Performance Appraisal 

Why Ratings Based Appraisals Fail 

Summary: Rating employee performance is the most common method of recording and evaluating employees. Unfortunately, while it's easier and less time consuming than other methods, it is almost useless in helping employees actually improve their performance. Here's the scoop on the limitations of ratings based appraisals.

---

In the January edition of The Public Sector Manager / Workplace2001 newsletters, we discussed why the use of RANKING procedures to compare employees to each other can create disastrous consequences. Fortunately, ranking systems for performance appraisal aren't used that frequently compared to the use of RATING systems. Unfortunately, RATING systems are also problematic, and are used in a huge number of organizations. 

First, what is the difference between a ranking and a rating system? A ranking system evaluates employees based on whether they are better, equal or worse than their peers. It is a comparison. A rating system compares employee performance to some set of criterion, and produces either a number or a letter grade that supposedly represents the employee's level of performance. With a RANKING system it isn't really possible for everyone to be ranked as excellent (or at the top of the heap), even if all employees are excellent. A rating system permits everyone to be rated highly, if they warrant it. 

Rating systems are so popular that computer programs have been developed to undertake the evaluations. In addition, most 360 evaluation processes are based on ratings systems, with the ratings obtained by not just the supervisor, but peers, customers, etc. 

The question is whether they "work". 

Problem One: Appearance of Objectivity 

In our organizations we have legal and philosophical pressure to evaluate employees in an objective, consistent and fair way. There is no question that being objective is critical. Because of our desire to conform to those needs, we create systems that use numerical scales (for example 1-5) to evaluate employees. As an example, the University of California uses the following rating categories (in addition to some other components):   

• Job knowledge: Evaluate the use of information, procedures...etc required for current jobs.
• Quality: Evaluate the accuracy, completeness, etc of work.
• Planning/Organization: Consider areas such as varying work demands, developing efficient measures,...
• Initiative: Consider the self-starting ability, resourcefulness, and creativity applied to the duties of the position.

If you look carefully at the criteria above, you will see that they don't eliminate subjective judgements at all. One manager's idea of "self-starting ability" can be quite different than another's idea. How does one objectively evaluate "creativity". 

This wouldn't be a major problem except that often we act as if the ratings ARE objective. We make pay and promotion decisions on information that is at best quite subjective. We forget that any rating is only an indication of how one person (the manager) applies a fuzzy criterion. In terms of legal consequences, a poorly and badly designed set of criterion is probably not sufficient to protect an employer. Dismissing an employee based on, let's say, a low ranking on creativity is going to be really problematic unless one can justify that rating in terms of hard, concrete events (failed to create a new product between January and December). But if we use the criterion above, we don't need rankings. 

Ratings systems give people a false sense of security, protection and objectivity. 

Problem Two: Development Issues 

One function of performance appraisals is to help employees develop so they can contribute more effectively. Do rating systems, in and of themselves, contribute to employee development? The answer is No. 

In order for staff to develop and learn they need to know what they need to change, where (specifically) they have fallen short, and what they need to do. If a manager assigns a 1 (unsatisfactory) on a scale of 5 to the dependability criterion, what information does that convey (by itself) to the employee? Not much. It just says the manager is dissatisfied with something. In order to make it meaningful and promote growth, far more information must be added to the appraisal process. When were they undependable? In what very specific ways? What changes need to occur? Those are the critical growth questions. 

One argument offered by ratings proponents that the manager can use the rating scale as a springboard to discuss those details. That's true. However, why do the ratings. A manager dealing with an employee who is habitually late can simply document the lateness, and discuss with the employee what needs to be done to remediate the problem. No numbers, and no very rough, subjective categories. 

Simply put, ratings, on their own, do not convey sufficient information for people to improve. And since they don't do that, why use them? 

Problem Three: Fairness Issues 

If, as we suggest, ratings systems are too subjective (but appear objective) and ratings do not help employees get better, there are some serious problems from the position of the employee. 

First, since the criteria for ratings are so often loose, most employees are going to resist being classified at the low end of the scale. Employees who are low rated are more likely to resist the subjective evaluation of the boss, argue, claim personal vendettas, etc. Simply put, they are easy to argue with, just because performance compared to vague criteria are unmeasurable. So, the manager says performance is unsatisfactory (1) and the employee believes it is excellent (5). Where do you go from there. 

It is far more sensible to eliminate the rating completely, and use critical incident reports or firm, measurable objectives where there is less possibility for interpretation. Which is less likely to cause resistance on the part of an employee? Telling someone you think they rate an unsatisfactory classification for dependability, or providing employees with an attendance sheet that documents that they were late eight times in the month? 

Why Is Rating So Popular? 

If ratings are not objective, are not needed to promote employee development (and productivity), and create friction and argument in the workplace, why are they so popular? 

The answer is simple. Organizations can use a common, "one-size-fits-all" form that can be administered quickly and easily. It doesn't cost as much as an Management by Objectives approach which has potential for providing objectivity and the perception of fairness. It also doesn't provide objectivity and fairness. 

It's cheap and it doesn't take a whole lot of time. Or is that really true? In a short-term perspective it IS true. However if a rating system doesn't help people do better, are there costs that are incurred as a result of having such a system? Probably. A poor system is expensive later. In legal issues, grievances, and the cost of performance problems that are not addressed using a rating system. 

Final Comments 

In a short article we can't complete a fuller picture of all the issues. We invite those interested to order our white paper entitled "Performance Management - Why Doesn't It Work" for a more detailed, but focused discussion. In closing let's consider some of the following regarding performance appraisal. 

1. Many organizations report that once a person's salary is no longer tied to doing rating type appraisals, they cease to be done. The reason: Badly implemented systems cost too much in terms of time, money and discomfort on both staff and management sides. 

2. 360 Appraisals (rankings from multiple sources) are worse than regular manager-employee rating systems. They create more subjective data, with rankings from one source contradicting ratings from another. Hugely expensive.   

3. Where rating systems appear to be succeeding (and the value they add is not usually assessed), they work IN SPITE of the ratings. A good manager can make a rating system work. A poor manager who relies solely on the ranking system is going to do more damage with it than if they did nothing at all. Conflict, bad feelings and argument are going to occur.  

Next month we will be completing this series by addressing the question: 

If ranking systems aren't good, and rating systems aren't good, how do we do performance management?