9. Growth, Capital Formation and Technological Choice: Explanation of Long-Term Factors for Recent Decline of Economic Growth in China

Naughton (1995), McKinnon (1994), Zhang (1997).

Sachs and Woo (1994), Lin et al. (1994).

This article does not intend to discuss the statistical problems of China's economic growth rate. Interested readers can refer to the existing research literature. Rawski, for example, has done a series of studies on recent declines in economic growth (www.​pitt.​edu/​~tgrawski/​). He even reexamined China’s GDP growth rate in 1997 and 1998, adjusting the growth rate of the official data below 5.7% (Gary et al. 2000, Rawski 2001). In addition, in recent years there have been many research articles published on the reliability of China's official GDP growth rate data since the reform. For example, some economists found that in 1978–1991, the average growth rate of GDP of China was overvalued by 1 percentage points, while in 1992–1997, it was overvalued by 2.5 percentage points (Meng and Xiaolu 2000).

For the process of industrialization in East Asia and the comparative analysis of rural enterprises, see the fourth part of this paper.

Young (2000), 4; Chen (1999), Zhang (1999), Zhang (2001), Lo (1999).

It should be noted that, according to information provided by Easterly and Fischer (1994), due to the long-term results of epitaxial growth, the capital-output ratio during 1950–1987 before the disintegration of the Soviet Union had an annual increase of about 2.5%.

Solow (1962), Dension (1967).

Zhang and Han (1996) provided a literature review of the empirical research on the mode of economic growth in the Soviet Union. See Zhang and Han (1996).

(1992).

Radelet and Sachs (1997).

To know more about the main literature on the empirical study of the changes in the total factor productivity of China's industrial sector since the reform, refer to the literature review provided by Zhang and Shi (2000). Moreover, among the recently published books, O Keijiro, Liu Deqiang and Murakami Naoki made an empirical study on the efficiency changes of state-owned and township enterprises and a comparison of the relative efficiency. In fact, in the second chapter of their work they also provided an overview of relevant literature. See Zhang and Shi (2003), Keijiro et al. (2000), Chen et al. (1988), Jefferson et al. (1992, 1996).

He (1992).

It should be pointed out that our estimation results are strikingly similar to those of Li and others (1992). See Li et al. (1992).

We also noted that in 1977–1988, the average TFP growth rate was about 4.1%, and the contribution of TFP growth to output growth was about 41.9%. That was the golden period of the growth of TFP in Chinese economy. In 12 years, TFP maintained a positive growth.

Jefferson and Zheng (1995).

Murakami et al. (1995, 1995).

In 1999, for example, GDP grew by 7.1%, of which 4.3 percentage points were contributed by industrial growth (Macroeconomic Research Team of the Institute of Economics of the Chinese Academy of Social Sciences, 2000). Therefore, to observe and explain the growth of GDP, we can focus mainly on the growth of industrial added value.

For the account and sources of these price index as well as data, see “Data Appendix” of this article.

Table 9.3 is quoted from Chen (1999). But it needs to be reminded that since the author did not mention the source of the data and how the data was processed in his paper, his data may be the nominal value not adjusted by price. Of course, based on our calculations in Table 9.1, we know that the use of nominal values does not actually affect the validity of his basic conclusions.

Zhang (2000).

Byrd and Zhu (1990).

Ostuka (1996).

Otsuka (1998).

If the definition of large enterprises was changed from the original value of fixed assets of 100 million yuan to 50 million yuan, the number of large enterprises in the textile and machinery industries would have soared to 471 and 491. See State Bureau of Statistics: Chinese Industrial Economic Statistical Yearbook (1992), Beijing, China Statistics Press, 1992.

We know that, according to the new classical theory of the production function and by using the framework of section two, we can conclude that the growth rate of labor productivity can be broken down into growth rate of capital-labor ratio and weighted sum of TFP growth rate.

In calculating the actual labor productivity, we used the industrial net output value index provided by Lo (1999). See Data Appendix of this article.

For example, the World Bank’s survey (Byrd 1992) and Chen (1999) provided many typical cases about the rise of township enterprises. As they pointed out, in the early days, township enterprises were even established through informal channels such as “personal relations” with state-owned enterprises. See Byrd (1992).

Patrick and Rohlen 1987.

This data was calculated from the basic data provided in Table 9.1 (1999) by Chen Jianbo.

Ranis (1995, Chinn 1979).

Mead (1984).

Hayami (1998).

In the literature of development economics, this simple and traditional product of rural industrial production is called “Z- goods”.

Ranis and Stewart 1993.

Watanabe (1979).

Dollar and Sokoloff (1990).

Zhang (2001) made theoretical discussions and empirical tests on the scale of China's industries and economies of scale. We find that the higher profitability shown by China's big business sector does not seem to be the result of economies of scale, but is more likely to be a “sector effect”. Because the most skilled labor force and engineering personnel in China's economy are concentrated in those sectors, they have higher skills in learning and mastering technoglies than those of small and medium-sized enterprises.

Hallagan and Zhang (1998), Zhang and Halagan (1998).

This figure was indirectly obtained from the World Bank's research publications. See Byrd and Lin (1990).

The Fudan University’s Economic Research Centerconducted a large survey on the “horizontal cooperation” between the state-owned enterprises in Shanghai and township enterprises in other parts of the country in 1987, which provided us with substantial information. This information is worth careful analysis because it can help us understand the complex motivations and ways of cooperation between township and state-owned enterprises. See the report of the survey: Enterprise Reform and the Development of A New Road: A Collection of Investigation Reports of Horizontal Cooperation of Industrial Enterprises in Shanghai.

With regard to sub-sectors, in 1990, the number of township enterprises entering the ferrous and non-ferrous metal smelting and rolling processing industries were 5,648 and 3,839; the number entering transportation equipment manufacturing industry and electronic and communication equipment manufacturing industry were 7,603 and 3,506. In the iron and steel industry dominated by the state-owned enterprises with the scale of increasing returns, township enterprises also occupy a considerable proportion. These township enterprises mainly produce the final products (steel products). The Chinese Statistical Yearbook (1995) shows that in 1994, the number of township enterprises producing steel reached 17,455, providing 26% of the total output value of iron and steel industry. Sabin (1987).

Hallagan and Zhang (1996) developed a theoretical model of “excessive entry” for welfare analysis by which they discussed the welfare consequences of excessive entry. Recently, Young(2000), in his statistical analysis on the similarity of regional production structure in China, also seems to support this view.

Weitzman and Xu (1996), Nee (1992).

In fact, time mode of the entire industrial sector is similar to that of the state-owned sector losses, see Zhang (1998).

Data comes from the Data Appendix of this article.

Liang et al. (2000).

Strictly speaking, almost all explanatory variables in our model are not exogenous variables, but endogenous ones. However, it is difficult to construct simultaneous models because of the lack of proper variable definitions. We can only assume that these variables are predetermined and/or determined by institutions.

Another possibility is that, because the capital stock is the nominal value, therefore, the estimated value of coefficient of the scale of township enterprise is positive, which, to a certain extent, also reflects the fact that the newer and more expensive machines have a higher quality.

So, there is obviously a lack of empirial basis to attribute the continued decline of the Chinese economic growth rate in recent years to the decrease of total demand (among which consumption demand and investment demand accounted for more than 90%). Also, there is confusion of concept in theory.