Alternative Approaches to Estimate Returns to Scale in DEA- Models

This paper presents a number of approaches to estimate returns to scale properties of general multiple-input muliple-output technologies. Both qualitative scale efficiency-properties and quantitative scale elasticity-properties are covered. The estimation approaches considered are based on data envelopment analysis (DEA)-models using primal production data on inputs and outputs. We propose an approach to estimate scale elasticity based on the primal efficiency problem. This alternative approach is derived as a direct approximation of the differentials of the efficiency measures in the scale elasticity definition. The different approaches to estimate scale efficiency and scale elasticity are exemplified and compared using a small dataset.