EXPORT-ORIENTED [Asian Economic Journal 2000, 1998, FOREIGN Vol. 14 12 No.INVESTMENT: 3] 1] CHINA 55 Costs and Benefits of Export-Oriented Foreign Investment: The Case of China* Yun-Wing Sung Chinese University of Hong Kong The costs and benefits of export-oriented FDI have been discussed by Helleiner (1973, 1998), Watanabe (1972), Sharpston (1975), and others. Processed exports generated from FDI have constituted over half of the exports of Singapore, Malaysia, Philippines, and China. Despite the importance of processed exports, empirical studies of their costs and benefits are difficult due to lack of data, especially on transfer earnings. Data on the division of benefits between the source and host countries are scarce and unreliable. This paper examines the costs and benefits of export-oriented foreign investment for China. China has been highly successful in exporting and in attracting FDI, especially export-oriented FDI from Hong Kong. Since 1993, China has become the second largest recipient of FDI in the world after the US, and Hong Kong has become the world’s fourth largest source of FDI after the US, UK, and Germany. China’s processed exports are largely re-exported via Hong Kong. As a result, good data on the total valueadded of processed exports for Mainland China and for Hong Kong are available. It is found that the rate of value-added for Mainland China is relatively low compare with that for Hong Kong, indicating transfer pricing and absence of linkages in the mainland. This appears to be due to the rigidity of China’s economic system which hampers backward and forward linkages. The mainland is thus dependent on Hong Kong for many services in the value-added chain. However, the rate of value-added for China has increased substantially since 1996, indicating an increase in both backward and forward linkages. I. Introduction Export-oriented FDI (Foreign Direct Investment) has received considerable attention in the trade and development literature as it has been important in the growth of many NIEs (Newly Industrializing Economies). In the early 1970s, Helleiner (1973) pointed out that multinational firms were already deeply involved * This a revised version of a paper prepared for the 6th Convention of the East Asian Association, Kitakyushu, 4–5 September 1998. The research is partially supported by a grant from the Research Grants Council of Hong Kong (RGC Ref. No. CUHK 165/94H). The author wishes to thank an anonymous referee for very helpful suggestions and comments. ASIAN ECONOMIC JOURNAL 56 Table 1 Share of Foreign Affiliates in Total Exports of East Asian Host Countries (per cent) Host Country Hong Kong (1984) South Korea (1978) Taiwan (1986) Singapore (1988) Malaysia (1986) Philippines (1983) Thailand (1980) Primary Sector – – 18.5 – 32.6 – – Secondary Sector Tertiary Sector Electrical/ electronic products Chemical products Motor vehicles Total – – 43.6 – – – – – – 30.8 – – – – – – – – – – – 16.5 24.6 18.5 86.0 51.2 51.5 37.3 – – 3.6 – 15.9 – – Note: Year in brackets refers to year of data. Source: UNTCMD (1992, Table Annex 9). in the expansion of manufactured exports from low-income countries, and many American electronics, machinery, and transport equipment firms shipped parts to low-wage countries for labour-intensive assembly or processing. In many cases, such manufacturing was undertaken in export-processing zones or similar enclaves. For the first-tier East Asian NIEs of Hong Kong, Singapore, South Korea and Taiwan, the share of exports of foreign affiliates was not high except for Singapore (Table 1). However, for the next tier of export-oriented East Asian exporters including Malaysia, Philippines and Thailand, foreign affiliates accounted for between 37% and 52% of their exports in the secondary sector. Foreign investment also accounted for over half of China’s exports, as will be seen later. The costs and benefits of export-oriented FDI have long been discussed in the literature. The seminal article of Helleiner (1973) was followed by Sharpston (1975), and Helleiner (1988). Helleiner (1973) distinguished between short-run allocative effects and long-run growth effects of export-oriented FDI. I.1 Allocative effects While import-substituting FDI may lead to immiserization due to tariff protection, export-oriented FDI will not result in negative value-added (at international prices) as the outputs are sold in the world market and most of the inputs are imported at international prices (Helleiner, 1973, p. 14). Processed exports usually have a low rate of value-added because most of the inputs are imported. This may not be inefficient as comparative advantage and specialization may require substantial amounts of imported inputs, especially in small economies. However, the profits and value-added in processed exports may be underreported due to transfer pricing. It may be difficult to prevent transfer pricing in EXPORT-ORIENTED FOREIGN INVESTMENT: CHINA 57 processed exports because the parent firm controls both the imported inputs and outputs, generating twice as many opportunities for transfer pricing as does import-substituting manufacturing (Helleiner 1973, p. 13). Moreover, ascertaining the price of specialized components commissioned by the parent firm can be very difficult for the simple reason that the parent firm may be the only outlet for such components. It should be noted that transfer pricing leads to losses only in government revenue. The host country earns at least the value of the wage bill from processed exports. As the opportunity cost of unskilled labour is very low in developing economies because of unemployment, the use of unemployed labour to earn scarce foreign exchange is quite attractive (Helleiner, 1973, p. 14). However, host countries usually provide infrastructure to attract FDI, and the large number of low-wage countries competing for ‘foot-loose’ manufacturing implies that host countries have little bargaining power. It is possible for the costs of infrastructure to exceed the benefits (Helleiner, 1973, p. 14). I.2 Growth effects Besides the above cost-benefit calculus which deals with allocative efficiency that is static, Helleiner (1973) also discussed dynamic growth effects, including ‘linkage’, ‘learning’, and ‘dependence’ effects. Backward linkage may fail to develop because the exporting firm enjoys special concessions for imported inputs. Domestic intermediate inputs into the Kaohsiung Export Processing Zone in Taiwan were only 4% of the value of its exports in 1969 (Helleiner, 1973, p. 16). Processed exports, which are intensive in unskilled labour, do not seem to promise much in terms of learning effects. However, besides modern factory discipline, management skills are also transferred, as management of an indigenous labour force would not be undertaken by expensive expatriates for long. Moreover, the rise in income from processed exports usually raises the demand for education and acquisition of more sophisticated skills. The first-tier Asian NIEs have been able to upgrade their skills tremendously, showing that labourintensive exporting is not a blind alley. Processed exports may imply dependence on foreign investors. Processed exports may constitute an enclave as they are dependent neither on local resources nor on local markets. The host country has very little control on sources of inputs and market outlets. The bargaining strength of host country is low as processed exports are foot-loose. I.3 Empirical studies on costs and benefits of processed exports Despite the importance of processed exports, empirical studies of their costs and benefits are difficult to carry out because of lack of data. Data on transfer earning are very difficult to obtain. ASIAN ECONOMIC JOURNAL 58 Data on exports of foreign affiliates in the host country can usually be obtained from industrial censuses. However, exports of foreign affiliates include both processed exports which use largely imported inputs, and also exports that use largely local inputs. There are usually no data to distinguish the two kinds of exports. Data on the benefits of exported-oriented FDI for the source country are even more difficult to obtain. The earnings from FDI are available from balance of payment accounts, but such earnings include export-oriented as well as importcompeting FDI and further disaggregation is usually not possible. Moreover, data on earnings from FDI are usually quite unreliable. Transfer earnings further exacerbate the problem. Data on the division of earnings between the host and source countries are all but non-existent. The case of China is interesting as it has been highly successful in exporting and in attracting FDI, especially export-oriented FDI. From 1978 to 1997, China’s exports have grown at an average annual rate of 16.7%.1 Processed exports have increased from nothing to over half of China’s exports. China’s rank in world trade rose from the 32nd in 1978 to 11th in 1992, and rose to 10th in 1997. Since 1993, China has been the second largest recipient of FDI in the world after the US. Moreover, good data on the division of earnings between the host and source economies are available for China, permitting an insightful study of the costs and benefits of processed exports. II. Forms of processed exports in China The growth in China’s manufacturing exports was especially rapid after 1984, when the State Council approved two schemes allowing duty-free import of components and raw materials for use in export industries. The first scheme is called ‘process and assembling’, and the second scheme is called ‘process with imported materials’. In this paper, the first scheme is called contractual processing and the second scheme is called ordinary processing. II.1 Contractual processing In contractual processing, Chinese firms process raw materials or components supplied by foreign firms for a processing fee. The processed outputs belong to the foreign firms and are exported by them. The foreign firms usually provide the required machinery, product design and technical assistance. The foreign funds involved in contractual processing usually involve OFI (Other Foreign Investment)2 rather than FDI. Unlike FDI, OFI does not give 1. China’s exports grew by a mere 0.5% in 1998 because of the Asian Financial Crisis (AFC). However, China’s exports rebounded sharply in the second half of 1999. The adverse effects of the AFC appear to be over. 2. The other forms of OFI are issuing shares, leasing and compensation trade. EXPORT-ORIENTED FOREIGN INVESTMENT: CHINA 59 legal control of the enterprise to the foreign investor. According to Chinese terminology, FDI includes investment in the three types of ‘foreign-invested enterprises’ (FIEs), namely, fully foreign-owned enterprises, equity joint ventures and contractual joint ventures. The distinction between FDI and contractual processing is not sharp because the foreign partner in contractual processing often has de facto though not de jure control of the Chinese firm. However, the foreign funds involved in contractual processing constitute commercial credit rather than FDI because the Chinese partner legally controls the operation and usually pays for foreign machinery and technical assistance with labour services used in making goods for the foreign partner. In this article, the term ‘foreign investment’ includes FDI and ‘other foreign investment’ but excludes portfolio investment and loans. According to Chinese regulations on foreign investment, the output of contractual processing should all be exported, whereas FIEs are permitted to sell a certain portion of their output in the domestic market. Firms undertaking contractual processing tend to be small and labour-intensive. FIEs are usually bigger and they involve long-term commitment. Contractual processing was popular in the early days of the open era. As a result of China’s maturing investment environment and the attraction of China’s domestic market, many contractual processing operations have been converted to FIEs in the 1990s. This has led to a decline in OFI and an acceleration in FDI in China (Sung et al., 1995, p. 78). II.2 Ordinary processing The bulk of ordinary processing is undertaken by FIEs. As in contractual processing, ordinary processing also involves the processing of imported materials and components for export. Unlike contractual processing, the outputs and imported inputs are owned by the firms in China rather than firms outside China. Instead of earning a processing fee as in the case of contractual processing, the firms involved sell the processed exports for a profit. However, as these firms are often subsidiaries of foreign firms, the prices of inputs and outputs are often determined by the foreign parents. The distinction between contractual processing and ordinary processing is thus largely legal rather than substantive. II.3 Growth of processed exports China’s high tariffs and deep exemptions for imported inputs used in export processing create strong incentives for this kind of trade. Detailed statistics on process exports are available from China’s Customs Administration because their imported inputs are entitled to tariff exemption. Table 2 shows the growth of China’s processed exports, including both contractual processing and ordinary processing. Both categories are large and both have grown rapidly. In recent