The paper focuses on hierarchically structured organisations with a large set of operating units. While the central body in such organisations faces asymmetry of information concerning the operating costs of the units, it may wish to incentivise them through benchmarking and target setting to operate as efficiently as possible. If a standard Data Envelopment Analysis (DEA) approach is used for this purpose, each operating unit could estimate its own efficient targets. However, this decentralised scenario is not necessarily appropriate for a centralised organisation in which a central body wishes to optimise the performance of the system of units as a whole. On the other hand, a top-down imposed set of targets is often not suitable as they would be too demanding for some units and too lax for others. This paper proposes a DEA-based approach for incentivising the units of a hierarchically structured organisation in order to optimise the performance of the units collectively while at the same time the targets are not too demanding for inefficient units. The proposed approach is also extended so that incentive levels for operating units are determined over time, taking into account any changes in their productivity. Accordingly, the central management can strike a balance between not spending too much on incentives on the one hand and encouraging the operating units to reveal their true cost function on the other. We illustrate our approach using data from a set of German savings banks.
Bibliographical note© 2018, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/
Funding: Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – AH 90/5-2.
- Centralised Management
- Data Envelopment Analysis (DEA)
- Incentive Regulation
Afsharian, M., Ahn, H., & Thanassoulis, E. (2019). A Frontier-based System of Incentives for Units in Organisations with Varying Degrees of Decentralisation. European Journal of Operational Research, 275(1), 224-237. https://doi.org/10.1016/j.ejor.2018.11.036