Openness and Economic Performance... - United Nations Information ...

Journal of Chinese Economic and Business Studies, Vol. 1, No. 1, 2003, pp. 71–95

Openness and Economic Performance: A Comparative Study of China and the Asian NIEs

SHUJIE YAO and ZONGYI ZHANG

ABSTRACT Being the world’s largest developing economy, China’s successful economic performance since 1978 has had a powerful impact on the global economy. Its open policy features an evolutionary process, involving the gradual liberalization of foreign exchange, international trade and foreign direct investments. This paper evaluates how this evolutionary process has contributed to China’s economic success in comparison with the development experiences of the Asian newly industrialized economies (NIEs). It concludes that despite the economic crisis in 1997–98, China and the NIEs represent a successful development model, which is built upon openness and huge investments in physical and human capital. Key words: Openness; China; Asian NIEs.

1. Introduction China has achieved spectacular economic growth and social progress since economic reform and openness started in 1978. In many respects, it resembled the Asian newly industrialized economies (hereafter, the NIEs) from the 1960s to the 1990s.1 Although it is still a poor country in terms of per capita income, being the largest industrializing economy, its rapid growth over a prolonged period of time has made a powerful impact on the global economy in general and the Asia– Pacific region in particular. Moreover, the recent reunification of Hong Kong and Macao will greatly reinforce its position as one of the most influential economies in the world. China’s success has been due to Deng Xiaoping’s open-door policy, which followed closely the development path of the NIEs. Before economic reforms, the apparent lack of modern capitalism under Mao’s leadership, particularly during the Cultural Revolution, was a main constraint on economic performance. However, continuing economic growth and rapid industrialization in the NIEs Shujie Yao, Middlesex University Business School, The Burroughs, Hendon, London, NW4 4BT, e-mail: [email protected] and Zongyi Zhang, School of Management, Chongqing University, China Journal of Chinese Economic and Business Studies ISSN 1476-5284 print: ISSN 1476-5292 online ß 2003 Taylor & Francis Ltd http://www.tandf.co.uk/journals DOI: 10.1080/1476528032000039758

72

S. Yao and Z. Zhang

from the 1960s had a significant impact on the Chinese people and the communist leadership. Both politicians and academics wondered why their Chinese cousins in Singapore, Taiwan and Hong Kong could do things so well, but they themselves on the mainland could not. The death of Mao and the downfall of the Gang of Four provided a unique opportunity for Deng and his followers to change the economic development strategy. Hence, people’s energy and efforts were directed away from class struggle towards economic development and modernization. Policy reforms started from the countryside and agriculture. Enormous productivity gain in agriculture allowed rapid accumulation of surplus food and raw materials for industrial and urban development. The majority of the rural population was lifted out of absolute poverty in less than a decade of agricultural reform (Yao, 2000). In addition, over one-quarter of the rural labour force was transferred out of agriculture to rural industries and services. From the late 1980s, China entered a period of large-scale rural industrialization (Rozelle, 1994). In the urban sector, China required a development model, but the focus was on openness and market reform. Import-substitution and self-reliance were replaced with export-led, foreign technologies and international business practices. Mandatory planning was gradually replaced with free market competition. Hence, the market, rather than the plan, set prices of goods. Ideological reforms encouraged the development of non-state ownership. As a result, private, collective and foreign enterprises flourished, forcing the state sector to increase efficiency and productivity (Groves et al., 1994; Hay et al., 1994). One important aspect of market liberalization was the gradual reform of the foreign exchange market. Without bringing the official exchange rate towards the free market equilibrium level, China would not have been able to promote international trade and to attract foreign capital so successfully. Yet it is the rapid growth in exports and foreign direct investments (FDIs) that has brought about sustainable growth for such a long period of time. Unlike the NIEs, the Chinese currency (Renminbi, RMB) was unconvertible and overvalued persistently for decades, up to the late 1980s (Lardy, 1995; Chou and Shih, 1998).2 Currency overvaluation prohibited international trade and foreign investments. After 1979, the RMB was gradually devalued. By 1994, it was made convertible in a current account transaction for international settlement. The dual exchange system introduced in 1979 was replaced by a single exchange system from January 1994.3 It needs to be stressed that China’s success depended largely on its ability to manage the timing, sequence and scope of liberalization, arising from Deng’s philosophy of pragmatism and gradualism. A typical example of China’s gradualism is its partial openness in its stock and foreign exchange markets. Despite gradual devaluation and the introduction of a unified exchange rate from 1994, RMB was kept unconvertible in capital account transaction. The exchange of RMB to foreign currencies was also subject to tedious regulatory and administrative control. In the Shanghai and Shenzheng Stock Exchanges, shares are divided two types: A and B, with domestic investors trading A-shares and foreign investors trading B-shares. The number of companies that can issue B-shares and the volume of B-shares are highly regulated. This control mechanism effectively protected the foreign exchange and stock markets from speculative attacks by foreign investors and short-term capital flight. In retrospect, it was this gradual and pragmatic approach to openness that saved China from the recent economic crisis that severely crippled many Asian NIEs. In the recent literature, there are numerous studies examining the linkage between openness and economic performance in Asia. Most studies find that

Openness and Economic Performance

73

FDI and international trade contribute positively to economic growth (Harrold, 1995; Lardy, 1995; Pomfret, 1997, for China; Sengupta and Espana, 1994, for South Korea; Yue, 1999, for Southeast Asia; and Dowling, 1997, for the Asian high performing economies; Greenaway, 1998, for the developing countries in general). FDI involves a significant component of technological transfers and spillover, as inferred from the new theory of endogenous growth (Romer, 1986). Export-orientation forces producers to respond to international competition. In the NIEs, the outward looking strategy has been associated with relatively free labour and capital markets (Balassa, 1988). Following the NIEs, China started to exploit its comparative advantage by focusing on labour-intensive manufacturing for exports. The experience of China and the NIEs was in marked contrast to the inward orientation of development in most Latin American countries, which have been marked by considerable distortions in labour and capital markets, debts, budget deficits and economic stagnation (Brohman, 1996). Many recent econometric analyses have focused on understanding the causality between the dependent and independent variables. For example, does FDI cause GDP to grow, or vice versa? Most empirical results support the argument that FDI (or trade) can promote output growth. In the Chinese case, since foreign exchange policy is a precondition for the rapid growth in FDI and exports, it implies that the exchange rate mechanism must also have played an important role in economic performance. To prove this, we establish an econometric model system to examine how the real exchange rate, FDI and exports interact and how they contribute to economic growth in the Chinese regions using provincial level data for the period 1978–95. The rest of this paper is organized as follows. Section 2 assesses the economic performance and openness of China and the NIEs in comparison with the rest of the world in the last two to three decades. Section 3 presents some empirical analyses on the linkage between openness and economic growth. Section 4 discusses the recent economic crisis and recovery in Asia. It intends to emphasize that, despite the crisis, the remarkable achievements of the NIEs cannot be ignored and their successful development based on openness and investments in physical and human capital is still a useful example for the future development of other developing economies. Drawing on China’s experiences in reforming its foreign exchange and stock markets, it also stresses the importance of pragmatism and gradualism in openness and financial sector reforms. The final section offers some conclusions.

2. Economic Performance and Openness: A Brief Review According to the latest statistics (Table 1), China is classified as a middle-income economy by the World Bank, but it is still a poor country compared with the firsttier NIEs: Taiwan, South Korea, Hong Kong and Singapore.4 In 2000, for example, China’s per capita gross national income (GNI) in nominal terms was $840, which was only 3.2% of Hong Kong’s, or 16.3% of the world average. Even in PPP terms, China’s per capita GNI was only 15.3% of Hong Kong’s and 53.6% of the world average (NBS, 2000; World Bank, 2002). However, China shares a number of common features with the NIEs in terms of openness, investments in physical and human capital and, to some extent, strong governance. These features have made China and the NIEs different from many of the low-income countries in Africa, Latin America and south Asia.

74

S. Yao and Z. Zhang Table 1. Main economic indicators of selected economies in 2000 Countries

China, P.R. Hong Kong, China Indonesia Japan Korea, Rep Malaysia Singapore Taiwan, China* Thailand Low Income Middle Income High Income World

Population (million)

Nominal GNI/head (US$)

PPP GNI/head (US$)

1261 7 210 127 47 23 4 22 61 2459 2693 903 6054

840 25950 570 34210 8910 3380 24740 12333 2010 420 1970 27510 5150

3940 25660 2840 26460 17340 8360 24970 12333 6330 1990 5650 27450 7350

Notes: *Figure for Taiwan is GNP/head in 1998. PPP GNI is gross national income converted to international dollars using purchasing power parity rates. An international dollar has the same purchasing power over GNI as a US dollar in the United States. Low Income ($785 or less), Middle income ($785–$9655), High income ($9656 or more). GNP ¼ gross national product, GNI ¼ gross national income. Sources: World Bank (2002, pp. 12–14). Data for Taiwan is from NBS (2000, pp. 854).

Table 2. Economic and population growth in selected economies

China Hong Kong Indonesia Japan Korea Malaysia Singapore Taiwan Thailand Low Income Middle Income High Income World

GNP/head annual growth 1965–97 (%)

Population annual growth 1965–97 (%)

6.8 5.7 4.8 3.6 6.7 4.1 6.3 7.5a 5.1 1.4 2.2 2.3 1.4

1.7 1.8 2.0 0.8 1.5 2.6 1.9 n.a. 2.1 2.4 1.7 0.8 1.8

Sources: The United Nations (1997) for Taiwan. World Bank (1999), pp. 24–26, for all the other economies.

The annual growth rate of per capita GNP ranged from 4.1% in Malaysia to 7.5% in Taiwan during 1965–97 (Table 2). This is in sharp contrast to the world’s average of 1.4%. Over the reform period after 1978, China’s economic growth was almost unmatched by any other economy in the world. Its gross domestic product (GDP) more than quadrupled in 20 years (1978–95). Real per capita disposable income more than tripled in the cities and almost quadrupled in


75

the countryside (Yao, 2000). From 1994, China has become the second largest recipient of foreign direct investment (FDI) after the USA (Nolan and Wang, 2000). The latest figures show that China became the world’s ninth largest trading (and exporting) economy in 1999, moving from 23rd in 1978 (People’s Daily, 2000). Due to the Asian economic crisis, China’s exports plummeted in late 1998 and early 1999, but recovered strongly from mid-1999. Overall, total exports in 1999 reached a record high of $194.9 billion, up 6.1% from 1998, imports were $165.7 billion, up 18.2%. In the first three months of 2000, China’s total imports and exports were $60.3 billion, up 47.1% from the previous year (People’s Daily, 2000). High economic growth in China and the NIEs has been associated with, and enhanced by, relatively slow population growth. In the NIEs, low population growth reflected the general negative trend of women’s fertility as per capita income rises over time. In China, stringent family planning policy has been induced, which restricted each urban family to have only one child and each rural family to have two children. China’s family planning policy may not be popular among its populace and may have some undesirable long-term consequences, but the actual outcome of birth control has a number of advantages. First, it helps improve the average living standards of the population given a certain level of economic activities. Second, it reduces the pressure on employment and the natural environment. Third, it helps enhance human capital accumulation as more resources can be allocated to education and heath care. High economic growth in China and the NIEs has led to a significant reduction in the level of poverty. Poverty reduction in the NIEs was much more significant than in some large Latin American and South Asian countries during the 1970s and 1980s. In China, more than 75% of the rural population lived in poverty in 1978. This proportion rapidly declined to less than 13% by 1996 (Table 3). The most common features of China and the NIEs for their economic success are a high rate of savings, export push, and investments in physical and human capital. High saving rates may reflect the traditional culture of the Chinese and the East Asian people rather than a result of government policy. This traditional culture has been extremely helpful for capital accumulation and education in the developing economies. All the NIEs had a savings/GDP ratio of more than 30%. In Singapore and China, the saving rate is as high as 40–50%, which compares favourably to 22%, the world average, and 17% in the low-income economies (Table 4). In some Sub-Saharan African countries, the saving rate is negligible. Needless to say, saving is a prerequisite for domestic investments, which in turn, is an important factor for positive economic growth. In 1965–97, most NIEs and China maintained an average annual growth rate of 7.7–10.9% in physical investments, compared with 3.9% in the low income economies and 3.2% of the world average. The endogenous growth theory suggests that investment in human capital is as important as physical investment. Accumulation of human capital is reflected in a number of areas, particularly in health and education. Improvement of people’s heath has led to a significant increase in life expectancy and a large reduction in child and infant mortality. In China and the NIEs, life expectancy is significantly higher, and child mortality lower than in the low-income economies (Table 5). In education, the campaign to popularize primary and secondary school education has been highly successful. By 1997, almost all the Chinese children were able to receive primary education and over 70% (rising from 46%

76

S. Yao and Z. Zhang Table 3. Changes in poverty incidence Percentage of population below poverty line Economies

Period

First year

Last year

Indonesia Malaysia Singapore Thailand China (rural)

1972–82 1973–87 1972–82 1962–86 1978–96

58 37 31 59 75

17 14 10 26 6.7–13

Brazil Colombia Costa Rica India Morocco Pakistan Sri Lanka

1960–80 1971–88 1971–86 1972–83 1970–84 1962–84 1963–82

50 41 45 54 43 54 37

21 25 24 43 34 23 27

Notes: This table uses economy-specific poverty lines. Official or commonly used poverty lines are used. In other cases the poverty line is 30% of mean income or expenditure. The poverty line is measured in PPP dollar terms at about $300–$700 a year in 1985 except for Malaysia ($1420) and Singapore ($860). Apart from Malaysia, China, Costa Rica and Sri Lanka, measures for entries are per capita incomes. The measures of entries for other countries are per capita expenditures. The data for China are just for the rural population as few urban people are considered to be poor. Sources: World Bank (1993, Table 1.1, pp. 33) for all countries except China, and Yao (2000) for rural China.

Table 4. Investments and savings

China Hong Kong Indonesia Japan Korea Malaysia Singapore Thailand Low Income Middle Income High Income World

Investment annual growth % 1965–97

Savings as a percentage of GDP 1997

10.9 7.7 9.2 4.7 12.4 10.1 9.6 9.0

42.7 30.6 30.6 30.5 34.2 44.4 51.2 35.7

3.9 2.1 3.1 3.2

17.0 26.2 21.4 22.2

Sources: World Bank (1999, pp. 174–176).

in 1980) of the relevant aged children were able to receive secondary education. In the low-income economies in the same year, the gross enrolment rate in secondary schools was only 43% (World Bank, 1999, pp. 78–80). The popularization of primary and secondary education in China and the NIEs led to a significant reduction in adult illiteracy. Young–aged population illiteracy was almost

77

Openness and Economic Performance Table 5. Investments in human capital Life expectancy at birth (years)

Child mortality (age 1–5), deaths/1000

Illiteracy rate (age 1Il5–24) as %

1980

1997

1980

1997

1980

1997

China Hong Kong Indonesia Japan Korea Malaysia Singapore Thailand

67 74 55 76 67 67 71 64

70 79 65 80 72 72 76 69

167 118 238 100 272 190 157 245

151 83 214 73 103 151 102 171

10 4 11 0 0 12 3 4

3 1 3 0 0 3 1 2

Low Income Middle Income High Income World

52 65 74 63

59 69 77 67

320 195 143 218

265 168 100 167

47 11 0 24

32 5 0 16

Sources: World Bank (1999, pp. 110–112 (life expectancy and child mortality), pp. 78–80 (young people illiteracy rate)).

Table 6. Integration with the world economy Trade in goods as % of PPP GDP 1987 China, P.R. Hong Kong Indonesia Japan Korea Malaysia Singapore Thailand Low income Middle income High income World

1997

FDI inflows as % of PPP GDP 1987

1997

FDI inflows current prices ($billion) 1990

1997

6.8 125.0 11.1 20.8 36.6 49.4 200.7 16.9

8.5 250.4 13.7 25.0 44.9 90.0 290.7 29.7

0.2 – 0.1 1.2 0.5 0.7 10.0 0.4

1.2 – 0.7 1.0 1.2 2.9 14.3 1.0

3.5 – 1.1 1.8 0.8 2.3 5.6 2.4

44.2 – 4.7 3.2 2.8 5.1 8.6 3.8

7.0 10.3 27.4 20.6

8.4 18.6 38.7 29.6

0.1 0.3 2.2 1.5

0.3 1.4 3.1 2.4

1.1 22.6 167.0 192.7

10.6 150.0 233.9 394.5

Sources: World Bank (1999, pp. 324–326 (Trade & FDI)).

eliminated in Korea and Singapore by the late 1990s. It was reduced to 1–3% of the relevant age population in China and other NIEs. This is in contrast to 32% in the low-income economies. Another common feature of China and the NIEs is their integration with the global economy. This is reflected in their export-push development strategy and the absorption of foreign capital. Hong Kong and Singapore are the most open economies measured by the trade/GDP ratio. The volume of trade is two to three times that of PPP GDP. The trade/GDP ratios are also high in other NIEs, ranging

78

S. Yao and Z. Zhang Table 7. Top ten recipients of FDI in the developing world (1998)

All LDCs China Brazil Mexico Argentina Poland Malaysia Chile Thailand Venezuela Russia Fed Rest of world

FDI $billion

As a percentage of total

160.0 44.8 25.6 10.7 5.9 5.9 5.3 5.3 5.1 4.0 3.2 44.2

100.0 28.0 16.0 6.7 3.7 3.7 3.3 3.3 3.2 2.5 2.0 27.6

Sources: World Bank (1999), World Development Indicators, World Bank (Washington, DC), pp. 323.

from 16.9% in Thailand to 90% in Malaysia. These ratios are significantly higher than the average ratio of the low-income economies (Table 6). In terms of international trade, China is still far less open than the NIEs. In 1997, for example, the trade/PPP–GDP ratio was only 8.5%. However, the figures in Table 6 may understate China’s openness in two respects. First, because China is a large country, it is difficult, or impossible, to achieve the same level of trade/GDP ratio as that of the small city economies of Hong Kong and Singapore. Secondly, the GDP figures are calculated in PPP terms. If they were calculated in nominal terms, the trade/GDP ratio in 1997 would be 31% for China, rising from 8.81% in 1977 (Pomfret, 1997, Table 1). Another indication of openness is the absorption of foreign capital. In 1978, there was little FDI in China. By 1995, China became the second largest recipient of FDI after the USA. The most rapid growth in FDI was registered in the late 1980s and 1990s. The FDI/PPP–GDP ratio rose sixfold from 0.2% to 1.2% in 1987–97. Total FDI inflows rose almost 13-fold from $3.5 billion in 1990 to $44.2 billion by 1997 (Table 6). In 1998, China accounted for 28% of the total FDI flowing into the developing countries (Table 7). China’s absorption of foreign capital followed the recent development path of Malaysia, Korea and Singapore. In general, the NIEs and China are much more open to FDI than most of the low-income and middle-income economies over the last three decades.

3. How does Openness Affect Growth? 3.1. Growth and Openness: An International Perspective One relevant study on the NIEs is by Sengupta and Espana (1994). They use time series data to estimate an augmented Cobb–Douglas production function including export as an explanatory variable for GDP growth. In their sample, there are two NIEs (Korea and Taiwan), three mature industrialized countries (Japan, Germany and Belgium) and one developing economy (the Philippines).


79

Table 8. Externality effect of exports on growth: 1967–86 Economies

Intercept

I

Korea Taiwan Japan Belgium Germany (FRG) Philippines

208.9 6.8* 7.7 314.8* 190.0* 0.7*

0.151* 0.094* 0.615* 1.148 1.651 0.336*

I2 n.a. n.a. 0.384* 0.001 0.003 n.a.

L_ 0.024* 0.050 0.119* 0.631* 0.028* 0.001

X_ 0.401* 0.438* 0.116 0.333* 0.610* 0.389*

R2 0.614 0.933 0.690 0.419 0.447 0.407

DW 1.96 1.56 1.98 1.95 2.44 1.52

X/Y (%) 30.6 56.7 12.9 59.0 28.3 21.6

Notes: (1) The dependent variable is GDP growth. (2) I and I2 respectively denote investment and its squared term, L_ and X_ respectively labour and export growth, X/Y export/GDP ratio. (3) The ‘*’ sign denotes significance at the 5% level or below. Source: Sengupta and EspaZa (1994, Table 4).

Figure 1. GDP and export growth in APEC economies (1990–95), Source: Pan (1998), Figure 4.9.

The regression results show that, in all cases except Japan, export had a positive and significant effect on GDP growth (Table 8). Their model does not include FDI and foreign exchange rate, which are particularly important for the Chinese economy under transition. In the next subsection, we will extend the model to include all these variables for China. Figure 1 shows further evidence on the positive relationship between GDP and export growth. Over the period 1990–95, China achieved the fastest growth in GDP and exports among the APEC (Asian Pacific Economic Corporation) members. The NIEs are among the best performers. The mapping of cross-countries growth shows a clear and positive relationship between GDP and exports. Some recent studies on FDI focus on how it is determined but not on how it is linked to economic growth. Liu et al. (1997) shows that FDI is determined by GDP growth, international trade, and other variables. Some more descriptive studies (e.g. Chen et al., 1995; Lardy, 1995) suggest that FDI has an important role in China’s

80

S. Yao and Z. Zhang

Figure 2. Trade/GDP and FDI/GDP ratios in 1995 in the APEC economies, Source: Pan (1998), Figure 4.10.

economic performance. However, there have been no studies that quantify the triangular relationship between FDI, export and economic growth. However, in a cross-country study, Pan (1998) finds a positive relationship between the ratios of trade/GDP and FDI/GDP among the APEC economies (Figure 2). Three of the NIEs (Singapore, Hong Kong and Malaysia) had the highest FDI stock/GDP and trade/GDP ratios. 3.2. Growth and Openness: Evidence from the Chinese Regions Before economic reforms, import-substitution and rigid price control were two major features of China’s development strategy. For example, there was little FDI in as late as 1982, four years after the inception of economic reforms in 1978. The total amount of FDI was only $0.64 billion in 1983, but increased gradually thereafter, reaching $4.37 billion by 1991. After Deng’s famous tour to South China in 1992, there was a sudden surge of FDI. The total inflow jumped to $11.3 billion in that year, rising to $27.5 billion in 1993, $33.8 billion in 1994 and $44.8 billion in 1998 (Table 9). It is obvious that the history of FDI is much shorter in China than in the NIEs. In addition, FDI is highly concentrated in a number of provinces along the eastern coast (Guangdong, Shanghai, Tianjin, Fujian, Shangdong, Jiangsu, Zhejiang, Hainan, Liaoning and Hebei). In 1995, for example, the eastern region accounted for over 88% of total FDI, with Guangdong alone accounting for 27%. One important reason for the skewed distribution of FDI across regions is the early reforms that focused on opening up four special economic zones in Guangdong and Fujian in 1980, 14 coastal cities in 1984, Hainan Island in 1988 and Shanghai Pudong Development Zone in 1989. Of course, there are fundamental reasons why the coastal cities were selected as open zones. Compared to the inland areas, the coastal regions had a more productive agricultural and industrial basis,


81

Table 9. Chinese FDI, exports and structure, 1980–99 Year

Actual FDI ($billion)

Total exports ($billion)

Export by foreign invested firms as % of total

Manufactured exports as % of total

1980 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

0.2 2.0 2.2 2.6 3.7 3.8 3.8 4.7 11.3 27.5 33.8 37.7 41.7 44.2 44.8 n.a.

18.2 27.4 30.9 39.4 47.5 52.5 62.1 71.8 85.0 91.8 121.0 148.8 151.1 182.7 183.7 194.9

0.0 1.1 1.6 3.0 5.2 8.3 12.5 16.8 20.4 27.5 28.7 31.7 n.a. n.a. n.a. n.a.

49.7 49.4 63.6 66.5 69.7 71.3 74.4 77.5 79.9 81.8 83.7 85.6 85.5 86.9 n.a. n.a.

Sources: Lardy (1995, Tables 1, 6 and 7) and Pomfret (1997, Tables 2 and 3) for data up to 1995. Data for 1996 and 1997 are derived from NBS (1998). Data for 1998–99 are from People’s Daily, 14 March 2000.

a more efficient transportation system, better environmental and human resources, and above all, easier access to China’s largest investors, especially Hong Kong.5 China’s strategy to open its market for foreign investors coincided with, and is reinforced by, its effort to promote exports. Right from the inception of economic reform, China emphasized that foreign invested firms (foreign-owned companies, equity joint ventures, and cooperative ventures) must produce a large proportion of their outputs for exports. As a result, many foreign invested firms were concentrated in the export processing and manufacturing areas in the special economic zones, the open cities and Hainan Island. This investment strategy produced some spectacular results. First, total exports increased dramatically from only $18.2 billion in 1980 to $194.9 billion by 1999. Secondly, the share of manufactures exports in total exports rose from 49.9% in 1980 to 86.9% by 1997. Rapid expansion of exports and a significant change in the export mix resembled the export performance of Korea and Taiwan in the 1960s and 1970s, and of the second-tier NIEs in the 1980s and early 1990s. Thirdly, foreign-invested firms played an important role in China’s export drive. Their exports were negligible in the first half of 1980s but increased to $46.9 billion, or 31.7% of total exports by 1995 (Table 9). Although China’s success in attracting FDI and promoting exports can be explained by many factors, it is important to note that gradual reform in the foreign exchange market played a critical role. Without devaluing the RBM, it would be impossible to make the Chinese market attractive for foreign investors whose transactions involve frequent exchanges of foreign currencies and the RBM. In the early 1980s, when the official exchange rate was substantially lower than the black market rate, foreign investors had little incentives to invest in China.6 In the late 1980s, the government established a few official swap markets to facilitate the reallocation of foreign exchange and to maintain a dual exchange rate mechanism. The swap market was an official channel allowing investors to change

82

S. Yao and Z. Zhang

foreign currencies into RBM at a higher rate than the official exchange rate. It was the first important step of the government to provide incentives to attract FDI and promote exports. As the foreign exchange reserve increased rapidly in the early 1990s, the dual rate system was abolished in January 1994, and so were the swap markets. By then, the RBM had been gradually devalued towards its market equilibrium level (Chou and Shih, 1998, Table 2). The official exchange rate was devalued from 1.68 yuan per dollar in 1978 to 8.321 yuan in 1995, or by almost 400%. In real terms, when adjusted by the US and Chinese consumer price indexes, the real exchange rate rose by over 200% in the same period (Table 10). Turning to the growth theory, China’s systematic reforms in the foreign exchange market, its efforts to promote exports and to attract FDI all have the same objective, that is, creating a better environment for economic growth. If we consider economic growth as the core of the model, then output is determined by physical inputs (physical capital and labour), the internal production environment (human capital, transportation, institutions, and the like), and the external environment (FDI, export and foreign exchange mechanism). This economic growth model can be illustrated in Figure 3. Output (GDP) is basically determined by two physical inputs: labour and capital. However, the efficiency of input usage, or economic performance, is further determined by two sets of factors: external and internal. The external factors are related to openness, including FDI, export, and the foreign exchange mechanism. The internal factors include human capital, infrastructure, location, and institutions Table 10. Nominal (official) and real exchange rates of RMB/US$ Year

Nominal exchange rate

Real exchange rates

1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

1.680 1.550 1.500 1.705 1.893 1.976 2.327 2.937 3.453 3.722 3.722 3.766 4.784 5.323 5.515 5.762 8.619 8.321

1.740 1.750 1.820 2.220 2.570 2.730 3.260 3.920 4.430 4.610 4.050 3.650 4.780 5.390 5.460 5.190 6.540 5.660

1995/1978

4.953

3.253

Notes: REt ¼ OEt (CPIUS/CPIChina), RE and OE respectively denote real and official exchange rates, CPI is consumer price index using 1990 as the base year. Sources: NBS (1996), for the official exchange rates and China’s CPI. US Department of Commence (1980–96), for the US CPI.


83

Figure 3. Economics growth and production environment.

(e.g. government policy, legal regulations, etc). Economic performance varies significantly across countries due to the variations in both internal and external factors. Some recent cross-country studies reveal that human capital, saving and population growth are the three main variables responsible for inter-country growth differences (Islam, 1995; Sala-i-Martin, 1996). However, few studies have considered all the internal and external factors in a cross-countries, or cross-regions analysis. This may be due to some difficult technical problems, such as multi-collinearity and simultaneity, which cannot be reconciled in a single regression model. To overcome these problems, instead of estimating one single equation, we have three regression models (with FDI, Export and GDP as the dependent variables) run in a seemingly unrelated equation system. We also use the dynamic system approach with appropriate instruments in the DPD (dynamic panel data estimation) proposed by Arellano and Bond (1998). Since it is difficult to obtain comparable data for different countries, we use the panel data for the Chinese provinces over the period 1978–95. Each province can be treated as a separate economy because there are significant differences among the provinces regarding the degree of openness and economic performance. The empirical results are presented in Table 11. The detailed explanations on data, estimation technique and model specification, are provided in the appendix. Here we just highlight the important points of the results. GDP is defined as a function of capital stock, employment, human capital, real exchange rate, FDI, export and transportation. Other variables include a dummy variable representing the eastern economic area (East), a dummy variable representing the unusual economic boom in 1992–95 after Deng’s famous tour to South China, and a time trend.

84

S. Yao and Z. Zhang

Table 11. Regression results based on a panel data of 30 Chinese provinces in 1980–95 GDP Xs Constant Labour Capital Human Exchange FDI Export Transport East D9295 Time

Export

s

t-values

Xs

s

t-values

Xs

s

t-values

17.55 0.409 0.472 0.054 0.102 0.006 0.111 0.038 0.033 0.021 0.011

1.79 12.13 8.85 2.69 9.76 1.69 13.29 1.68 0.29 2.68 2.01

Constant GDP Exchange East D9295 Export(1)

0.783 0.187 0.141 0.288 0.256 0.793

3.19 4.85 1.91 3.55 4.55 18.79

Constant GDP Wages Exchange Transport Human East D9295 FDI(1)

3.044 0.165 0.114 0.626 0.157 0.044 0.373 0.529 0.798

4.15 2.33 0.92 2.65 1.63 0.52 3.57 4.96 18.93

Test-Statistics R2 Wald Sargan M1 M2

FDI

1.56 0.864

Test-Statistics 0.981 p ¼ 0.000 p ¼ 0.889 p ¼ 0.119 p ¼ 0.388

R2 Wald Sargan M1 M2

1.295 0.718

Test-Statistics 0.931 p ¼ 0.000 n.a. p ¼ 0.195 p ¼ 0.473

R2 Wald Sargan M1 M2

0.948 0.749

0.919 p ¼ 0.000 n.a. 0.343 0.454

Notes: All the variables are in natural logarithms. Values are measured in 1990 constant prices. Detail explanations on data, model specification, estimation techniques, test-statistics are provided in the appendix. Data sources: NBS (1996).

The results suggest that, apart from capital and labour, both internal and external factors have a significant impact on GDP. For the external factors, the effects of export and real exchange rate are much stronger than those of FDI. The output elasticity of export is over 0.11, of real exchange rate 0.10, and of FDI less than 0.01. For the internal factors, human capital is more important than transportation and location. The output elasticity of human capital is 0.054. The location effect (East) is insignificant. This may be due to the fact that regional productivity differences have been largely explained by the variations in other external and internal factors. The estimated coefficient on the time trend indicates that the Hicks-neutral technological progress was 1.1% per annum. It reflects a location-invariant macro productivity shock over the data period. Export is defined as a function of GDP, real exchange rate, location and the lagged dependent variable. The dummy variable for the period 1992–95 is also included. All the explanatory variables are significant at the usual 5% level. Referring to the regression results on GDP, it is clear that exports and GDP are interdependent. The long-run elasticity of exports with respect to GDP is 0.89, suggesting that a 10% rise in GDP will lead to an 8.9% rise in exports, other conditions remaining unchanged. Real exchange rate is another important factor affecting exports. In other words, without reforming the exchange market, China’s exports would have been severely disadvantaged.7 Although the location factor is not significant in the GDP equation, it is significantly associated with exports. This indicates that the eastern area is far more export-oriented than the inland provinces, even after controlling for all the other factors.


85

Figure 4. Linkage between openness and output in the Chinese regions, 1980–95. Notes: The values are long-run elasticities. Arrows indicate the direction of impact. Source: Table 11.

FDI is defined as a function of GDP, effective wages (nominal wages adjusted by productivity), real exchange rate, transportation, human capital, location and the lagged dependent variable. The dummy variable for 1992–95 is also included. Like exports, FDI is mainly determined by GDP, with a long-run elasticity of 0.825. The real exchange rate also plays an important role in FDI. The location factor is significant, indicating that the eastern area is much more successful in attracting FDI than the rest of the country. The effects of human capital and wages on FDI are positive but not significant, implying that wage differentials across provinces were not a significant concern of foreign investors in the data period. In reality, wage differences may just reflect the variation in labour quality. The quantitative relationship between GDP and the three factors of openness are summarized and illustrated in Figure 4. In the model, the real exchange rate is treated as an exogenous variable. The results suggest that it had a significant and sizeable effect on the three endogenous variables: FDI, export and GDP. It is clear that the gradual devaluation of RMB towards its real equilibrium exchange rate with the US dollar over the data period has been one of the most important factors responsible for China’s success in attracting FDI, promoting export, and above all, stimulating economic growth. FDI and exports have a simultaneous relationship with GDP. FDI inflows and exports stimulate GDP growth, which, in turn, provides a solid basis for attracting more FDI and an export push. Such an interaction among these three economic variables formed a virtuous circle of openness, growth, more openness and more growth. 4. Economic Crisis and Recovery Only four years after the World Bank published The East Asian Miracle (World Bank, 1993), all the high performing economies in East Asia were plunged into a deep financial and economic crisis in late 1997 and 1998. The crisis was first triggered by the collapse of the Thai baht and stock market. The contagion soon spread into Malaysia, Indonesia and Korea. All these countries were so severely affected that the International Monetary Fund (IMF) was asked to pump in huge sums of money to rescue the economies of these countries.8

86

S. Yao and Z. Zhang

The Asian crisis brought down the world GDP growth from 4.8% in 1997 to just 1.9% in 1998. Japan was already in recession before the crisis, but the crisis hit Japan particularly hard, and its recession contributed to 0.5 percentage points of the decline in world GDP in 1998. Since Japan was the power engine in East Asia, its own recession accelerated the decline of the NIEs. The worst hit economies were Thailand, Indonesia, Malaysia and Korea. In 1998, output fell by 13% in Indonesia, 8% in Thailand, 6.7% in Malaysia and 6.0% in Korea. Even Hong Kong and Singapore were forced into negative growth in 1998. Only Taiwan and China managed to maintain a positive growth, but at a much lower rate than before the crisis (Table 12). The scale of the crisis was unprecedented in these economies during their recent development history. In early 1998, the crisis seemed to spread into the already fragile Russian and other eastern European economies. A worldwide economic depression looked imminent as the US and western-European stock markets suffered severe losses in mid 1998. In the third quarter of 1998, there was a strong speculation that China might have to devalue the RBM to maintain a competitive edge against the NIEs. It also looked as if the potential collapse of the RBM and the Chinese economy would serve a final blow and trigger a global crisis. Suddenly, China’s ability and willingness not to devalue its currency became the centre of world attention. In late 1998, Mr Zhu Rongji, China’s premier, promised that China would not devalue the RBM. In early 1999, the central government also ordered the central bank, the People’s Bank of China, to issue 100 billion yuan investment bonds to boost its domestic economy in the face of a depressing external market. For the first time in over 20 years, China experienced a significant negative growth in exports in late 1998 and early 1999. However, China’s effort not to devalue the RMB and its ability to achieve a GDP growth of 7.8% in 1998 played a significant role in easing the crisis in Asia. During the economic crisis, many economic analysts thought that the Asian miracle might have been over, and the ‘bubble’ may have burst. Paul Krugman shares the view that Asia’s crisis was due to corruption and crony capital (Krugman, 1998). He argues that the failure of the financial system in east Asia was due to excessive investments in risky and low returns (or unprofitable) projects. Table 12. GDP growth of the NIEs in economic crisis 1997–99 by quarter

1997Q1 1997Q2 1997Q3 1997Q4 1998Q1 1998Q2 1998Q3 1998Q4 1999Q1 1999Q2 1999Q3 1999Q4

China

Hong Kong

Indonesia

Korea

Malaysia

Singapore

Taiwan

Thailand

9.4 9.6 8.1 8.2 7.2 6.8 7.6 9.6 8.3 7.1 7.0 6.8

6.1 6.4 5.7 2.7 2.8 5.2 7.1 5.7 3.4 0.7 4.5 8.7

6.6 6.6 6.6 6.6 6.2 16.5 17.4 13.9 10.3 1.8 0.5 5.8

5.4 6.3 6.3 3.9 3.8 6.6 6.8 5.3 4.6 9.8 12.3 12.3

8.2 8.4 7.4 6.9 1.8 6.8 8.6 8.6 1.3 4.1 8.1 10.6

3.8 7.8 10.1 5.6 5.6 1.5 0.7 0.8 1.2 6.7 6.7 7.1

6.8 6.3 6.9 5.9 5.9 5.2 4.7 3.7 4.3 6.5 5.1 6.8

0.4 0.4 0.4 0.4 8.0 8.0 8.0 8.0 0.1 3.5 3.5 7.7

Notes: Average of latest 3 months compared with average of previous 3 months. The rates for Thailand in 1997 and 1998, for Indonesia in 1997 are annual rates. Sources: The Economist (1997–99, various issues), Economic Indicators.


87

It is therefore logical to infer that the bubble must burst once bad debts and financial losses accumulated to such a point that even the government cannot bear them. Krugman’s explanation is valid in the sense that the Asian economies have some fundamental problems in the financial institutions and governments’ soft-attitudes towards large state enterprises and banks. His hypothesis, however, may not be sufficient to prove the sudden occurrence of a deep crisis. Radelet and Sachs (1998) suggest that there were other important factors responsible for the sudden collapse of Thailand, Malaysia, Indonesia and Korea. In our view, two of their explanations are still valid today, (1) panicked reversals in capital flows, and (2) lack of a legal and regulatory framework to support a liberalized and wide-open financial system in the world market. Panicked reversals in capital flows were caused by the so-called herd behaviour of domestic depositors and foreign investors: if one decides to withdraw money in anticipation of a crisis, others will follow. The worst hit economies (Thailand, Indonesia, Malaysia and Korea) had the highest short-term debt-to-reserve ratios, which rendered them the most vulnerable to attacks by investors rushing to draw out money in panic. Panicked outflows of capital may be due to the premature opening of domestic financial markets to foreign investors. In Thailand, Indonesia, Malaysia and Korea, new banks and finance companies were allowed to operate without supervision and adequate capitalization. This financial environment was partly responsible for the capital withdrawals, panic and deep economic contraction that followed. In contrast, Hong Kong, Singapore and Taiwan had a much more mature financial market with a far better monitoring and supervision mechanism than Indonesia, Thailand, Malaysia and Korea. This explained why the former suffered far less during the crisis than the latter. Like the worst hit economies, China’s financial system had similar problems of corruption, crony capital, huge non-performing loans and losses by the state-own enterprises and banks. However, China was able to escape a similar crisis. In retrospect, it was not because China had better economic fundamentals than the crisis economies, but because it had a more pragmatic approach to financial market reforms. The Chinese government understood well that it could not entirely open up its financial market for foreign investors given the poor state of its financial institutions and the unpredictable volatility of stock markets. Foreigners were restricted to investing in the B shares in the Shanghai and Shenzheng stock exchanges. The RMB was gradually devalued but it was kept unconvertible in capital account transactions even after a unified exchange rate was introduced in 1994. Moreover, foreign banks and other financial institutions are not allowed to open branches in China without going through a tedious administrative procedure. China’s gradual reform and openness in the financial market resembled its gradual reform approach in the real sectors of the economy. In the middle of the Asian crisis, some observers suggested that Asia’s development was somehow a bubble, rather than a miracle, and when the bubble burst, the remarkable achievement of these economies in the past decades would be completely wiped out by the crisis. This view is obviously mistaken. In the past few decades, there have been enormous gains in income levels, health, and education. It is precisely the development strategies of export-push, FDI attraction and massive investments in physical and human capital that make these economies strong enough to withstand such an unanticipated shock without totally collapsing. No one was able to predict a deep crisis in Asia before it happened. Equally surprisingly, no one was able to predict a quick recovery. Many observers suggested

88

S. Yao and Z. Zhang

Figure 5. GDP growth rates of crisis economies in Asia, 1997–99 (Source: Table 12).

Figure 6. Real GDP growth rates, Argentina and Mexico, 1991–97 (Sources: Radelet and Sachs, 1998, Figure 1).

that it might take many years for the crisis economies to move from negative to positive growth. The latest statistics, however, show that none of the crisis economies suffered more than six consecutive quarters of negative growth. Figure 5 (derived from Table 12) shows that the quarterly GDP growth rates of the worst hit economies, Indonesia, Thailand, Malaysia and Korea, exhibit a clear V-shape pattern, which is not dissimilar to that experienced by Mexico and Argentina in their crisis years of 1994–96 (Figure 6). In most economies, GDP growth slowed from the first quarter to the last quarter in 1997. It then plunged into real contraction throughout 1998 (except China and Taiwan, and Singapore in the first quarter), but recovered into positive territory from the first quarter of 1999 in Thailand and Korea, and from the second quarter of 1999 in all the other crisis economies. By the end of 1999 and early 2000, the recovery trend appeared unstoppable. In the last quarter of 1999, GDP grew


89

12.3% in Korea, 10.6% in Malaysia, 6.8% in Thailand and 5.8% in Indonesia (Table 12). Singapore, Taiwan and China also achieved significant GDP growth. 5. Conclusions This paper compares the economic performance of China with that of the NIEs. There is no doubt that China and the NIEs shared a number of common features in their development strategy. One of the most important features is openness, including the attraction of FDI, export-push and financial market reforms. In the Chinese case, the gradual devaluation of the RMB was a prerequisite for its success in attracting FDI and promoting exports. The empirical results show a clear circular positive relationship among GDP, FDI, export and foreign exchange policy. Apart from openness, China and the NIEs had some other common features, particularly in savings and investment in physical and human capital. All these economies have achieved tremendous economic and social progress over the past three decades. Apart from the crisis year of 1998, GDP growth has been persistently higher than the world average, people have become far more educated, life expectancy is much higher, infant and child mortality has become much lower. It is precisely this economic and social progress that makes China and the NIEs capable of recovering quickly from the unexpected and devastating crisis in 1997–98. Of course, China and the NIEs have a number of fundamental economic problems, particularly in the financial market. Corruption, crony capital, financial repression, and the like, were the main causes for economic instability and periodic depressions. In the NIEs, their efforts to open up the financial market for foreign investors without an effective monitoring and control mechanism on the flows of short-term capital was one major factor responsible for the recent economic crisis. China’s pragmatism and gradualism proved to be highly successful in guarding against a similar disaster. Although China has been successful in terms of economic growth, it has a number of important weaknesses, including the inefficiency and loss-making of state-owned industries, the debt-ridden financial institutions, corruption and crony capital, uneven regional development, environmental degradation, and rising income inequality. These problems may greatly undermine its ability to maintain high economic growth and to avoid any deep economic recession, but the experiences and lessons learned from the crisis economies are useful for its future economic development. One such lesson is that the timing and scale of financial reforms need to be carefully designed. China joined the World Trade Organization (WTO) in December 2001. Joining the WTO means that China has to be far more open than is it today. The real challenges will come from the most vulnerable sectors of the economy: banking, insurance, capital and technological intensive manufacturing and agriculture. If China has to open all these sectors for international competition, there are potential dangers that it may not be able to cope with the cyclical shocks that are typical of international capitalism. Some sectors may be so unprofitable that a large number of firms have to be closed down, with thousands of people being made redundant. The stock markets and the financial sectors may have to withstand huge short-term capital flight and international speculation, as it was the case in the NIEs during the recent crisis. However, China has to be prepared for such challenges. The main conclusion drawn from this paper is that openness can promote economic growth, but openness may also bear some unforeseen risks. Hence, it is useful to stress the

90

S. Yao and Z. Zhang

timing, scale and sequencing of reform in the financial and stock markets. China was successful in the past because of gradualism, but this also implies that it still has a long way to go to become a truly open economy that can engage in fullscale competition with the industrialized economies. Notes 1. In this paper, the Asian NIEs include the first-tier economies of Taiwan, Singapore, South Korea and Hong Kong and the second-tier economies of Malaysia, Thailand and Indonesia. 2. Chou and Shih estimated that the RMB was overvalued by 57% in 1978. Overvaluation was brought down to 33% in 1983 and only 5% in 1984 due to successive devaluation (Chou and Shih, 1998, Table 2). 3. The evolution of foreign exchange liberalization is referred to in Chou and Shih (1998) and Pomfret (1997). 4. In this paper China refers to the mainland of China, excluding Taiwan, Hong Kong and Macao. Taiwan is the Taiwan province of China. Hong Kong and Macao are special administration areas of China. Korea is referred to as the Republic of Korea. 5. Hong Kong and Macao accounted for 54.31% of China’s accumulative FDI in 1984–95, Japan 8.2%, Taiwan 7.82%, USA 8.0%, Singapore 3.81%, Korea 2.85% and UK 2.34% (NBS, various issue, 1984–98). 6. Ding (1998) has a detailed study on the foreign exchange black market and exchange flight in China in this period. 7. A relevant study by Brada et al., (1993) also shows that devaluation (change in the real exchange rate) had a positive effect on the balance of trade in China. 8. The IMF signed emergency lending agreements with Thailand in August 1997 ($17 billion), with Indonesia in November 1997 ($35 billion), and with Korea in December 1997 ($57 billion). The actual amounts of disbursement were, however, much smaller. IMF also attached a number of conditions for the loans. Such conditions, including a drastic restructuring of the financial systems in the middle of a deep crisis, are blamed for aggravating, rather than easing the crisis (Radelet and Sachs, 1998).

References Arellano, M. and Bond, R.S., Dynamic Panel Data Estimation Using DPD98 for Gauss. London: Mimeo, Institute for Fiscal Studies, 1998. Balassa, B., ‘‘The Interaction of Factor and Product Market Distortions in Developing Countries,’’ World Development, 1988, 11(2). Brada, J., Kutan, A.M., and Zhou, S., ‘‘China’s Exchange Rate and the Balance of Trade,’’ Economics of Planning, 1993, 26, pp. 229–242. Brohman, J., ‘‘Postwar Development in the Asian NICs: Does the Neoliberal Model Fit Reality?’’ Economic Geography, 1996, 72(2), pp. 107–130. Chen, C., Change, L., and Zhang, Y., ‘‘The Role of Foreign Direct Investment in China’s Post-1978 Economic Development’’, World Development, 1995, 23(4), pp. 691–703. Chou, W.L. and Shih, Y.C., ‘‘The Equilibrium Exchange Rate of the Chinese Renminbi,’’ Journal of Comparative Economics, 1998, 26, pp. 165–174. Ding, J., ‘‘China’s Foreign Exchange Black Market and Exchange Flight: Analysis of Exchange Rate Policy,’’ The Developing Economies, 1998, 36(1), pp. 24–44. Dowling, P., ‘‘Asia’s Economic Miracle: A Historical Perspective,’’ Australian Economic Review, 1997, 30(1), pp. 113–123. The Economist, 1997–1999. (various issues) Economic Indicators. Fleisher, B.M. and Chen, J., ‘‘The Coast-Noncoast Income Gap, Productivity, and Regional Economic Policy in China,’’ Journal of Comparative Economics, 1997, 25, pp. 220–236. Greenaway, D., ‘‘Does Trade Liberalization Promote Economic Development?’’ Scottish Journal of Political Economy, 1998, 45(5), pp. 491–511. Groves, T., Hong, Y., McMillan, J. and Naughton, B., ‘‘Autonomy and Incentives in Chinese State Enterprises,’’ Quarterly Journal of Economics, 1994, 183–209. Harrold, P., ‘‘China: Foreign Trade Reform: Now for the Hard Part,’’ Oxford Review of Economic Policy, 1995, 11(4), pp. 133–146.


91

Hay, D., Morris, D., Liu, G., and Yao, S. Economic Reform and State-Owned Enterprises in China 1979– 1989, Clarendon Press Oxford, 1994. Islam, N., ‘‘Growth Empirics: a Panel Data Approach,’’ Quarterly Journal of Economics, 1995, 110, pp. 1127–1170. Krugman, P., Bubble, Boom, Crash: Theoretical Notes on Asia’s Crisis. Mimeo: MIT, 1998. Lardy, N.R., ‘‘The Role of Foreign Direct Investment in China’s Economic Transformation,’’ China Quarterly, 1995, pp. 1065–1082. Liu, X., Song, H., Wei, Y., and Romilly, P., ‘‘Country Characteristics and Foreign Direct Investment in China: A Panel Data Analysis,’’ Weltwirtschaftliches Archiv, 1997, 133(2), 311–329. National Bureau of Statistics (NBS), China Regional Economy: A Profile of 17 Years of Reform and Open-up. Beijing: China Statistical Press, 1996. National Bureau of Statistics (NBS), Statistical Yearbook of China. Beijing: China Statistical Press, 1998. National Bureau of Statistics (NBS) Statistical Yearbook of China. Beijing: China Statistical Press, 2000. Nolan, P. and Wang, X., ‘‘Reorganising Amid Turbulence: China’s Large-Scale Industry,’’ in S. Cook, S. Yao, and J. Zhuang, eds, The Chinese Economy Under Transition, Basingstoke: Macmillan, 2000. Pan, S., ‘‘Asia Pacific Economic Cooperation and Regionalism in the World of Globalisation and Regionalisation,’’ Unpublished PhD Dissertation, UK: University of Sheffield, 1998. People’s Daily, 14 March, 2000, pp. 1 (China becomes the 9th largest exporting nation in the world). Pomfret, R., ‘‘Growth and Transition: Why has China’s Performance been so Different?’’ Journal of Comparative Economics, 1997, 25, pp. 422–440. Radelet, S. and Sachs, J.D., ‘‘The East Asian Financial Crisis: Diagnosis, Remedies, Prospects,’’ Brooking Papers on Economic Activity, 1998, 1, pp. 1–90. Rozelle, S., ‘‘Rural Industrialisation and Increasing Inequality: Emerging Patterns in China’s Reforming Economy,’’ Journal of Comparative Economics, 1994, 19, pp. 362–391. Romer, P., ‘‘Increasing Returns and Long-Run Growth’’, Journal of Political Economy, 1986, 94(4), pp. 1002–1037. Sala-i-Martin, X., ‘‘The Classical Approach to Convergence Analysis’’, Economic Journal, July 1996, 106, pp. 1019–1036. Sengupta, Jati K. and Juan, R. Espana, ‘‘Exports and Economic Growth in Asian NICs: An Econometric Analysis for Korea,’’ Applied Economics, 1994, 26, pp. 342–357. The United Nations, Economic and Social Survey of Asia and the Pacific. Bangkok: Economic and Social Commission for Asia and the Pacific, 1997. US Department of Commerce, Economics and Statistics Administration, and Bureau of Census, 1980–1996, Statistical Abstract of the United States. World Bank, The East Asian Miracle. Washington, DC: World Bank, 1993. World Bank, World Development Report. Washington, DC: World Bank, 2002. Yao, S., ‘‘Economic Development and Poverty Reduction in China Over 20 Years of Reforms,’’ Economic Development and Cultural Chang, 2000, 48(3), pp. 447–474, April. Yue, C.S., ‘‘Trade, Foreign Direct Investment and Economic Development of Southeast Asia,’’ The Pacific Review, 1999, 12(2), pp. 249–270.

APPENDIX This appendix provides the details on data, methods of regression and interpretation of results for Table 11.

A.1. Model Specifications The GDP Equation. The equation for GDP is a Cobb–Douglas production function, which includes the basic inputs of labour and capital, the internal environmental factors (human capital, location, and transportation), and the external environmental factors (export, FDI, and real exchange rate). The final form of the estimated equation is expressed in equation (A1).

92

S. Yao and Z. Zhang

lnðGDPÞit ¼ 0 lnðlabourÞit þ 1 lnðcapitalÞit þ 2 lnðhumancapitalÞit þ 3 lnðexchangeÞt þ 4 lnðFDIÞit þ 5 lnðexportÞit þ 6 lnðtransportÞit þ 7 ðeastÞ þ 8 D9295 þ 9 ðtimeÞ þ eit

ðA1Þ

The subscript (it) denotes province (i ¼ 1, 2, . . ., 30) and year (t ¼ 1978, 1979, . . . , 1995). The error term (eit) is assumed to be a stochastic white noise. All the variables are calculated in 1990 constant prices. Price indices are province specific. All the data required are available from the official statistics (NBS, 1996) except physical capital stock, which is derived from a special procedure to be discussed below. It is difficult to define a variable that represents human capital. In most empirical studies, it is approximated by the secondary school enrolment rate of the relevant aged population (Islam, 1995). In this study, however, we find that the secondary enrolment rate does not perform well in our model. We suspect that the main difference in education between provinces and over time is the number of students enrolled in higher education (Fleisher and Chen, 1997, use university graduates/population and a proxy for human capital). As a result, we use the ratio of the number of students enrolled in higher education over the number of students enrolled in secondary education to represent the changes in human capital. This ratio reflects the propensity that secondary school students can be enrolled in higher education. The rapid development in higher education reflects the rapid economic growth in China over the data period. In addition, the variation in higher education also represents the variation in the economic activities among the provinces. The values of exports and FDI are provided in US dollars in the official statistics. Since they are measured in US dollars, most economic analysts do not bother to deflate the values in current prices into values in constant prices (e.g. Liu et al., 1997). We feel that it is important to conduct an appropriate deflation. One relevant deflator is the US consumer price index. Therefore, we deflate the values of exports and FID in current prices (US dollars) by the US consumer price index based in 1990. We then convert the deflated values into equivalent values in RMB by multiplying the value with the official exchange rate in 1990 ($1 ¼ RBM 4.784). In regression analysis, it is not important to convert the values in US dollars into values in RBM because all the values will be logged. However, since all the other variables in the model are measured in RBM, it is useful to change these two valuables into RBM as well. In the earlier years of the data period, many provinces had no FDI. This means that many observations have zero values and cannot be logged. To overcome this problem, we replaced all the zero values with a very small but non-zero value. The exchange rate is the real exchange rate as calculated in Table 10. This variable is time-variant but location-invariant as all the provinces faced the same foreign exchange rate. Ideally, when we calculate the real exchange rate we should use the exchange rates and the price indexes of China’s main trading partners. However, since RBM follows the US dollar very closely (although RMB is not pegged to the dollar), we only use the dollar exchange rate and the US price index to calculate the real exchange rate. Transportation is measured as the equivalent mileages of railways, highways and waterways per 1000 sq km. The highway is the dominant means of transpor-


93

tation in terms of mileage. The ratios of the lengths of railways, highways and waterways are 1.00/16.84/1.90 at the national level. The simplest way to measure transportation is to add the total lengths of these three different means of transportation (e.g. Fleisher and Chen, 1997; Liu et al. 1997) However, the transportation capacity of one mile of railway is different from that of one mile of highway or waterway. As a result, it is necessary to convert railways and waterways into equivalent highways. The conversion ratios are derived from the volumes of transport per mile by each of the three means of transportation. At national average, the conversion ratios are 4.27/1.00/1.06. In other words, we multiplied railways by 4.27 and waterways by 1.06 to convert them into the equivalent length of highways. This method of conversion may not be perfect, as the relative capacity of different transportation means may not be the same in different provinces. However, any possible conversion errors may be small because highways account for a predominant proportion of the total transportation volume. One possible way to correct the conversion errors is to instrument the length of transportation. This is done through the DPD (dynamic panel data) estimation as discussed below. Finally, capital stock has to be calculated using the flows of investments according to equation (A2). Kitþ1 ¼ ð1 ÞKit þ

Iitþ1 pK i

ðA2Þ

where K is physical capital stock, I investment, pK the price index for investment, with the subscript for province (i) and time (t), and is a depreciation rate of capital stock. To work out the time series of capital stock, we need to know its initial level and the average depreciation rate of capital. Both of these values are unknown so it requires some reasonable assumptions before capital stocks are derived. The first assumption is that the initial capital stock in each province is equal to twice the level of initial GDP; that is, the capital stock in 1978 is equal to twice the GDP in that year for each province. This assumption implies that the capital share was 50% in 1978. This share is similar to the capital elasticity of output. Given that the Chinese economy was not quite capital-intensive, a capital elasticity of 0.5 is not unreasonable. The second assumption is that the average depreciation rate of capital stock is 7.5%. This rate is arbitrary but it implies that the average life of capital equipment is 13.3 years. Finally, since the price index of capital stock is not available, the GDP deflator is used to deflate investments. The Export Equation. Export is determined by GDP and the real exchange rate. There may be some other factors that can affect exports but these factors can be represented by a lagged dependent variable. The inclusion of a lagged dependent variable can also resolve the serial correlation problem that exists in the model without the lagged term (see below for more explanation). The location dummy for the eastern region and a dummy for the boom years in 1992–95 are also included in the regression. The final specification is given in equation (A3). lnðexportÞit ¼ 0 þ 1 lnðGDPÞit þ 2 lnðexchangeÞit þ 3 ðeastÞ þ 4 D9295 þ 5 lnðexportÞit1 þ eit

ðA3Þ

All these variables and the required data have been given above for equation (A1).

94

S. Yao and Z. Zhang

The FDI Equation. The basic explanatory variables for FDI are GDP, effective wages, real exchange rate, transportation and human capital. The two dummies for the eastern region and the period 1992–95 are also included. The lagged dependent variable is added to resolve the problem of serial correlation and to account for the effects of some other variables that may be omitted from the model. The final specification is given in equation (A4). lnðFDIÞit ¼ 0 þ 1 lnðGDPÞit þ 2 lnðwagesÞit þ 3 lnðexchangeÞit þ 4 lnðtransportÞit þ 5 lnðhumancapitalÞit þ 6 ðeastÞ þ 7 D9295 þ 8 lnðFDIÞit1 þ eit

ðA4Þ

The data and explanations for FDI, GDP, exchange, transportation, human capital and the two dummy variables are exactly the same as those given for equation (A1). The wages variable is the effective wage rate, which is the ratio of real wages divided by real GDP per employee in the respective year and province. Real wages are wages in current prices divided by the consumer price index for each province. All the explanatory variables except the wages variables are expected to have a positive effect on FDI. The effect of the wages variable on FDI may not be decisive. On the one hand, wages reflect the cost of production and should have a negative effect on FDI. On the other hand, wages may reflect the quality of labour and should have a positive effect on FDI. This is why real wages have to be divided by labour productivity (we call this ratio effective wages). The net effect of the wages variable will depend on the interaction of its negative effect as a cost of production and its positive effect as a measurement of labour quality.

A.2. Estimation There are three potential econometric problems if the equations for GDP, FDI and export are estimated separately using the OLS technique. First, the three equations are unlikely to be independent of each other and, ideally, they should have to be estimated in a simultaneous system. However, they cannot be run in a simultaneous system because there are not enough observations (and degrees of freedom) to generate reliable and robust results. As a compromise, they are run in a seemingly unrelated system. Secondly, equation (A1) has a simultaneity problem in OLS as some of the explanatory variables (exports, capital stock, labour and human capital) are unlikely to be exogenous. To solve this problem, we use GMM estimators in DPD to run the regression in a dynamic system (see Arellano and Bond, 1998, for more explanation). More precisely, capital and export are treated as endogenous variables, whose levels dated (t 2, t 3) are used as instruments for the difference equations, and differences dated (t 1) are used for the levels equation. We cannot treat any other right-hand side variables in the same way due to data restriction. As a result, labour, human capital, the real exchange rate and transportation are treated as exogenous variables, but to overcome the problem of multicollinearity, they are all instrumented. Thirdly, the export and FDI equations exhibit a serial correlation problem. To solve the problem, it is necessary to add a lagged dependent variable to the models.


95

A.3. Results and Interpretations The M1 and M2 are N(0, 1) distributed tests for first and second order serial correlation (see Arellano and Bond, 1998). In the GDP equation, there is no serial correlation without adding the lagged dependent variables. In the export and FDI equations, serial correlation is removed after adding the lagged dependent variables. Sargan is the p-value from a test of the validity of the instruments. The test result cannot reject the validity of instruments in the GDP equation. The Wald test gives the p-value for joint significance. The results are derived from the one-step estimates with robust test statistics. The goodness-of-fit (R2) indicates that more than 90% of the variation in the dependent variable is explained by the explanatory variables in all regressions. In the GDP equation, the t-statistics show that the estimated coefficients of labour, physical capital, human capital, the real exchange rate, exports, the dummy variable for the period 1992–95 and the time trend are significant at or below the usual 5% level. The estimated coefficients of FDI and transportation are significant only at the 10% level. In the export equation, all the estimated coefficients are significant at the 5% level. In the FDI equation, the estimated coefficients of GDP, the real exchange rate, the two dummy variables and the lagged dependent variable are significant at or below the 5% level. The estimated coefficients of human capital and wages have the expected sign but they are not significant.