Comparative Study Of Two TFP Calculation Method Of The DEA

Since the reform and opening up, China’s economy has experienced more than30years of rapid growth. However, in recent years more and more economists think thisgrowth which is mainly by increasing investment is “extensive mode”. There are alsostudies have shown that, since the late20th century, China’s total factor productivity(TFP) showing a clear downward trend. Along with the deepening of China’seconomic reform, the continuous improvement of industrial restructuring andupgrading, as well as the imbalance in the regional development of the of China’seconomy phenomenon, in-depth exploration of the quality of China’s economicgrowth, TFP research deep into the industrial sector from the total economy, theregional level, will have an increasingly important significance.In reality, almost all production processes are involving intermediate inputs, andintermediate inputs takes a high percentage in total output. There are various methodsfor calculation of TFP, and all based on the production function estimates. Even if anon-parametric method which is not related to a specific production function form,will select input and output indicators based on the production function. For anindustry, there need a certain condition of exist the value-added production in theory. Thus, there is the choice of the selection of value-added production function and totaloutput production function when estimating the TFP of a industry or sector.Based on the existing research of the impact of intermediate inputs on economicgrowth and TFP, in this thesis, the data envelopment analysis (DEA) is selected. Werespectively based on the value-added production function and total output productionfunction to estimate the TFP. Use China’s inter-provincial panel data from1997to2007, respectively estimated for28inter-provincial autonomous regions in Chinaduring the period examined three industries TFP and its decomposition index. Andfrom the country, the region of eastern, central and western, inter-provincial differentlevels of the calculation results of the two methods conducted a comprehensivecomparative analysis. The different and common characteristics are summarized. Inthis thesis, the impact of the intermediate inputs on industry TFP is studied from theperspective of empirical analysis. At the same time, TFP trend of the reference periodof three industries in China were analyzed.Main conclusions of this thesis are as follows:Firstly, there is level promotion effect of intermediate inputs on industry TFP. TheTFP based on value-added production function exist underestimate. Theunderestimate ratio of the three industries are11.2%，6.7%and9.2%respectivelyduring the examined period.Secondly, results based on the gross output production function which consideredintermediate inputs is better stability. The results showed that, the TFP and itsdecomposition based on gross output production function of the volatility of the indexis less than the results based on the value-added production function.Thirdly, there is not trend effect of intermediate inputs on industry TFP. The TFPindex based on the two production function is highly correlated by the change in trend, namely intermediate inputs does not affect the trend the TFP of dynamic time.Fourthly, during the examined period, the two methods have consistently shownthat the technical efficiency contribution to TFP is greater than that of technologicalprogress of China’s primary industry, and scale efficiency is greater than the puretechnical efficiency; the technology progress contribution to TFP is greater than thatof technical efficiency, pure technical efficiency is greater than the scale efficiency ofChina’s secondary industry; the scale efficiency is greater than pure technicalefficiency of China’s tertiary industry, and the technical efficiency contribution to TFPis greater than that of technological progress based on the value-added productionfunction, while based on the results of the gross output production function is just theopposite.Fifthly, during the examined period, the two methods have consistently shown thatthe TFP index of the three industries of China in decreasing order is: the secondaryindustry, the tertiary industry, and the primary industry. The technical efficiency of thethree industries showed a downward trend, which is the important reason to pull downthe TFP level during the examined period.Sixthly, during the examined period, the three industries TFP of the two methodsshow that the TFP index of Eastern is higher than that of the Central and Westernregions. The three regions all show that the scale efficiency is greater than the puretechnical efficiency; the technical efficiency index is greater than technologicalprogress index of the primary industry of the Eastern, the Central and Western areopposite; the technology index is greater than the technical efficiency of the Easternpart of the second industrial, while the pure technical efficiency is greater than thescale of technical efficiency, the Central and Western are opposite.Lastly, dividing the time periods as1997-2002and2002-2007two periods,compared with the previous five years, the TFP of primary industry growth lowermore than half of the provinces, and the Central region reduced the highest proportion;the TFP of the second industry growth lower about half of the provinces; the TFP ofthe tertiary industry showing increased more than half of the provinces, and theEastern provinces increased the highest proportion while the Western provincesincreased the lowest proportion, namely, the Eastern, Central and Western regionincreasing as the ladder-state potential of the tertiary industry TFP growth.