Centralized allocation of human resources. An application to public schools

• An extended CDEA that deals with non-transferable inputs is proposed.
• An iterative procedure capable of reallocating resources is designed.
• The overall efficiency and the optimal size of a network are determined.
• The results show that efficiency can be improved by redesigning the network.

Centralized resource allocation requires planning how to implement changes in order to adapt resources to the allocated budget without losing outputs. This paper presents an alternative model to reallocate human resources in a public education network based on the so-called centralized data envelopment analysis. This stream has received less attention in the efficiency literature and evaluates the overall efficiency of a set of decision making units controlled by a central authority. In this study, an extended centralized data envelopment analysis dealing with non-transferable inputs and environmental factors is proposed to assess the global efficiency of a centralized education network. We then design an iterative procedure capable of reallocating resources without jeopardizing the level of efficiency. The proposed model is applied in a real case of public schools from Catalonia (north-east Spain). The results indicate the network could be improved to optimally redistribute educational resources without falls in the level of aggregated outputs. The study also provides useful information for accountability and decision making when implementing improvement programs in public schools.