Returns to scale in DEA models for performance evaluations

Abstract

The technology industry is a major driving force in Taiwan's economic development. The Taiwan government employed technology development programs (TDPs) to stimulate industrial technology research and development to enhance industry competitiveness. This paper uses Data Envelopment Analysis (DEA), a nonparametric approach, to evaluate the relative efficiency of decision-making units (DMUs) that use multiple inputs to produce multiple outputs, to evaluate the relative managerial efficiency of TDPs. The inputs are human resources and expenditures. The outputs are patents and technology outcomes. We investigated TDP performance over the period from 1999 to 2003. This study uses CCR and BCC models, which are DEA model variants, to calculate efficiency indexes. CCR is adopted under the assumption of constant returns to scale. BCC is used to understand the variable returns to scale, including the constant, decreasing and increasing. We found that TDPs in the material and chemical engineering fields in 1999 and 2001, machinery and aerospace fields in 1999, and communication and optoelectronics fields in 2002 had better performance than the other TDPs. We provide potential improvements for inefficient DMUs.