CHAPTER 7
Managing Performance and Accountability
It wouldn’t surprise me if this were the section of the book that you are most interested in.1 After all, whenever I speak to government managers about improving performance, they all want to know how they can hold their employees accountable. To do this, I think it is best if we first think in terms of roles and responsibilities, second about the performance management system itself, and third in terms of implementation of the system.
With respect to roles, every government manager needs to provide his employees with the training, tools, and expectations required so that they will both know what they need to do to succeed and be able to meet the requirements of the job. In other words, it is up to each employee’s supervisor to put her in a position where she can properly perform her job. This means not only setting up the training and ensuring the employee has a decent work space, access to the computer system, a copy of the code of conduct, and appropriate performance standards but also addressing any problems that might develop along the way (e.g., running organizational interference if technical problems develop, providing the employee with additional training or guidance if they struggle in a particular area).
The employee is responsible for coming to work, learning as much as possible, and preparing to become a journeyman employee. Her job is not to be a passive participant in the training phase; it is to be an active player. Therefore, if she believes that she is not receiving proper training or the right tools, she needs to let management know what the problems are and try to work together to find an appropriate solution.
The Performance Appraisal System
Now that everyone’s roles are clear, let’s turn our attention to the performance appraisal system. There are two key components: (1) the employee’s performance plan, which consists of the employee’s performance standards; and (2) then administering the system. We’ll cover the second area in the next chapter. Let’s look at each in more detail.
The Employee’s Performance Plan
The employee’s performance plan is part of the government’s overall performance appraisal system. In order to understand the system and how the performance plan fits within it, there are some key definitions that you should be aware of.
Critical element means a work assignment or responsibility of such importance that unacceptable performance on the element would result in a determination that an employee’s overall performance is unacceptable. Such elements shall be used to measure performance only at the individual level.
Noncritical element means a dimension or aspect of individual, team, or organizational performance, exclusive of a critical element, that is used in assigning a summary level. Such elements may include, but are not limited to, objectives, goals, program plans, work plans, and other means of expressing expected performance.
Performance rating means the written, or otherwise recorded, appraisal of performance compared to the performance standard(s) for each critical and noncritical element on which there has been an opportunity to perform for the minimum period.
Performance standard means the management-approved expression of the performance threshold(s), requirement(s), or expectation(s) that must be met to be appraised at a particular level of performance. A performance standard may include, but is not limited to, quality, quantity, timeliness, and manner of performance.
Progress review means communicating with the employee about performance compared to the performance standards of critical and noncritical elements.2
Writing Performance Standards
The key component of the appraisal plan is the performance standards, since they serve as the basis for measuring the performance of the employees. First and foremost, the standards need to be aligned with your organizational goals so that everyone is concentrating on achieving the same objectives. If these are not aligned, people will focus on all sorts of things that may not be in the organization’s best interests. Moreover, you may even find yourself in the uncomfortable position of having most, if not all, of your employees far exceeding their standards even as the organization fails to achieve its goals. This can happen if your organizational energy is diffused.
To me, a good visual analogy is the Death Star that was featured in the Star Wars movies.3 The Death Star was a round, planet-size machine that was capable of destroying another planet. When the Death Star became operational, all of the energy forces in the machine came together in one focused laser beam that produced an enormous amount of focused energy and ultimately destroyed another planet.
Figure 7-1. Align Your Forces Like the Death Star
That is exactly what you want your performance standards to do: Ensure that all of your organizational energy is focused in a laser-like manner to achieve your organization’s performance goals. When that doesn’t happen, you wind up with a lot of wasted energy and employees working in an inefficient manner.
To provide a real-world example, I recall working with an organization that was responsible for conducting certain types of investigations. It was clear to me that senior management wanted to produce a high-quality product, yet it didn’t measure the individual quality of its investigators. When managers started looking at the overall quality of the work being performed on a national basis, they quickly discovered that the quality of the investigations varied greatly by employee, in part because the performance standards sent a message that quality was not important and that employees would not be held accountable for poor quality. As a result, there was an enormous amount of rework going on. Management learned from this experience that if it wanted to consistently deliver a quality product, it needed to measure each employee’s accuracy and ensure that everyone was focused on the same things.
Performance standards are very hard to write, and supervisors often have a difficult time objectively appraising their subordinates. What usually happens is that supervisors follow the path of least resistance and write standards that are so vague and/or so generic that (1) the employee doesn’t really know what is expected of him; and (2) there is no real way to hold the employee accountable for poor performance because the standards do not specify what the employee is supposed to do in order to be considered fully successful.
Let’s look at several examples of typical standards that I have seen and analyze how they play out in the workplace.
Element: Writes reports and management studies on topics assigned
Standard: Reports are accurate and well written and are generally submitted on a timely basis. Recommendations contained in the studies are based on sound logic and supported by the appropriate facts.
Analysis: This type of standard is so vague that it is virtually impossible to appraise someone against it. As a result, the supervisor simply uses her judgment to decide how the employee is doing. Although some supervisors probably like that since it provides them with in their minds maximum flexibility, it (1) is unfair to the employee, because he never really has anything to go by in determining how he is doing; (2) places the supervisor in an awkward position if she wants to take action against a poor employee; and (3) provides no real basis for making distinctions between employees. With this type of standard, little documentation about employee performance is ever maintained and no one, especially the supervisor, looks forward to the end of the appraisal period because everyone recognizes that each employee’s appraisal will be based more on gut than on fact.
Element: Processes fiscal transactions
Standard: Processes a reasonable number of transactions per day with acceptable quality. Transactions containing errors are corrected within forty-eight hours.
Analysis: This standard is extremely vague. For example, what is a reasonable number of transactions per day? What is acceptable quality? How is it measured? The only real metric here involves correcting transactions within forty-eight hours, but how will that be measured, and how long is the employee expected to take to process fiscal transactions?
Element: Cooperation and organizational support
Standard: Serves as a team player and assists others when required. No more than three legitimate complaints will be received regarding the way the employee works with others and supports the organization.
Analysis: To me, as written this standard is very imprecise, to a large extent measures conduct, not performance, and tracks exceptions rather than day-to-day interactions. Moreover, how does one determine what a legitimate complaint is, since there is no real benchmark?
How does one go about writing solid and actionable performance standards? In my view, you need to write them in a way that allows you to determine how well each employee is actually doing his job and be able to make meaningful distinctions between people (i.e., who is doing an outstanding job, who is doing a satisfactory job, and who is doing a poor job). A good way to tell if you are on the right track is to compare your employees’ appraisals relative to the performance standards and then see how they match your gut. What I mean is that most people know who the best and worst performers are in an organization. That is never a secret. If the standards you put in place generally result in the people you believe are your best receiving high ratings and the ones you think are the weakest receiving relatively low ratings, you are probably at least off to a good start.
Try to write the standards using the SMART principles that were described in Chapter 5 when I talked about the development of organizational metrics. The more your standards are specific, measurable, and attainable, the more actionable they will be and the more credibility they will have with the employees, which is very important.
Think of the overall performance plan and the standards themselves as part of a mini–balanced scorecard for each employee; in other words, they should be a system for measuring everyone’s overall contribution. To do this, try to incorporate metrics that, where possible, at least track each employee’s accuracy, output, timeliness, and customer satisfaction.
This, of course, is easier said than done, since you need to find a way to easily capture this information. I suspect that at least some of this information is probably in your computer system (e.g., individual output and/or timeliness may be easily measurable if each employee is responsible for a specific product or output). If not, you can either try to build fields for this in your system, assign each employee a specific code or number and capture the work that way, or have each employee record what he completes in a centralized database that is subject to supervisory verification.
Employee Output. With respect to employee output, you should count the number of widgets that each employee completes (assuming that the job in question can realistically be measured in this way). However, if the employee completes different types of outputs, you should consider assigning a weight to each type of work so that every employee’s output is adjusted by the degree of difficulty of the work (e.g., end product A might be worth two standard man-hours, two points, two credits, two cases, or some other common measure; end product B might be worth four standard man-hours or points, and so on). In this way you can compare “apples to apples”; that is, you can measure the true output of each employee who performs the same type of work, regardless of the mix of their work.
In setting up a system to measure employee productivity, you need to determine what outputs will be measured; what their weights are, if any; what time, if any, will be excluded from the measurement period;4 what is the minimum level of productivity that will be accepted; and how you will capture the data.
This is an extremely important and sensitive measure because if too much output is expected of the employees, they will find shortcuts for achieving the standard, and quality will suffer. Conversely, if too little productivity is expected, the organization’s overall performance will decline because you will not be efficiently using your available resources.
Timeliness. Concerning timeliness, you need to know how long it takes an employee to produce a specific product or deliver a particular service. After all, your customers expect you to serve them as quickly as possible, and timeliness is always an important measure of your organization’s overall delivery of service. That is why you should try to set up your computer system so that it can track work from the time it is assigned to an employee until the time it is completed. Be clear as to exactly what you will be measuring (e.g., from the beginning of a certain point in the process until the end of another point) and what the minimum acceptable level of timeliness is. In this way, there will be no surprises and no secrets, and you will know exactly how long it is taking each employee to complete their work.
Accuracy. Accuracy can surely be measured as long as you do a random sample of each employee’s work and retain the review sheets for your records. To do this properly, I recommend that you identify what will be measured for accuracy, the criteria that will be used, the amount of work to be reviewed, the frequency of the reviews, the manner in which an employee can contest a negative quality determination, and the minimum level of acceptable quality. I would also record this information in a database so that it can easily be retrieved and you can identify any patterns that require correction.
Customer Satisfaction. Very few government organizations measure individual employee customer satisfaction. To some extent, that makes sense, because you don’t want to conduct surveys of the public asking how each government employee is doing. First of all, the public generally doesn’t know what each government employee does; second, even if they did, it would be cost prohibitive to conduct such surveys.
That having been said, you may want to consider conducting surveys of stakeholders that frequently interact with your employees. This will provide a good sense of how each employee is dealing with people from outside the immediate organization. It will also help drive the right behavior because once employees know that the way they treat stakeholders will be evaluated, they will be much more likely to exhibit the right behavior, or pay the price with a lowered appraisal or adverse award determination.
The same approach could also be taken for the service that each employee provides to his internal customers. If an employee also knows that his internal customers will evaluate the way he works with them, he will be even more likely to act appropriately.
I suggest that you use a modified 360-degree review5 of each employee to measure the way he provides service to his customers. Such an approach provides a more in-depth and fair review of each employee’s overall performance, since many more people contribute to the evaluation than just the first-line supervisor. To me, this makes much more sense than basing the evaluation of the employee’s performance on merely the supervisor’s judgment and perhaps a few “valid complaints.”
Such an approach can be a bit tricky, however, if co-workers are asked to do the reviews. They may be uncomfortable participating in such a review if it can result in the subject of the review being fired or affect her pay. That is why I do not recommend including co-workers in this process.
Another approach is to make this element noncritical, meaning the employee cannot be fired for failure to meet the minimum expected level of performance. This may reduce the concerns that could arise from these reviews.
Regardless of the way you set up the standard(s) for this element, make sure that everyone understands the rules, the expectations, and how frequently such surveys will occur. Note I recommend that surveys be administered at least twice a year and be averaged or weighted accordingly in order to give each employee the opportunity to improve.
The elements I have described are not all-inclusive; you can certainly add other ones if they make sense for your organization. However, I am recommending that, where possible, you include the four categories I just described because they provide you with a fair and well-balanced view of each employee’s performance, which is one of the key goals of a performance appraisal system.
In terms of writing the actual standards, let’s take the three examples I provided earlier and see if we can make them more meaningful and precise and more consistent with the SMART principles.
Element: Writes reports and management studies on topics assigned
Standard: A minimum of 85 percent of reports are submitted by the due date. Reports are accurate at least 85 percent of the time. Any required changes are submitted within two days 90 percent of the time.
Analysis: This standard is much more accurate and precise than the earlier one and lets the employee know exactly what is required to meet the minimum level of acceptable performance. Now you might be thinking to yourself that this is all well and good, but how will I ever track all of these reports and be able to accurately recall what the employee’s performance was in this particular category? My response is that such a thought indicates you have a problem in terms of how you are tracking your employee’s performance. However, don’t feel too bad, because I believe most supervisors have the same problem.
In order to address that issue, I recommend you design a simple one-page review sheet (see Figure 7-2.) that provides the employee with written and timely feedback on each report submitted, which you can retain in your employee file and refer to whenever you need to provide performance feedback (e.g., on the annual appraisal, during the mid-year review, or during any other feedback session).
Element: Processes fiscal transactions
Standard: Processes at least sixty transactions a day with an accuracy rate of at least 92 percent. At least 85 percent of transactions are processed within five days of receipt, and 95 percent of erroneous transactions are processed within two business days.
Analysis: This standard is much tighter and clearer than the original example I provided. Note that the number of fiscal transactions processed per employee should be relatively easy to track through the computer system. Moreover, for the purpose of this example, I am assuming that most transactions are relatively similar in complexity, so I am not assigning a different weight to each type of transaction. Although a few transactions are probably more difficult than the others, I am also assuming that everyone has a relatively similar work mix, and that the number of sixty transactions per day takes that mix into account.
Figure 7-2. Employee Review Sheet
Regarding timeliness, that too should be relatively easy to track in the computer system by employee, especially if you set up those parameters beforehand. Concerning accuracy, using a review sheet like the one in Figure 7-2 (without the references to timeliness) should suffice to track the quality of each person’s work.
Element: Cooperation and organizational support
Standard: Serves as a team player and assists others when required. Employee receives a score of at least 75 or higher on his modified 360-degree review. No more than three legitimate complaints will be received regarding the way the employee works with others and supports the organization (a legitimate complaint is an allegation that the employee did not appropriately serve the team, another member of the organization, a customer, or a stakeholder in a manner that was consistent with the organization’s core values and one that has been verified as being accurate by a member of management).
Analysis: By incorporating the modified 360-degree review, you have put in place a reliable way to measure how well the employee has both cooperated with and supported the organization. Moreover, because it is the sum total of the observations of a number of people, it is much more difficult for the employee to allege supervisory bias. Finally, since you have defined what constitutes a legitimate complaint, everyone knows what the criteria are, making it much easier to hold the employee accountable.
There are a couple of other approaches you should be aware that will make your standards even better. First of all, when an element contains several performance standards (such as the examples I gave for writing reports or processing fiscal transactions), it is important to indicate how the employee will be rated if she meets some but not all of the standards. In other words, if she meets two out of the three standards (say, timeliness and productivity) but not the third (accuracy), has she met the element or failed it? You need to be clear about this point. Are any of the standards more important than the others? Be clear regarding this as well.
Second, note that I incorporated several key components (timeliness, productivity, accuracy) into the elements because, in my example, the employee would have had more than one distinct task (i.e., writing reports, processing fiscal transactions). However, if the employee has one primary task (e.g., processing claims, servicing loans, rehabilitating customers, repairing machines, making decisions), then you could have one element devoted to productivity, one to timeliness, one to accuracy.
The point here is that there is more than one way to do this. The key is to design your standards in a way that makes sense for each job, and the best way to do that is to make the standards as simple, measurable, fair, and accurate as you can.
Third, recognize that every job is not so easy to measure and that some components of certain jobs may be close to impossible to assess. For example, it might be pretty tough to precisely measure what a research scientist does on an annual basis. Be aware that there is no governmentwide requirement that each aspect of every job be measured numerically. Sometimes you have to use a more generic measure. However, where possible, it is generally better to have standards to which you can assign numbers.
Another approach that I strongly recommend is to articulate at least two levels of performance for each standard: the minimally acceptable level and the far-exceeds level. The reason I advocate this approach is that it lets everyone in advance know what the rules are, so when it comes time to distribute rewards people already know how they are doing and whether or not they will receive an award.
To be fair, some people will disagree with this approach because they feel it reduces their flexibility, and, to some extent, that is true. However, I think that is part of the problem and why I advocate a more rigid approach. In my experience, government supervisors often give awards to the people they like and not necessarily to the people who are making the most important contributions. Such an approach tends to undermine the employees’ faith in management and its decisions, because employees learn that you get rewarded on the basis of whom you know instead of on what you do.
By establishing in advance the criteria required to receive an award, you set up a methodology whereby people are reliably rewarded for excellent performance and employees whose performance is merely average or worse will not receive an award. That having been said, there will still be plenty of room for flexibility because every employee who far exceeds his standards doesn’t have to receive the same amount of money. However, I would ensure that if someone’s performance exceeds the standard by 50 percent, he should receive more than another person who exceeds it by 25 percent. In addition, management should retain the right to give out awards for special acts or contributions. The point here is that employees need to see a clear relationship between the performance appraisal process and rewards and recognition. There is more on this subject in Chapter 9.