Bridging radial and non-radial measures of efficiency in DEA
- 405 Downloads
Data envelopment analysis (DEA) has been utilized worldwide for measuring efficiencies of banks, telecommunications, electric utilities and so forth. Yet, the existing models have some well-known shortcomings that limit their usefulness. In DEA we have two fundamental approaches to measuring efficiency with very different characteristics; radial and non-radial. We demonstrate a method for linking these two approaches in a unified framework called Connected-SBM (slacks-based measure). It includes two scalar parameters, and by changing the parameter values we can relocate the analysis anywhere between the radial and the non-radial models. An appropriate choice of these parameters can overcome the key shortcomings inherent in the two approaches, namely, proportionality and mixed patterns of slacks.
KeywordsEfficiency DEA Radial and non-radial slacks
Unable to display preview. Download preview PDF.
- Cooper, W. W., Seiford, L. M., & Tone, K. (2007). Data envelopment analysis: a comprehensive text with models, applications, references and DEA-solver software (2nd ed.). Berlin: Springer. Google Scholar
- Tsutsui, M., & Tone, K. (2007). Separation of uncontrollable factors and time shift effects from DEA scores. GRIPS Discussion Paper 07-09, GRIPS Policy Information Center. Google Scholar
- Tsutsui, M., & Tone, K. (2008). Comparison of adjusted management efficiency of vertically integrated electric power companies between Japan and the US. Journal of Socio-Economic Research, 56, 113–130. (In Japanese.) Google Scholar