Notice: This HTML version is intended only to facilitate reading. If you want to quote the paper, please use PDF version, which contains page numbers.


The Industrial Development of Western China in the First 5 years of Western Development Program

 

Ke Zhang , School of Public Policy & Management, Tsinghua University

 

 

  1. Introduction

Since 1999, China launched the program of western development and began to implement a series of development plans in 2000.  As the economic base, industrial sectors are one of the development focuses of the program. An important policy measure in the program is to increase investment.  As we know, investment, especially government investment, has been  the main source maintaining China’s high speed growth for a long time. Investment growth causes fast capital accumulation and then scale expansion of industries. However, capital productivity decides the impact of investment on promoting economic growth. Investing in low productivity sectors will not only waste capital, but also accumulate economic risk. Recently, China has realized that high inputs and relative low income are not a recipe for a sustainable growth mode. It’s important to improve industries’ productivity to grow more efficiently.

            Productivity is a measure of the efficiency of production:  the quantity or value of output in relation to a quantity or cost of input. Productivity is the economic measure of the level of technology and management and is the source of long-term competitive advantages. Any competition or industrial policy tries to promote productivity and the policy effect of the program is partially reflected in the growth and efficiency changes of local sectors. In order to keep sustainable development, western China should try to improve local industries’ productivity.

            This paper sheds light on the growth and productivity changes of the industry mix in the western region in order to assess the policy effect of the western development program. According to neoclassical economic growth theory, input factors, such as capital, labor, etc. and Total Factor Productivity (TFP) are the direct resources of growth. This paper applied the basic analytical framework in industrial sector level. Capital growth is used to show scale expanding. TFP is looked at as a measure of potential competitiveness. High and low TFP distinguish two different growth modes. In order to find the change of western industrial sectors, shift-share analysis is employed, which tells the difference between growth rate of local sectors and national ones. Based on data envelopment analysis, the productivity change of each sector is analyzed and compared.  This reflects the basic competitiveness of sectors and can predict the development potential. This paper also discusses the technological efficiency of local sectors and tries to explain the growth mode. The general finding is that most industrial sectors of western regions show productivity growth but the efficiency improvement is relatively slower than the technical frontier’s progress.

 

2. Methodology and data

2.1 Shift-share analysis basics

Shift-share analysis is one of the widely used techniques for examining regional growth and comparisons. The shift-share model decomposes economic change in a region into three additive components: the reference area component, the industry mix and the regional share. The decomposed variable may be income, capital stock, employment, value added, or a variety of other measurements.

The reference area component generally refers to the national economy and is so called the national share (NS), which measures the regional target variable change if it grew at the same rate as the nation. Capital stock is used as the target variable in this paper. The industrial mix (IM) is the industry composition of the region and refers to the capital stock change due to the difference between the growth rate of a particular sector and that of the whole national industry. Thus it reflects the degree to which the region specializes in sectors that are fast or slow growing nationally. The regional share (RS) reflects the regional impact on a particular sector, which is measured by the capital stock change due to the difference between the regional growth rate and the national growth rate of the industry. The total shift (TS) is the sum of the three components, reflecting the changing economic position of the region relative to the nation. These components are formulated as:

 

NS≡Kirgn                                                   (1)

IM≡Kir(gin-gn)                                                 (2)

RS≡Kir(gir-gin)                                                 (3)

TS≡NS+IM+RS                                               (4)

where the subscript i refers to industry i, r refers to region r, and n refers to the nation. Kir is the capital stock, and gir,gin, gn are the change rate of capital stock. This paper identifies the sectors which grow faster at western regional level than at national level through calculating local industries’ RS to see if RS>0 or not. These fast growing sectors are the main growth sources of the western region, whose level of productivity reflects the growth mode of such region. Total productivity increase generates efficient growth. However productivity decrease generates inefficient growth.

 

2.2 Data envelopment analysis (DEA) basics

The term of efficiency is about the utility of resources. It expresses the degree to which units best utilize their available resources to obtain maximum potential output. The modern efficiency measurement begins with Farrell (1957) who defined a simple measure of firm efficiency which could account for multiple inputs. The basic idea of the measure is to identify the production or cost frontier and quantify the difference between the unit to be examined and the frontier. The production or cost frontier reflects the maximum output given input or minimum input given output. Generally, data envelopment analysis (DEA) is a principal method to estimate frontiers which involves mathematical programming without assumed frontier function form. Fare, Grosskopf and Lovell (1994) discussed details of DEA methods.

 

            A relatively simple presentation of the DEA model is[1]:

Max.                                    (5)

s.t.   

                      

where xik and yrk are the i-th input and r-th output of the k-th decision making unit (DMU) respectively.  u and v are the weights that DEA will optimally assign to each input and output (Dinc and Haynes 1999). This paper employs DEA and Malmquist index approach to measure the locally fast growing sector’s total factors productivity (TFP) changes.

 

 

2.3 Data

This paper deals with inter-provincial sectors’ data, including all twelve western provinces: Ningxia, Shannxi, Gansu, Qinghai, Xinjiang, Inner Mongolia, Tibet, Sichuan, Chongqing, Yunnan, Guangxi, and Guizhou. Because of the accessibility and consistency of data, the analysis covers the change of selected sectors from 1999, the beginning of the western development strategy, to 2003. Twenty-three sectors (GB/T4754-2002 two digital classification code)[2] are analyzed, including part of manufacturing, energy, mining, metallurgy and chemical engineering industry, etc.

The data are from Yearbook of Industrial Economy Statistics edited by National Bureau of Statistics of China. Output is measured by sector’s value added. Inputs include capital stock measured by the sum of the average annual balance of capital and current assets and labor measured by each sector’s annual average employment.

 

 Table 1: List of 23 sectors

Code of sectors

Name of sectors

06

coal mining & preparation

07

petroleum & natural gas extraction

08

ferrous metals mining & preparation

09

nonferrous metals mining & preparation

13

food processing

14

food manufacturing

15

beverage manufacturing

17

textile manufacturing

22

paper making & manufactured goods

25

petroleum processing & coking

26

chemical materials & products manufacturing

27

medical & pharmaceutical products

28

chemical fibers

31

non-metal mineral products

32

smelting & processing of ferrous metals

33

smelting & processing of nonferrous metals

34

metal products

35

universal machine manufacturing

36

special purpose equipment & machinery manufacturing

37

transportation equipment manufacturing

39

electric equipment and machinery manufacturing

40

computer, electronic & telecommunications equipment manufacturing

44

power generation, steam and hot water production and supply

 

3. Analysis

3.1 fast growing sectors

Because investment is one of the main development measures, we use capital stock as the measure of  each sector’s scale. Based on equation 3, we calculated RS of every sector in all twelve western provinces. The outcome is reported in table 2 where each column represents a province and each row represents a sector. Table 2 shows that since 1999, every western province has several sectors expanding faster at provincial level than at national level. And, relatively, northwest provinces have more fast growing sectors than southwest provinces. It seems that the western region grows rapidly during the implementing of a western development strategy. On average, a western province has eight fast growing sectors, covering one third of the twenty-three sectors.

3.2 productivity changes

We have identified the fast growing sectors of each western province. The province with most of its fast growing sectors experiencing productivity increase is growing efficiently. Following Fare et al (1994), DEA is employed to measure TFP of each fast growing local sector, and output-based Malmquist index is employed to measure annual TFP changes of each fast growing local sector. Here we take capital stock and labor both into account as inputs because quantity of employees is an important input factor affecting productivity and traditionally Chinese enterprises employ more labor than the efficient level. Controlling the quantity of employees at an efficient level is a goal of economic reform of Chinese enterprises. When measuring scale of a sector, labor can be ignored because it doesn’t reflect the real level. But when measuring productivity or efficiency of a sector, labor must be considered.

 

 

 

 

 

 

 

 

 

Table 2: fast growing sectors in western provinces

code

province abbreviation

Northwest provinces

Southwest provinces

IM

XJ

QH

GS

NX

SX

TB

SC

CQ

YN

GX

GZ

06

 

 

 

 

 

 

 

 

 

07

 

 

 

 

 

 

 

 

08

 

 

 

 

 

09

 

 

 

 

 

 

 

 

13

 

 

 

 

 

 

 

14

 

 

 

 

 

 

 

15

 

 

 

 

 

17

 

 

 

 

 

 

 

 

 

 

22

 

 

 

 

 

 

 

 

25

 

 

 

 

 

26

 

 

 

 

 

 

 

 

27

 

 

 

 

28

 

 

 

 

 

 

 

 

 

 

31

 

 

 

32

 

 

 

 

 

 

 

33

 

 

 

 

 

 

34

 

 

 

 

 

 

 

 

 

 

35

 

 

 

 

 

 

 

 

 

 

 

36

 

 

 

 

 

 

37

 

 

 

 

 

 

 

 

 

 

39

 

 

 

 

 

 

 

 

 

40

 

 

 

 

 

 

 

 

 

 

 

 

44

 

 

 

 

 

 

 

 

Note: The following abbreviations means, IM-Inner Mongolia; XJ-Xinjiang; QH-Qinghai; GS-Gansu; NX-Ningxia; SX-Shannxi; TB-Tibet; SC-Sichuan; CQ-Chongqing; YN-Yunnan; GX-Guangxi; GZ-Guizhou

 

We calculated the means of annual TFP changes and reported the outcome in table 3. If TFP stays unchanged, the index equals 1. If TFP increased, the Malmquist index is more than 1, otherwise less than 1. Table 3 shows that about 90% fast growing sectors have indices more than 1, experiencing TFP increase. This is significant for the western region because it reflects that its growth is efficient.

 

 

Table 3: Means of annual TFP change index of fast growing sectors in western provinces

code

province abbreviation

Northwest provinces

Southwest provinces

IM

XJ

QH

GS

NX

SX

TB

SC

CQ

YN

GX

GZ

06

 

 

 

1.013

 

1.118

 

 

 

 

 

1.083

07

1.104

 

 

1.066

 

1.268

 

 

1.103

 

 

 

08

1.174

1.043

1.314

0.808

 

1.26

1.113

 

1.14

 

 

 

09

 

 

0.988

1.004

 

1.183

 

 

 

 

 

1.145

13

1.237

 

1.215

1.091

 

1.205

 

1.229

 

 

 

 

14

1.206

1.345

 

 

 

 

1.332

1.161

 

 

 

 

15

 

0.899

0.843

0.935

0.995

1.148

 

1.022

 

1.071

 

 

17

1.16

 

 

 

2.279

 

 

 

 

 

 

 

22

1.075

 

 

 

1.16

 

 

 

 

1.135

1.052

 

25

1.155

1.455

 

1.32

0.985

1.327

 

1.262

 

1.132

 

 

26

 

1.169

1.028

 

 

 

 

1.122

 

1.079

 

 

27

1.213

 

1.139

 

1.132

1.029

0.85

 

 

1.075

1.057

1.007

28

 

 

 

1.27

 

 

 

1.183

 

 

 

 

31

1.172

1.082

1.103

0.991

1.116

1.054

 

1.084

1.163

 

 

0.997

32

1.166

1.379

1.08

 

 

 

 

 

 

1.199

1.223

 

33

1.254

 

 

 

1.073

1.239

 

1.165

 

1.179

 

1.19

34

 

 

 

 

 

 

 

 

1.088

 

 

1.114

35

 

 

 

 

0.979

 

 

 

 

 

 

 

36

 

 

 

1.116

 

1.142

 

1.236

1.032

 

1.27

 

37

 

 

 

 

 

 

 

1.66

1.44

 

 

 

39

 

 

 

0.986

1.2

 

 

 

1.185

 

 

 

40

 

 

 

 

 

 

 

 

 

 

 

 

44

 

0.938

 

0.815

1.076

 

 

 

 

 

1.011

1.093

 

Note: The following abbreviations means, IM-Inner Mongolia; XJ-Xinjiang; QH-Qinghai; GS-Gansu; NX-Ningxia; SX-Shannxi; TB-Tibet; SC-Sichuan; CQ-Chongqing; YN-Yunnan; GX-Guangxi; GZ-Guizhou

 

Following Fare et al (1994), we divided the TFP change index into both an efficiency change index and a technology change index. This will indicate that a local sector’s TFP change comes from two factors. One is the most advanced local sector’s technical progress, which represents the technical frontier’s movement. Another is the local sector’s technical efficiency improvement, which represents the achievement of such a sector in “catching up” to the most advanced one. Figures 1 and 2 reflect that local sectors’ TFP changes show different modes. Forty-four percent of the local sectors with TFP increases experienced technical efficiency decreases in the circumstance of technical frontier progress. We can conclude, therefore, that these sectors’ efficiency improvements are relatively slower than their most advanced peer’s technical progress. They still have the potential to make their management and technical learning or innovation more efficient, even though their TFP has increased. But such a lag in progress is a potential threat to their long-term TFP increase and, thus, sustainable growth. Among local sectors with TFP decrease, 80% show an efficiency decrease while their most advanced peers improved technology. Inefficient management or technical progress is the main reason for the TFP decrease.



We also compared the referenced regions with an index of the national mean level and of the relatively developed province level, including Beijing, Shanghai, Jiangsu, Zhejiang, Guangdong, Fujian, and Shandong. Table 4 reported the outcome, from which we can find that the TFP change index of western provinces’ sectors is almost the national mean level. Also, it’s clear that Beijing, Shanghai and Guangdong have few advantages among these twenty-three sectors, which shows the difference of industrial structure between western provinces and Beijing, Shanghai, and Guangdong. But Shandong, Fujian, Jiangsu and Zhejiang have more fast growing sectors respectively than western provinces, although the TFP change index is similar. It seems that western provinces still develop a little more slowly than Shandong, Fujian, Jiangsu and Zhejiang.


          Table 4: Means of annual TFP change index of reference provinces

code

Provinces abbreviation

BJ

SH

JS

ZJ

GD

FJ

SD

NM

06

 

 

 

 

 

1.124

1.147

1.15

07

 

 

0.923

 

1.313

 

1.119

1.116

08

 

 

1.337

 

 

 

 

1.16

09

 

 

 

 

1.224

1.145

1.154

1.11

13

 

 

 

 

 

1.043

1.144

1.185

14

 

 

 

 

 

1.067

1.11

1.168

15

 

 

 

1.071

 

 

1.06

1.062

17

 

 

1.122

1.119

1.054

1.09

1.154

1.161

22

 

 

1.299

1.102

1.081

1.083

1.201

1.13

25

 

1.304

1.218

1.146

 

 

 

1.206

26

 

1.116

1.145

1.071

1.213

1.236

1.21

1.162

27

1.124

 

1.175

1.118

 

1.082

1.137

1.08

28

 

 

1.108

1.183

 

1.046

1.169

1.125

31

 

 

 

1.147

 

1.165

1.138

1.142

32

1.193

 

1.251

1.267

1.254

 

1.228

1.253

33

 

 

 

 

 

1.154

1.247

1.206

34

 

 

1.141

1.144

 

1.132

1.12

1.134

35

 

1.127

1.147

1.12

1.133

1.242

1.199

1.195

36

 

 

 

1.149

1.146

1.214

 

1.185

37

 

 

1.116

1.092

 

1.409

1.236

1.201

39

 

 

1.125

1.101

 

1.156

1.217

1.174

40

 

1.032

1.158

1.077

 

1.03

1.212

1.218

44

 

0.984

 

 

 

1.001

0.932

1.008

 

Note: The following abbreviations means, BJ-Beijing; SH-Shanghai; JS-Jiangsu; ZJ-Zhejiang; GD-Guangdong; FJ-Fujian; SD-Shandong; NM-National mean

 

 

4. Conclusion

Twenty-three industrial sectors in 12 western provinces constitute a total of 276 local sectors, among which 101 local sectors developed faster than the national level in terms of capital stock. Or, in other words, 37% of the total local sectors in 12 western provinces evidenced rapid development in investment growth. Through DEA, we find that in 90% of the 101 fast growing local sectors’ TFP has increased from 1999 to 2003. And 40% of these sectors with TFP increases showed efficiency improvements when their most advanced peers realized technical innovation. This is critical for western provinces since these sectors showed strong evidence in improving efficiency and, thus, narrowing the gap between them and their advanced peers. But, at the same time, the western region must realize that about 50% of 101 fast growing local sectors experienced efficiency decreases, which is the main reason for TFP decrease and the main weakness of 44% of sectors with TFP increases. Such sectors should pay more attention in improving management level and building their ability in technical innovation while sustaining their growth.


Reference

Coelli, Tim, Perelman Sergio. “A Comparison of Parametric and Non-parametric Distance Functions: With Application to European Railways”. CREPP Discussion Paper. 1996.

Fare, Rolf, Shawna Grosskopf, Mary Norris and Zhongyang Zhang. “Productivity Growth, Technical Progress and Efficiency Change in Industrialized Countries”. American Economic Review. Vol. 84, No.1 (March 1994), 66-83.

Fare, Rolf, Shawna Grosskopf and C.A.Knox Lovell. Production Frontiers. Cambridge University Press, 1994.

Farrel, Michael. “The Measurement of Productivity Efficiency”. Journal of the Royal Statistical Society. Series A (General), Vol. 120, No. 3 (1957), 253-290.

Haynes, Kingsley, Mustafa Dinc. “Productivity Change in Manufacturing Regions: A Multifactor/Shift-share Approach”. Growth and Change. Vol. 28 (Spring 1997), 201-222.

Dinc, Mustafa, Kingsley Haynes. “Sources of Regional Inefficiency, An Integrated Shift-Share, Data Envelopment Analysis and Input-output Approach”. The Annals of Regional Science. Vol. 33, issue 4 (1999), 469-489.

Rigby, D. L. and W.P. Anderson. “Employment Change, Growth and Productivity in Canadian Manufacturing: An Extension and Application of Shift-share Analysis”. Canadian Journal of Regional Science XVI:1 (1993)

Stimson, Robert, Roger Stough and Brian Roberts. Regional economic development : analysis and planning strategy. New York : Springer, 2002.

National Bureau of Statistics of China. Yearbook of industrial economy statistics. 1998- 2004. Beijing: China Statistic Press.

 


Appendix

Malmquist productivity index,

                                      (1)

Decomposition of Malmquist index,

                          (2)

The first term of right side of equation 2 is efficiency changes and the second term is technical changes.

To calculate equation 1 we must calculate the four component distance functions, which will involve four LP problems. Here we begin with assuming CRS technology and the output-orientated LP used to calculateis as following,

                         subject to

                                                                                               (3)

 

The remaining three LP problems are simple variants of this one,

       subject to

                                                                                     (4)

and

 subject to

                                                                                        (5)                      

 

The above approach can be extended by decomposing the (CRS) technical efficiency change into scale efficiency and pure (VRS) technical efficiency components. This will involve calculating two additional LP problems, which are repeating LP’s (2) and (3) with the convexity restriction () added to each. Then we can use the CRS and VRS values to calculate the scale efficiency residually.

 

 



  Ke Zhang is a PhD Candidate in the School of Public Policy & Management, Tsinghua University, Beijing. The author wishes to acknowledge the help of reviewer.

[1] Mustafa Dinc, Kingsley Haynes. “Sources of Regional Inefficiency, An Integrated Shift-Share, Data Envelopment Analysis and Input-output Approach”. The Annals of Regional Science. Vol. 33, issue 4 (1999), 473.

[2] The code and name of all twenty-three sectors are listed in table 1.

 

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.

(c) Washington Institute of China Studies

We encourage you to visit CEOpedia - The Management Online