The basic agency model contends that incentives lack power and efficiency when the employee has little control over the measures on which his compensation is based, or alternatively when the employee can control certain measures of performance but those measures give a noisy indication of the value-adding efforts provided by the employee. If a division manager is rewarded on the basis of how well his division does-measured by the division's earnings-and if those earnings are largely outside of the control of the manager, then little will be gained from the incentive system. The division manager might, on the other hand, have a lot of control over the cost of manufacturing goods his division produces. But a lower cost of manufacture may not necessarily correlate with higher corporate profits; indeed, actions taken to lower cost-of-manufacture-for instance, reducing the variety or quality of output, refusing rush orders-may mean lower corporate profits. This, essentially, is the multitask problem in slight disguise.
The impact of uncontrollable environmental variables on tangible measures of performance can sometimes be controlled by comparative or relative evaluation. To provide incentive compensation for a division manager, we might measure performance of the division relative to other divisions in the firm or other units in other firms that compete in the same basic market. Or we might compensate a salesperson based on how its sells compare with those of other comparable salesmen in the organization. The idea is that although the division manager may be unable to control her absolute level of performance, because she can't control economy wide variables, those same variables affect the performance of her "peers," and her division's relative performance gives a better idea how well she herself did. The salesperson may have had a very strong year because of the economy as a whole, or because her firm has introduced a great new product; because these same factors affect the entire sales staff, the relative performance of a particular salesperson gives a better indication how she her-self did. When performance is measured relative to others outside one's own organization, the term benchmarking is used. When compensation depends on how well A work relative to fellow employees B, C, and D, we say that the firm is using a tournament compensation scheme; although to be precise, in a true tournament, A's compensation will depend solely on her ordinal ranking relative to B, C, and D. leia todo o artigo
