1. Introduction

Since China embarked on economic reform and adopted the opening-up policy in the late 1970s, It has achieved a high economic growth for several decades. The key features of Chinese economic development are capital-driven and labour-driven, such as large-scale inward foreign capital and migration of peasant labour to the cities. Neoclassical growth theory suggests that apart from contributions by capital and labour, technological progress also plays an increasingly important role in economic growth. As for China, globalisation, marketization and infrastructure improvement are often regarded as main causes of the rapid economic growth. Therefore, empirical evidence is of interest to determine which factors, among those oft-cited factors, are crucial to Chinse economic growth.

In this paper, based on provincial panel data, a combination of a nonparametric approach of data-envelopment analysis (DEA) Malmquist index and a panel data model is applied to get the empirical results to determine the key factors. Section II reviews existing literature, while section III presents the DEA Malmquist model for Total Factor Productivity (TFP) estimation and the empirical panel models related to the economic growth. In the economic growth model, labour, capital, TFP, FDI stock, marketization, globalisation and infrastructure are used as explanatory variables in the model. Section IV estimates provincial TFP series using the DEA Malmquist, and discusses Chinese technological growth. Section V provides the empirical findings from the growth model. Finally, section VI concludes with a discussion on policy implications.

2. Literature Review

Some authors point out that China’s remarkable performance is mainly attributed to factor accumulations of capital and labour ^{[20, 22, 29]}; productivity growth, represented by total factor productivity, is low. Young ^[26] finds that TFP growth in China’s non-agricultural economy is 1.4 per cent per year. Zheng’s ^[29] estimation, based on UNIDO productivity database, shows that the average annual rate of change in TFP during the period 1962 to 2000 was 0.5 per cent, with its contribution to economic growth only 7.9%. Since TFP is sensitive to estimation approaches, to avoid any ex ante assumptions for estimation, Färe et al. ^[6], Coelli (1996) and Fu ^[9] used a nonparametric approach of data-envelopment analysis, to estimate TFP.

An important question arises from these findings. Is productivity progress in host countries promoted by foreign direct investment (FDI)? Evidence of FDI technology spillovers is found in empirical studies by Kokko et al. ^[15], Blomstrom and Kokko ^[3] as well as in case studies by Larrain, Lopez-Calva and Rodriguez-Clare ^[16]; Moran ^{[18, 19]}. Empirical verification by Hale and Long ^[13] reveals positive FDI technology spillovers for technologically advanced native firms, but find no, or negative, spillovers for backward ones. Focusing on countries other than China, some researchers found positive technology spillover effects from FDI for developed countries ^{[10, 17]}, while others found insignificant or negative results for less developed countries ^{[1, 11, 23]}.

With regard to institutional factors and economic regime, the most outstanding changes in China are the introduction of market mechanism and the opening-up policy ^[29]. Market reforms in China are characterized by the increasing non-state-owned share of the economy. Hence, in the study of Zhang (2001), the degree of market reform is indexed by a ratio of non-state employees to total employed labour force. Besides those factors considered above, Démurger ^[5], Fu et al. ^[8] and Fleisher et al. ^[7] regarded public infrastructure as a determinant of economic growth and technological progress in China.

The purpose of this study is to verify the impacts of those factors on Chinese economic growth with empirical results, by creating a model that combine all those factors in one framework.

3. Methodology

Productivity is not present in statistical sources and it is a key factor used in the evaluation of the quality of economic growth. Following Färe et al. ^[6], Coelli (1996) and Fu ^[9], TFP is estimated using the DEA approach. The formula for calculating the output-oriented Malmquist index is shown in equation (1). This non-parametric method has following advantages: The decision-making unit (DMU) can be technically inefficient, the form of production function may be unknown, and neutral technical changes are not necessary. Productivity changes can be decomposed into efficiency and technical changes, denoted as effch and techch, respectively, in equation (1). effch equals the ratio outside the brackets.

(1)

For a DMU i in period t, can be estimated using the following linear programming under constant returns to scale:

(2)

where x, y, K and M represent the input variables, the output variables, number of input and output variables, respectively. is a scalar and is a N × 1 vector of constants. N denotes the number of DMUs. X and Y are N × 1 vector of x and y, respectively. can be estimated by changing the period subscripts of x and y from t to t + 1 and maintaining the period subscripts of X and Y to t. The formulas for and can be obtained by similar adjustments. By adding a restriction of , the results for variable returns to scale (VRS) are obtained, otherwise constant returns to scale (CRS) is assumed. DEA Malmquist model is implemented with a DEA Toolbox created by Álvarez et al. ^[2].

Based on Solow growth model ^[24], the growth of an economy can be attributed to the growth of labour, capital and, in particular, technological progress. Therefore, these three factors are included in our model. Other factors, such as the introduction of a market mechanism and the opening-up policy are also considered according to the study of Zheng ^[29]. Based on existing literature, which are presented in previous section, FDI stock and infrastructure are incorporated in the model to assess the impacts of foreign investment and infrastructure investment. Thus, a model in Cobb-Douglas form is constructed and shown in the following equation.

(3)

In equation (3), the period is denoted by t and the subscript i represents different provinces in the country. L represent labours, K stands for capital stock, MTFP is the DEA Malmquist estimate of TFP, FDIS is the stock value of FDI, M represents marketization, G denotes globalisation and Infr is an abbreviation for infrastructure. All variables are at a provincial level and vary across provinces and time. and stand for the output elasticities of labour force and capital stock, respectively. The exponents of other factors are their associated output elasticities. The output elasticity of each input factor is not limited at quantity. Hence, constant returns to scale are not assumed.

Value of marketization is generated by calculating the ratio of non-state employees to total employed labour force, following the method of Zhang (2001). Globalisation index uses total international trade, i.e., import and export volume to represent the connection with the global economy. The length of the regional highway network is used as a proxy for public infrastructure development. Capital stock is calculated, based on the perpetual inventory method, which is demonstrated in Equation (4).

(4)

In Equation (4), K is the capital stock, I denotes gross capital formation, is the depreciation rate and i, t indexes, regions and periods, respectively. According to Zhang et al. ^[28], for China, takes the value of 9.6%. Capital stocks in the initial year are also taken from Zhang et al. ^[28]. The labour force is represented by the number of persons employed at the end of the year.

Applying logarithm on both sides of equation (3) and inserting an error term , the econometric equation specified in equation (5) is obtained for the growth model. stands for the individual effect and hence is allowed to vary across provinces and it also represents the initial provincial levels of economic growth and catches any effects of omitted time-invariant and cross-section fixed variables.

(5)

4. TFP Estimation and Discussion on Technological Growth

The cumulative growth rates of Malmquist productivity index (MTFP), technical changes (TC) and technical efficiency changes (TEC) estimated with DEA Malmquist are presented in Figure 1. The base year for the cumulative growth rates is 1990. To embark on a new, efficient and sustainable development pattern, the Government highlighted technological and innovative industries in recent years. This is reflected in the TFP growth rates generated from Equation (1) and (2), which reaches a cumulative growth rate of 52.9% in 2013.

Download as

• PPT
  PowerPoint Slide
• PNG
  Larger image(png format)

Veiw figureFigures index

•  NEW
  View larger figure in new window

View next figure

Figure 1. Cumulative growth rates of TFP, TC and TEC estimated with DEA Malmquist

Figure 1 shows that the technical-change component in the Malmquist index exceeded 1 in 2002, and the efficiency-change component of Malmquist index stayed around 1. The change in the Malmquist productivity index, which is comprised of technical-change and efficiency-change components, rises approximately simultaneously with the technical-change index, and both reached a cumulative growth rate around 50% in 2013. This indicates that productivity growth is achieved mainly through technical progress, and efficiency change does not contribute much to the productivity growth. This is similar to the conclusions by Zheng and Hu ^[30]. The empirical results for TFP components suggest that technical efficiency should be improved in China.

The crossing point in 2002 in Figure 1. indicates that in this year, Chinese economy switched from capital-driven and labour-driven to technology-driven. Productivity increased quickly and positively after this year. The contribution of technology to economic growth has become significantly since this year and this will be verified in the next section.

5. Regression Results from the Growth Model

Data used in the growth model are from provincial Statistical Yearbooks, 1991-2014. MTFP is calculated in accordance with the DEA Malmquist methods mentioned above. The summary statistics for those variables are shown in Table 1, and the regression results of the growth model are given in Table 1. Meanings of those variables can be found in Equation (3) and (5). Based on a Hausman test, fixed effects are used for the panel data.

Table 1. Summary statistics of variables

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View next table

Table 2. Regression results from the growth model

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View previous table

Labour (L), capital (K) and technology are found to have significant and large effects on Chinese economic growth. The elasticity of K and L are 0.606 and 0.473, which means 1% increase of K and L will induce 0.606% and 0.473% increases of GDP, respectively. The coefficients of other explanatory variables have similar meanings. The ratio of the elasticity of K to the elasticity of L is found to be around 6:4. This is contrary to the general benchmark of 4:6 ^[21]. Solow (1957) also finds that United States’ output elasticity of K is around 0.35, which indicates that the elasticity of L is even larger than 0.6 in some industrialized countries. However, other researchers in China also found similar elasticities. Chow ^[4] and Zheng and Hu ^[30] used a value of 0.40 for the output elasticity of labour. The labour shares estimated by Hu and Khan ^[14] were 0.386 and 0.453 during the pre-reform and reform periods, respectively. Zhang and Shi ^[27] obtained 0.391 for the elasticity of L. All these results suggest that capital formation in China plays a key role in economic growth. Calculations based on this output elasticity show that the contribution of capital stock is very high. Annual study shows that in 1991, capital formation accounted for some 68.7%growth in China and this percentage stayed above 50% between 1991 and 1999.

The capital-driven feature of Chinese economic growth is induced by increasing foreign investments attracted by the booming economy and by the quick domestic capital accumulation in Chinese economy. During the period 1991-2013, the average growth rate of fixed capital formation was 13.7%. The detailed growth rates for this period are shown in Figure 2. In 1995 and 2004, the growth rates of capital investment are over 19%. Economic growth, which depends heavily on physical capital investment, could cause overheating of the economy, and intensive Government-motivated investments could probably crowd out private investments.

Download as

• PPT
  PowerPoint Slide
• PNG
  Larger image(png format)

Veiw figureFigures index

•  NEW
  View larger figure in new window
• PREV
  View previous figure

Figure 2. Fixed capital formation growth rate in China

Technological progress, represented by the productivity progress, has a significant and large impact on the economic growth. The quantity of its coefficient (0.54), i.e., the output elasticity of technology, is between those of the capital (0.61) and labour (0.47). This suggests that technological progress play a key role in Chinese economy, and it is consistent with the results of positive TFP trend in previous section.

The coefficients of marketization, globalisation and FDI stock are insignificant in Table 2. This indicates that regardless the origins of the investment, from foreign countries, from state-owned firms or from private-owned firms, and regardless of target markets of products, to domestic market or to international one, the economic activities that can promote productivity, increase capital or labour involved in the production, can generate positive effects on the growth.

The coefficient of infrastructure is significant and this result reveals a crucial feature of Chinese development. As China is a big country and transportation costs are a key factor considered by most investors, most local governments pay great attention to the construction of public infrastructure to attract investments and accelerate the development of economy. In addition, public infrastructure can also solve problems caused by unequal distribution of natural and human resources, integrate cities in a well-connected area.

6. Conclusion

In this paper, we analyse the effects of capital, labour, technology, globalisation, marketization and infrastructure on economic growth, based on a DEA approach and a panel data model. Our empirical results indicate that three factors, capital, labour and technology, are crucial and significant for Chinese economic growth. Chinese economic growth in the past years was mainly capital-driven, with a largest output elasticity of 0.61 and high growth rates of capital formation. Technology, represented by productivity estimated with a DEA Malmquist approach, has a second large output elasticity of 0.54, which is larger than the output elasticity of labour (0.47). This implies that while economic growth in China in the past was capital-driven and labour-driven, it is currently becoming more technology-driven. This trend is important for a sustainable development of China. The significant coefficient of labour is partially due to the migration of rural workers from agricultural sectors to industrial sectors and the improvement of labour quality.

The technical change is a major and significant contributor to the productivity growth, especially after 2002. Another component of the productivity, technical efficiency change, lacks progress during the focal period. This means that efficiency improvement has larger potential in future development of China. To avoid the inflation and economic bubbles caused by the high quantity of investment, and to achieve a more sustainable growth, it is suggested that future growth should rely more on technological progress, especially on efficiency improvement of technology.

Although marketization, globalization and FDI do not present direct significant effects on economic growth, the contributions of these factors should not be ignored. The investment from foreign countries, state-owned or non-state-owned enterprises and the international trade with other countries are all highly correlated with the capital formation, employment and productivity improvement. In other words, they have significant indirect effects on the growth.

Infrastructure construction appears to have positive effects on the regional development. With regard to transportation costs, natural and human resource disparities in China, infrastructure construction is a major approach for mitigating such costs and disparities. On the other hand, once public infrastructure becomes saturated, focus of public construction should be moved to the investment in indigenous R&D investment to ensure more innovative and technical progress.

Acknowledgements

The work on this paper has been supported by the Collaborative Innovation Center for 21st-Century Maritime Silk Road Studies and the Foundation for High-level Talents in Higher Education of Guangdong.

References

[1]	Aitken, B. and Harrison, A. E, “Does domestic firms benefit from direct foreign investment? Evidence from Venezula”, The American Economic Review, 89(3). 605-618.Mar.1999.
	In article		View Article
[2]	Álvarez, I. C., Barbero, J. and Zofío, J. L, “A Data Envelopment Analysis Toolbox for MATLAB”, Working Papers in Economic Theory 2016/3, Department of Economics, Universidad Autónoma de Madrid, Spain, 2016.
	In article
[3]	Blomstrom, M. and Kokko, A, How foreign investment affects host countries, World Bank PRD Working Paper, No. 1745, Washington, D.C, 1997.
	In article
[4]	Chow, G, “Capital formation and economic growth in China”, Quarterly Journal of Economics, 108(3). 809-842.Mar.1993.
	In article		View Article
[5]	Démurger, S, “Infrastructure development and economic growth: An explanation for regional disparities in China”, Journal of Comparative Economics, 29(1). 95-117.Jan.2001.
	In article		View Article
[6]	Färe, R., Grosskopf, S., Norris, M., and Zhang, Z, “Productivity growth, technical progress, and efficiency change in industrialized countries”, American Economic Review, 84(1). 66-83.Jan.1994.
	In article
[7]	Fleisher, B., Li, H. and Zhao, M. Q, “Human capital, economic growth, and regional inequality in China”, IZA Discussion Papers, IZA DP No. 2703, 2007.
	In article
[8]	Fu, F., Vijverberg, C. C. and Vijverberg, W. P. M, “Public infrastructure as a determinant of intertemporal and interregional productive performance in China”, IZA Discussion Paper, No. 1019, Feb.2004.
	In article
[9]	Fu, X, “Exports, technical progress and productivity growth in Chinese manufacturing industries”, Applied Economics, 37(7). 725-739. Jul.2005.
	In article		View Article
[10]	Globerman, S, “Foreign direct investment and ‘spillover’ efficiency benefits in Canadian manufacturing industries”, Canadian Journal of Economics, 12(1).42-56.Jan.1997.
	In article
[11]	Haddad, M. and Harrison, A, “Are there positive spillovers from direct foreign investment? Evidence from panel data for Morocco”, Journal of Development Economics, 42(1). 51-74. Jan.1993.
	In article		View Article
[12]	Hale G. and Long, C, “FDI spillovers and firm ownership in China: Labor markets and backward Linkages”, Federal Reserve Bank of San Francisco Working Paper Series, 2006-25, 2006a.
	In article
[13]	Hale, G. and Long, C, “What determines technological spillovers of foreign direct investment: Evidence from China”, Yale University, Economic Growth Center Discussion Paper, No. 934, 2006b.
	In article
[14]	Hu, Z. L., and Khan, M. S,”Why is China growing so fast?”, IMF Staff Papers, International Monetary Fund, Washington, DC,1997.
	In article
[15]	Kokko, A., Tansini, R. and Zejan, M.C, “Local technological capability and productivity spillovers from FDI in the Uruguayan manufacturing sector”, Journal of Development Studies, 32(4). 602-611.Apr.1996.
	In article		View Article
[16]	Larrain, B. F., Lopez-Calva, L. F., and Rodriguez-Clare, A, “Intel: A case study of FDI in Central America”, CID Working Paper, No. 58, Harvard University, 2000.
	In article
[17]	Liu, X., Siler, P., Wang, C. and Wei Y, “Productivity spillovers from foreign direct investment: evidence from UK industry level panel data”, Journal of International Business Studies, 31(3), 407-425. Mar.2000.
	In article		View Article
[18]	Moran, T, Foreign direct investment and development, Institute for International Economics, Washington, D.C, 1998.
	In article
[19]	Moran, T, Parental supervision: The new paradigm for foreign direct investment and development, Institute for International Economics, Washington, DC, 2001.
	In article
[20]	Ozyurt, S, Total factor productivity growth in Chinese industry, 1952-2005. LAMETA, Universtiy of Montpellier Working Papers, no. 07-13, 2007.
	In article
[21]	Romer, P. M, “The origins of endogenous growth”, Journal of Economic Perspectives, 8(1). 3-22.Jan.1994.
	In article		View Article
[22]	Sachs, J. D. and Woo, W. T, “Understanding China’s economic performance”, NBER Working Paper Series, no 5935, 1997.
	In article
[23]	Singh, R, “Government introduced price distortion and growth, evidence from twenty nine developing countries”, Public Choice, 73(1). 83-99. Jan.1992.
	In article		View Article
[24]	Solow, R. M, “A contribution to the theory of economic growth”, Quarterly Journal of Economics, 70 (1).65–94.Jan.1956.
	In article
[25]	Solow, R. M, “Technical change and the aggregate production function”, The Review of Economics and Statistics, 39(3). 312-320. Aug.1957.
	In article		View Article
[26]	Young, A, “Gold into base metals: Productivity growth in the People’s Republic of China during the reform period”, Journal of Political Economy, 111(6). 1220-1261. Jun.2003.
	In article		View Article
[27]	Zhang, J. and Shi, S, “Change of total factor productivity in Chinese economy: 1952-1998”, World Economic Forum (Chinese), (2): 17-24. Feb.2003.
	In article
[28]	Zhang, J., Wu, G. and Zhang J, “The estimation of China’s provincial capital stock: 1952 -2000”, Economic Research Journal (Chinese), (10). 35-44. Oct.2004.
	In article
[29]	Zheng, Y, “Productivity performance in developing countries: Country case studies of People’s Republic of China”, Research paper of UNIDO on World Productivity Database, 2005.
	In article
[30]	Zheng, J. and Hu, A, “An empirical analysis of provincial productivity in China (1979-2001)”, Working Paper in Economics (SwoPEc), No.127, 2004.
	In article