Total Factor Productivity Analysis Based On Malmquist Index And Directional Technology Distance Function

Since Caves etc present Malmquist index in order to analysing Total Factor Productivity in 1982, nonpapametric Malmquist productivity index attracted more attention. With research work developing the sources of TFP have enjoyed a great deal of interest among researchers. So a series of models have been proposed. Also a lot of demonstration analyses were applied to models, the illustrations provided some cases. However on one side all of models make the same assumption: supposed some scale returns. And computing Malmquist index and decomposing factors based on supposed scale returns. On the other side the traditional Malmquist index models are based on input or output distance function proposed by Shephard. In which models either input or output change and decision making unit approaching production frontier by a radial direction, so the two limitations result in imprecise measurement. Chambers, Chung and Fare presented directional technology distance function method in 1996 and 1998. The method combines input function with output function, considers input and output changing at the same time and decision making unit approaching production frontier with a nonradial way. Therefore this paper focuses on presenting modified Malmquist index model based on directional technology distance function after analyzing Malmquist index methods either in China or abroad synthetically. Firstly identifying decision making unit's scale returns based on directional technology distance function so as to solving traditionalscale returns assumed and making measure more precise. Secondly after identifying scale returns presenting 3 Malmquist index models based on directional technology distance function: Malmquist index defined on t period, Malmquist index defined on t+1 period and IM Malmquist index. And measuring factors which affect TFP defined on t period: technical change and technical efficiency change. Then making illustration of models according to properties of directional technology distance function. Thirdly defining decomposing factors of cost Malmquist index on directional technology distance function. It is more important to figure out allcative efficiency. Combined with it Malmquist index methodology becomes more perfect. Finally applying model in this paper to demonstration analyses of TFP and affecting factors about 19 APEC countries. The demonstration analysis is mainly about Asian countries during two decades 1980-1990 and 1999-2000. Especially compared the East Asian Newly Industrialized economies(NIEs) with the Association of South East Asian Nations(ASEAN), the NIEs"s productivity growth are higher than ASEAN during 1980-1990 and result are completely reverse during 1990-2000.