Refine
Document Type
- Working Paper (3)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
- Leistungsmessung (3) (remove)
Institute
- Wirtschaftswissenschaften (3) (remove)
Do we measure what we get?
(2008)
Performance measures shall enhance the performance of companies by directing the attention of decision makers towards the achievement of organizational goals. Therefore, goal congruence is regarded in literature as a major factor in the quality of such measures. As reality is affected by many variables, in practice one has tried to achieve a high degree of goal congruence by incorporating an increasing number of these variables into performance measures. However, a goal congruent measure does not lead automatically to superior decisions, because decision makers’ restricted cognitive abilities can counteract the intended effects. This paper addresses the interplay between goal congruence and complexity of performance measures considering cognitively-restricted decision makers. Two types of decision quality are derived which allow a differentiated view on the influence of this interplay on decision quality and learning. The simulation experiments based on this differentiation provide results which allow a critical reflection on costs and benefits of goal congruence and the assumptions regarding the goal congruence of incentive systems.
This paper studies a setting in which a risk averse agent must be motivated to work on two tasks: he (1) evaluates a new project and, if adopted, (2) manages it. While a performance measure which is informative of an agent´s action is typically valuable because it can be used to improve the risk sharing of the contract, this is not necessarily the case in this two-task setting. I provide a sufficient condition under which a performance measure that is informative of the second task is worthless for contracting despite the agent being risk averse. This shows that information content is a necessary but not a sufficient condition for a performance measure to be valuable.
When performance measures are used for evaluation purposes, agents have some incentives to learn how their actions affect these measures. We show that the use of imperfect performance measures can cause an agent to devote too many resources (too much effort) to acquiring information. Doing so can be costly to the principal because the agent can use information to game the performance measure to the detriment of the principal. We analyze the impact of endogenous information acquisition on the optimal incentive strength and the quality of the performance measure used.