There are no doubt some occasions in which employees will believe it is fair for a supervisor to hold a particular co-worker to a lower performance standard-for instance, a colleague whose entire family was just killed in a car crash, or who lost a finger, or who has a learning disability, or who is brand new to the job. However, appraisal systems that permit evaluators to establish a separate baseline for each person being evaluated can invite perceptions of subjectivity and bias among those being evaluated. They certainly invite ferocious politicking and the possibility of corruption.
Besides worrying about perceived (and actual) justice in comparative evaluation schemes, you have to worry about the impact that such schemes can have on the behavior of those being evaluated. It can happen that some likely behavioral reactions:
1 – Politicking and possible attempts at corruption may occur at the time the criteria for evaluation are established or handicaps are employed. We know of no fool-proof response to this problem, but for several reasons it often helps to have criteria set and handicaps determined by a committee of supervisors:
a) it is usually harder to corrupt or influence a committee than an individual; b) individual supervisors are less likely to be capricious or biased when they have to defend their actions in a group; and c) committees can stiffen the backbone of individual supervisors, by giving them a “cover” or excuse for decisions that are made.
2 – Those evaluated will attempt to game the system, in terms of assignments they accept, tasks they perform, and so on. For instance, when a particular division of a firm is in trouble, it may be worthwhile for the firm to send one of their better managers to the rescue. But with ill-constructed incentive systems that engage in comparative evaluation using measures of divisional performance, asking a manager to take an assignment in a poorly performing division may be asking that manager to take a cut in compensation. The same is true when a professional is asked by a partnership to develop a new practice.
3 – When the comparison group lies entirely within a single organization, the group that is competing may collide against the evaluator’s interests. This sort of collusion is particularly likely when those who are in competition are socially similar to each other and different from their superiors, and when they are in close communication with one another. Stories of restrictive work-rate norms, enforced on the shop floor by social sanctions (or worse), and are known to most managers with a background in production. The somewhat pejorative term rate-buster was invented precisely to describe those who exceed the rate set informally by the workgroup.
4 – Even if the group doesn’t collide explicitly, conformity – a form of implicit collusion-may result. Individuals who are being evaluated may seek safety in numbers, especially when the scheme penalizes relatively poor performance more than it rewards relatively good performance. Complaints are often voiced, for example, that institutional money managers are more concerned with following what their counterparts are doing than with conducting their own fundamental analysis of stocks in the portfolios they manage. Such an outcome can sometimes be desirable-for instance, in guardian jobs, where “collective wisdom” is a good guide to acceptable performance. But in many cases this kind of conformity can substitute groupthink for individual judgment, impeding good performance.
5 – Cooperative efforts among those being comparatively evaluated may be hindered. Because individuals are being evaluated relative to one another, one person’s improvement comes at the expense of someone else in the same evaluation pool. Organizations relying on comparative evaluation schemes use a variety of approaches to deal with the potential problem of impeding cooperation among the employees who are being comparatively evaluated.
One tactic is to add a subjective assessment of employee cooperation to the measure of dmv handicap parking permit absolute or relative performance, to temper any dysfunctional tendency for employees to seek maximum performance at the expense of their co-workers. For instance, the large year-end bonuses for which the piece-rate employees compete are based on an assessment of their relative productivity, but bonus formulas also incorporate subjective assessments of workers’ cooperation, obtained from multiple managers within the plant. This approach can be a healthy corrective to the potential divisive effects of relative performance evaluation.
However, cooperation is often harder to measure than other dimensions of performance. This can make it very tricky to devise evaluation and reward schemes that will foster the desired balance between cooperation and “sticking to your knitting” to maximize individual output. Another means of trying to preserve cooperation among potential rivals in a relative performance evaluation tournament is to offer longer-term rewards for cooperation, such as promotions, which might offset any shorter-term incentive to withhold cooperation in order to fare well in the short run.
Relative performance schemes may also provide poor incentives near the end of the evaluation period. Think about a race or competition, in which the prize money is determined simply by what place you finish in. If you get very far behind or very far ahead of the rest of your competitors, there may not be strong incentives to perform at your highest level: If you are way ahead of the pack, you might simply wish to coast and protect your lead; if, on the other hand, you are dead last in the contest by a wide enough margin, you may conclude that there is no point in putting in any more effort. Yet obviously in many (if not most) settings, organizations will wish individuals to continue to strive to do their best, even if they are clearly destined to finish first or last in a relative performance contest. A good partial antidote to these problems is to avoid schemes that have the flavor of a pure tournament: To the one victor go all the spoils. Instead, rewards should be sensitive to how far in front or behind individuals are, so that finishing a “close third” yields better rewards than does a “distant third,” let alone a distant fifth.