Measuring technical progress with data envelopment analysis

The purpose of this study is to develop a nonparametric method for measuring productivity growth of a group of operations over multiple time periods. The method applies data envelopment analysis (DEA) to estimate technical progress via efficiency evaluation through the construction of piecewise linear frontiers to approximate production surfaces pertaining to different time periods. A measurement of the productivity improvement that an individual operation can achieve is made by adjusting the operation to the pertinent production frontiers of different time periods. The paper contains the following sectors. Section 2 illustrates a basic DEA model and defines a production technology to lay the foundation for the discussion. We show how a practical piecewise linear frontier can be constructed to approximate the theoretic production surface. Section 3 extends the basic DEA model and formulates new models to evaluate the productivity of operations in different time periods. By referring to the average rate of productivity progress achieved by its peer group, an individual operation's pace in productivity change can be estimated. While Section 3 focuses on estimating total factor productivity growth, Section 4 gives a parallel discussion on the estimation of the productivity progress attributable to part of the factors. To show the tractability of our method, Section 5 presents an application using panel data. The last section concludes with some remarks concerning the implications of this study and directions for future research.