Multi-period data envelopment analysis models and resource allocation: A case study

Abstract

Data envelopment analysis is a general tool for efficiency and performance measurement of the set of decision making units based on solving linear optimization problems. The paper deals with using this group of models for resource allocation analysis among Czech economic faculties within 7 years period from 2008 until 2014. The resources in the period t are divided among faculties according to their teaching and research performance in the period (t - 1). The teaching performance is measured by the number of students and number of graduated in the given period. The research performance is defined according to the national rules by the number of so called RIV points. The overall performance level in the given period is estimated using data envelopment analysis model with constant returns to scale assumption. The input in the period t is partly based on the efficiency or super-efficiency scores from the preceding period. Super-efficiency score given by Andersen and Petersen model is used for discrimination among efficient units. The results of the allocation in the last period are compared to the ones calculated to the standard methodology. Numerical experiments are realized using own MS Excel add-ins and LINGO modelling and optimization system.