On interim performance evaluations and interdependent period outcomes

Abstract

Performance feedback is an integral element of an accounting system, and firms provide this feedback at varying frequencies to their employees. This paper explicates the impact of an interim performance evaluation on the principal’s surplus using a dynamic two-period agency model. Two settings are discussed: single-purpose use, wherein accounting information is used solely for control purposes, and dual-purpose use, in which accounting information is used for production and control. Results demonstrate that the optimality of interim performance evaluations depends on the use of information and the interdependence of period outcomes. Furthermore, neither setting entails strict dominance with regard to carrying out interim evaluations or not. It implies that an interim evaluation can be optimal even if the optimal course of action does not depend on it. It further suggests that refraining from the interim review can be optimal even if that information is required to determine the optimal effort level.

On costless‐renegotiation proofing in binary agency models

“…16 16 Work by Lukas (2023), who uses a similar model setup, also suggests that the effect of stochastic interaction may not be obvious. He investigates whether frequent evaluation (performance evaluation in each period) or infrequent evaluation (performance evaluation once at the end of Period 2) is optimal.…”

“…However, after Period 1, the probability of a high outcome in Period 2 is independent from prior effort choices so that only second-period effort is relevant, as (27) and (28) show.14 One could easily extend the list. Setting g e2 ¼ b e2 creates the boundary case of stochastic independence, and it would be on a par with the special case of no complementarity mentioned in footnote 10.15 For purposes of exposition, the benefit of full commitment is slightly differently modeled than in sections before.16 Work byLukas (2023), who uses a similar model setup, also suggests that the effect of stochastic interaction may not be obvious. He investigates whether frequent evaluation (performance evaluation in each period) or infrequent evaluation (performance evaluation once at the end of Period 2) is optimal.…”