidence." It is now only "substantial evidence." Supervisors also do not know that an action of this kind is not likely to be overturned on appeal.
4. Fully successful and above reflect a passing grade in a pass/fail system. Fully successful is defined above. Highly successful and outstanding are arbitrarily defined levels. Highly successful means only that most (over 50%) of the performance elements have been exceeded.
Is it too much to expect that a large number of workers are beyond being just tolerated and perform in a highly successful manner?
Outstanding is defined as exceeding all established performance elements. That is a tough standard by anyone's perception.
Unfortunately, however, a worker who makes a significant contribution in one area cannot be rated outstanding if any other performance standard is not exceeded. Consider the 3M employee who developed and advocated for acceptance of the "Post-it notes." Would any worker in the private sector not be considered outstanding for such a contribution?
As you can see, the adjective ratings are closely defined in each instance. We cannot extend their meaning then, to common every day understandings of these words.
It would seem from Mr. Light's writings that he uses the federal performance review process as simply a means for bureaucrat bashing. What's even worse, he seems to take on the system from a partisan point of view.
The remedies he proposes are the wrong answers for a very complex process. Capping the number of outstanding ratings has been done in the past, to the detriment of many hard working and competent people. More money for incentives sends the wrong signal to the many people who are intrinsically motivated to work for the public interest-the carrot and stick is not an effective approach to leadership in the long term.
Quality management guru W. Edwards Deming makes a strong and compelling case to eliminate performance ratings altogether. In every work situation-public or private-there is always a small number of workers, however, that can't be helped.
In such cases we must have an effective process to separate them from the workforce for the efficiency of the service. It is not likely that we would need such a system to deal with teams or units. There are many other ways to recognize group contributions, but the performance system is not necessary for this purpose.
There is always room for differing points of view. Ms. Lachance provides effective counter arguments to the issues raised by Mr. Light. The difference seems to be in a spirit of good faith that is necessary to make us listen to both sides of a question. We can't think much of scholarship when a critic forgets to fairly consider all sides of an issue-not just his own bias.
Robert W. Crittenden
Retired federal manager
The current five-level rating systems works great as long as everyone gets an Outstanding rating. The five-level rating system is basically the same as the two level pass/fail system. If you get a 1 (Outstanding) you pass; if you get a 2 or lower rating, you fail.
With approximately 70 percent of employees at Fort Sill getting an outstanding rating of 1, anyone with a lower rating is below average and in the bottom 30 percent. Any rating below average needs to be justified with examples of how an employee can exceed measurable performance standards, advance notice that the employee is not exceeding the standard, and a chance to improve performance.
As for removing dead wood, the dead wood in the employee ranks has already been removed during the downsizing and reorganization process. Also, any employee that gets rated below 3 (satisfactory) gets harassed, documented, and fired. That is why there are so few federal employees still working that are rated below 3.
The remaining dead wood is in the supervisor and management ranks. It should not be easy for them to fire federal employees because it is usually the supervisor that should be fired, not the federal employee.
Federal employees want a supervisor that treats them with dignity and respect in a non-hostile work environment. They do not want a rating system that is used for punishment, not for rewards. The current rating system lowers employee morale and creates conflict between employees and supervisors.
Performance systems should be abolished, employees should be vested with job protection, and formal judicial action should be required to remove a federal employee from his/her job. Charges should be substantiated with facts, not the supervisor's unsupported opinion of exaggerated, distorted, enormous impact of minor infractions of bureaucratic rules and regulations.
The chain of command in the government agency will always support the supervisor and should not be the judge in matters of such importance and permanent impact on an employee's life as removal from the government service.
Ronald W. Ballew
Vice President for Reynolds Army Community Hospital
National Federation of Federal Employees (NFFE) Local 273
Fort Sill, Okla. (U.S. Army)
W. Edwards Deming, a believer in people, said it best in Out of Crisis-"The performance of anybody is the result of a combination of many forces-the person himself, the people that he works with, the job, the material that he works on, his equipment, his customer, his management, his supervision, environmental conditions (noise, confusion, poor food in the company's cafeteria) These forces will produce unbelievably large differences between people. In fact, apparent differences between people arise almost entirely from the action of the system that they work in, not from people themselves. A man not promoted is unable to understand why his performance is lower than someone else's. No wonder; his rating was the result of a lottery. Unfortunately, he takes his rating seriously."
I therefore recommend that less emphasis be put on changing people and more emphasis be put on changing the environment where we work. I recommend that the focus be placed on development and not on evaluation; that our appraisals be a learning tool, not a grading tool.
Having bosses as both evaluator and coach creates a challenge too. Are employees willing to share their weaknesses at the same time their being evaluated? Are employees going to negotiate performance objectives that challenge them or ones that are easily met? I don't think so. The risks for sharing weaknesses and not meeting a challenging objective are too high.
I recommend that the appraisal system value differences rather than award conformance; it is our differences that expand our possibilities. Current performance ratings are directly proportional to the employees ability to conform to the boss' personal preferences.
Place more emphasis on team and customer; put less emphasis on individual performance. The appraisal system shifts our focus from team and customer to our evaluators/boss. By putting the focus on team and customer, I begin existing for a far greater purpose than myself or my boss. A win for the team and customer is far greater than a win for an individual or boss. In many cases an individual or boss win results in a loss for the team and customer.
Design the appraisal system so that we're not only accountable for the work we do, but also for the processes we own. We must despise complacency if we're to compete in this new government revolution. If we continue to exist as we always have, we will cease to exist.
Fort Carson, Directorate of Public Works
I think different appraisal systems for each agency put federal employees at a serious disadvantage.
Federal employees are working in an era where change is a certainty. [Reductions-in-force] and reorganization are facts of life, and many employees must look outside their agencies if they want a promotion. The flexibility to move from agency to agency is a must. In this environment, a pass/fail system can hobble an employee who is competing against individuals rated with a five-level system.
A "pass" cannot look as good as a top-tier rating of a five-level system. An individual rated under a pass/fail system may have his/her application scored less than one with a performance rating which shows excellence. It is also difficult to compare a three-level system to a five-level system - the top rating in a five-level system will still look better than the top rating of a three-level system.
The additional performance levels allow an accurate rating to be made if the process is done honestly. As far as documenting poor performers is concerned, any of the appraisal systems now used can be effective. The problems of motivation, weeding out poor performers, and rewarding excellence are not addressed by changing the number of rating levels.
The variety of performance appraisal systems now in use has resulted in chaos, not flexibility. From agency to agency, and even division to division, the lack of standardization means that no one beyond the employee's own organization really knows how to interpret the results. The current trend to change performance appraisal systems is simply change for the sake of change.
U.S. Coast Guard
I agree with Mr. Light's position. Regarding Mr. Howard's comment about 360 degree feedback, I caution managers who are considering using it to use caution. There are differences of opinion as to whether they should be used for evaluation purposes. Many experts feel they should be used solely for developmental purposes. An excellent description of the issues and debate is contained in "Should 360-Degree Feedback Be Used Only For Developmental Purposes?" published by the Center for Creative Leadership .
Federal Energy Regulatory Commission
I agree with Mr. Light that the performance appraisal system that attempts to rank employees is often a meaningless task. Given this type of system, the pass/fail option is a superior alternative because less time is wasted in achieving the same result.
I believe that in order to have a truly effective system, employee performance needs to be aligned with organizational performance as outlined by the Government Performance and Results Act (GPRA). Further, designing a "new" system without an understanding of variation, systems thinking, psychology and theory of knowledge-principles Dr. W. Edwards Deming referred to as profound knowledge, will tend to lead to the same set of problems we are discussing in this forum.
Department of Defense
Author: Success Through Quality
(See Government Executive book review)
The old appraisal system created a lot of problems with people wanting to get an excellent or outstanding rating.
Often appraisals were structured to fit the needs of who should be promoted without reflecting the abilities of the employee performance.
The pass/fail appraisal [system] should stay. It a better system than we ever had. However, I do believe that a third tier of outstanding should be added. To achieve that tier, documented evidence needs to establish the reason why a person should get an award or promotion.
The appraisal system is misused under the current system of rating. The addition of a third tier would allow the employee to justify the reasons why the outstanding rating would be justified.
True, depending on the type of job a person would be able to compete for, we need to have a means and method that if a person meets the outstanding rating and puts in for another job that further consideration for that job needs to be addressed based on the criteria in the announcement.
What about those persons that fail? First of all we need to do joint labor management relations to establish when a person would fail and what needs to be accomplished to meet the passing grade in that position. The union must be made to understand that in order for any agency to deliver quality of service, that union-under partnership or not-has an equal responsibility to resolve problems at the lowest levels.
We need to work together to resolve what is best for the agency, but Congress and OPM seem to have other ideas.
Jim L. Jones
Social Security Administration
I can't believe that there can ever be a fair performance rating system.
There are always going to be personalities involved, be them good, bad, or indifferent, and there will always be a supervisor, and no matter how good or bad they are, or no matter how hard they try, there will always be favorites, and there will most certainly always be a whipping boy.
I would suggest doing away with all rating systems, and make it easier to fire any employee that does not perform their duties in a satisfactory manner. For the most part, we are all adults and have pride and self respect and shouldn't need an appraisal for doing a job. However, I do believe in rewarding an employee for performances above and beyond their work scope.
Thank you for letting me sound off.
Joe Chandler Sr.
USMC Logistics Base
Having been in management close to 20 years, I can attest to your assessment that the appraisal system is broken. To be perfectly honest, I'm not sure of the fix.
What you have failed to mention is the major impact that unions have played in this ongoing charade. Agreements negotiated with the NTEU representing Treasury employees, have been nothing short of comical in terms of truly seeking fair and accurate performance systems. Their real objective is to sabotage an already beleaguered system at any cost in order to gain an advantage for their constituency regardless of the cost. Try rating an employee where the current year rating is lowered (still favorable) and watch the fireworks!
I prefer the pass-fail system because it seems to be the least destructive of appraisal systems. The forced-ranking and capped-outstanding ideas are especially damaging to collaboration, which I thought was one of our most important goals.
I've been a civil servant over thirty years and have been through several performance appraisal systems. The first was as an apprentice. The grades really reflected how you got along with the foreman and/or journeyman you were assigned to. The second was the unsat/sat/outstanding system. I remember the foreman having us all form a line and as you reached the computer printout, you found your name with sat beside it and you initialed off, a painless approach for supervisor and employee.
The next evolution was the five step approach. This method was really tough because money was tied to it and you received 'guidance' in forms of percentages if you had too many outstandings or no satisfactories. During this time the satisfactory rating became something bad or low so the easy way out was to grade people highly successful.
My next experience was the PMRS (Performance Management and Recognition System) where you wrote down what you had accomplished that year and the supervisor graded you how he/she wanted to anyway. We're now using the pass/fail system. The pass/fail appraisal is further broken down into improvement needed, fulfilled requirements and significant strengths. I find that this is sufficient for a meaningful discussion between employee and supervisor. Much of the success or failure is in the hands of the supervisor. Many of them don't like to tell an employee where they need improvement.
I've found over the years that the employee actually finds it refreshing to hear the truth . . . normally they know where they can improve anyway. It's the frank discussion that can help the employee and supervisor. I think a combination of the 'write down what you accomplished this year' and the pass/fail would be the way to go. Finally, I'd be willing to go to a pay for performance as soon as the President, Congress and Senate allow me to grade them on their performance.