复旦大学：《发展经济学》阅读材料与文献_MACHINERY AND CHINA'S NEXUS OF FOREIGN TRADE AND ECONOMIC GROWTH.pdf

734 D. Lo and TM h chan Adopting an analytical approach that is typical of most available studies on the topic, Lardy(1992, p. 691), for one, frames his question this way Has the expansion of foreign trade been achieved largely through a state-driven export strategy in which sales on the international market are viewed simply as a means of financing much needed imports? This would imply, as in the pre- reform era, that exports were selected without much consideration of China's comparative advantage and that, as a result, expanding exports might contribute This approach is strictly in line with the standard neoliberal theory about export-led growth -that is, the thesis of trade regime neutrality, where the clear logical link between allocative efficiency and free trade rests with the principle of(endowment determined) comparative advantage(see, e.g., World Bank, 1987). Hence, the isolation of allocative efficiency as the only channel through which foreign trade could contribute to growth centralizes the question as to whether Chinas trade-growth nexus has been largely accounted for by a trade regime that is heading toward neutrality, for which Lardy gives an affirmative answer We find this approach problematic, even misleading in terms of policy implications We do not dispute the observation that, to date, the expansion of China's exports has been largely accounted for by labour-intensive products that accord with its ' given comparative advantage. Nor do we dispute the judgement that the exports expansion has contributed to economic growth. What we want to argue is: ()that the expansion of labour-intensive exports has had its intrinsic problems which could retard economic growth; (ii) that the main channel through which exports contribute to reform era, is largely unrelated to allocative efficiency; and (iii) that in recent years the export momentum, and hence the contribution to growth, has been increasingly sustained by exports that do not accord with China's'given'comparative advantage Taken together, our arguments imply that the Chinese nexus of foreign trade and economic growth contradicts, rather than validates, the neoliberal thesis of trade gime neutrality which explicitly rules out any form of strategic integration of late developing countries with the world market. I Our focus on the machinery sector serves to substantiate the above arguments. Thi ector encompasses most of China's fast-growing industries over the reform era, with the electronics industry being a representative one. During 1980-95, the machinery sector as a whole grew at an average annual rate of 17 per cent and increased its output share in Chinese industry from 17. 4 to 23.8 per cent, while the electronics industry alone grew at a rate of 23 per cent and increased its output share from 1.5to 3.7 per cent. Perhaps more important, the development of the electronics industry and the machinery sector in general has been especially illustrative of the contradictions is-a-vis relative scarcities--among industries'. As for strategic integration with the world market, it recisely to a situation where the trade regime involves selective discrimination: providing variable effective ates of protection, and export incentives, for different categories of goods. Thus, the division between the wo is essentially about the international specialization of an economy: should it strictly follow its comparative advantage, or should it attempt to create competitive advantage over specific industries 01998 John Wiley Sons, Ltd J.mnt.Dev.10,733-74901998)

Adopting an analytical approach that is typical of most available studies on the topic, Lardy (1992, p. 691), for one, frames his question this way: Has the expansion of foreign trade been achieved largely through a state-driven export strategy in which sales on the international market are viewed simply as a means of ®nancing much needed imports? This would imply, as in the prereform era, that exports were selected without much consideration of China's comparative advantage and that, as a result, expanding exports might contribute little or nothing to economic growth. This approach is strictly in line with the standard neoliberal theory about export-led growth Ð that is, the thesis of trade regime neutrality, where the clear logical link between allocative eciency and free trade rests with the principle of (endowmentdetermined) comparative advantage (see, e.g., World Bank, 1987). Hence, the isolation of allocative eciency as the only channel through which foreign trade could contribute to growth centralizes the question as to whether China's trade-growth nexus has been largely accounted for by a trade regime that is heading towards neutrality, for which Lardy gives an armative answer. We ®nd this approach problematic, even misleading in terms of policy implications. We do not dispute the observation that, to date, the expansion of China's exports has been largely accounted for by labour-intensive products that accord with its `given' comparative advantage. Nor do we dispute the judgement that the exports expansion has contributed to economic growth. What we want to argue is: (i) that the expansion of labour-intensive exports has had its intrinsic problems which could retard economic growth; (ii) that the main channel through which exports contribute to growth has been the ®nancing of technology imports and hence, just like the prereform era, is largely unrelated to allocative eciency; and (iii) that in recent years the export momentum, and hence the contribution to growth, has been increasingly sustained by exports that do not accord with China's `given' comparative advantage. Taken together, our arguments imply that the Chinese nexus of foreign trade and economic growth contradicts, rather than validates, the neoliberal thesis of trade regime neutrality which explicitly rules out any form of strategic integration of late developing countries with the world market.1 Our focus on the machinery sector serves to substantiate the above arguments. This sector encompasses most of China's fast-growing industries over the reform era, with the electronics industry being a representative one. During 1980±95, the machinery sector as a whole grew at an average annual rate of 17 per cent and increased its output share in Chinese industry from 17.4 to 23.8 per cent, while the electronics industry alone grew at a rate of 23 per cent and increased its output share from 1.5 to 3.7 per cent. Perhaps more important, the development of the electronics industry and the machinery sector in general has been especially illustrative of the contradictions 1 Pack and Westphal (1986, p. 88) gives a standard de®nition of trade regime neutrality: `that policies should not selectively discriminate Ð that is, for tradeables, vis-aÁ-vis world prices, and for non-tradeables, vis-aÁ-vis relative scarcitiesÐ among industries'. As for strategic integration with the world market, it refers precisely to a situation where the trade regime involves selective discrimination: providing variable eective rates of protection, and export incentives, for dierent categories of goods. Thus, the division between the two is essentially about the international specialization of an economy: should it strictly follow its comparative advantage, or should it attempt to create competitive advantage over speci®c industries. #1998 John Wiley & Sons, Ltd. J. Int. Dev. 10, 733±749 (1998) 734 D. Lo and T. M. H. Chan

Machinery and China's Trade-Growth Nexus 735 embodied in the impact of trade on growth and the spectacular expansion of non- labour-intensive exports in recent years This anized in owing way. 2 presents evidence indicating that Chinas exports expansion is not reducible to the outcome of its ' natural comparative advantage, but is increasingly underpinned by the competitiveness of its industry. Section 3 turns to the growth side, presenting a schematic account of the growth pattern that centres around technological imports and development. On this basis, Section 4 offers a discussion of possible factors that have contributed to the formation of competitiveness of the dynamic industries. Section 5 concludes the pap 2 CHINAS MANUFACTURING EXPORTS: COMPARATIVE OR COMPETITIVE ADVANTAGE? Is Chinas exports expansion in the reform era reflecting mainly the realization of its given'or'natural comparative advantage? Lardy's affirmative answer is based on the observation that manufacturing products, which are considered to be more labour intensive in production than primary products, have accounted for a fast increasing share of total exports. Similarly, Panagariya(1993)corroborates by noting that within total exports light industrial products have shot up while heavy industrial products have declined, and light industry is considered to be labour-intensive relative to heavy industry Moving down to more disaggregated levels of the commodity composition of Chinas exports, however, the picture becomes complex. The data in Table I highlight the fact that, along with the increasing share of manufactures in total merchandise exports, there is an even more conspicuous expansion of the share of machinery and transport equipment in total manufacturing exports, from 8 per cent in 1988 to Table 1. Chinas manufacturing exports and imports (SITc classification), 1980-96 Sources:zIN(1996.pp.581-582);zZ(1997,p.128) (a)1980(b)1988c)1996(b)/(a)c)/(b) Exports(USS100 m Total merchandise (SITC 0-9) 1511 Total manufactures(SITC 5-8) 1291 Machinery and transport equipment 2.61 (SITC 7) Imports(USS100 m) T 1388 77 2.51 Total manufactures 131 452 3.45 2.51 Machinery and transport equipment 16 48 3.27 3.28 Total manufactures 0.69 Machinery and transport equipment 0.16 0.64 Notes: Strict king, manufactures are defined as SITC (standard international trade categorie category 68(non-ferrous metals)are not always available. Note that startin reviously under category 9(commodities not classified elsewhere)have been re- ous categories from 0 to 8. Hence, there is some degree of non-comparability of da hat year. 01998 John Wiley Sons, Ltd J.mnt.Dev.10,733-749(1998)

embodied in the impact of trade on growth and the spectacular expansion of nonlabour-intensive exports in recent years. This paper is organized in the following way. Section 2 presents evidence indicating that China's exports expansion is not reducible to the outcome of its `natural' comparative advantage, but is increasingly underpinned by the competitiveness of its industry. Section 3 turns to the growth side, presenting a schematic account of the growth pattern that centres around technological imports and development. On this basis, Section 4 oers a discussion of possible factors that have contributed to the formation of competitiveness of the dynamic industries. Section 5 concludes the paper. 2 CHINA'S MANUFACTURING EXPORTS: COMPARATIVE OR COMPETITIVE ADVANTAGE? Is China's exports expansion in the reform era re¯ecting mainly the realization of its `given' or `natural' comparative advantage? Lardy's armative answer is based on the observation that manufacturing products, which are considered to be more labourintensive in production than primary products, have accounted for a fast increasing share of total exports. Similarly, Panagariya (1993) corroborates by noting that within total exports light industrial products have shot up while heavy industrial products have declined, and light industry is considered to be labour-intensive relative to heavy industry. Moving down to more disaggregated levels of the commodity composition of China's exports, however, the picture becomes complex. The data in Table 1 highlight the fact that, along with the increasing share of manufactures in total merchandise exports, there is an even more conspicuous expansion of the share of machinery and transport equipment in total manufacturing exports, from 8 per cent in 1988 to Table 1. China's manufacturing exports and imports (SITC classi®cation), 1980±96. Sources: ZTN (1996, pp. 581±582); ZTZ (1997, p. 128). (a) 1980 (b) 1988 (c) 1996 (b)/(a) (c)/(b) Exports (US$100 m) Total merchandise (SITC 0±9) 181 475 1511 2.62 3.18 Total manufactures (SITC 5±8) 90 331 1291 3.68 3.90 Machinery and transport equipment (SITC 7) 8 28 353 3.50 12.61 Imports (US$100 m) Total merchandise 200 553 1388 2.77 2.51 Total manufactures 131 452 1134 3.45 2.51 Machinery and transport equipment 51 167 548 3.27 3.28 Exports±imports ratio Total merchandise 0.91 0.85 1.09 Total manufactures 0.69 0.73 1.14 Machinery and transport equipment 0.16 0.17 0.64 Notes: Strictly speaking, manufactures are de®ned as SITC (standard international trade categories) 5 to 8 less 68. But data of category 68 (non-ferrous metals) are not always available. Note that starting from 1992 items that were previously under category 9 (commodities not classi®ed elsewhere) have been re-classi®ed under the various categories from 0 to 8. Hence, there is some degree of non-comparability of data before and after that year. #1998 John Wiley & Sons, Ltd. J. Int. Dev. 10, 733±749 (1998) Machinery and China's Trade±Growth Nexus 735

736 D. Lo and.h chan 27 per cent in 1996. It is of course well known that machinery and transport equip- ment is a broad category encompassing both labour-intensive and capital-intensive products. Yet the critical issue is whether these products are more, or less, labour intensive than other manufactures. In this regard, the fact that major world exporters of machinery and transport equipment are unexceptionally developed countries, with the top five(Japan, the US, Germany, France and the UK)accounting for almost 60 per cent of the world total in 1994 (World Trade Organization, International Trade 95, p. 102), should be taken for reference Moving further down, Table 2 gives more disaggregated data classified under Chinas statistical system. Note that within the total of manufacturing exports the share of mechanical and electronic products increased from 19 per cent in 1988 to 37 per cent in 1996. Put another way, in this period, the former category increased by hree times while the latter by seven times. Meanwhile, electronic exports increased by nine times while computer exports, which are a further disaggregated category under electronics, by 59 times. Again, it is noted that major word exporters of electronics (and computer products)are either developed or newly industrializing economies which conceivably have international comparative advantage very different from Chinas. Thus what Table 1 and Table 2 indicate at least at an intuitive level is an export pattern where more capital-intensive (or more high-tech)industries tend to expand faster. 2 p D): and (1 April 1997, p 1): GJSB (Il January (a)1988 (b)1996 (b)/(a) Exports(USS100 m 331.10 129141 Mechanical and electronic products 61.50 482.10 Electronic products 19.15 Computer products l08 Imports(USS100 m) 452.07 Mechanical and electronic products 207.00 Electronic products 79 3 echanical and electronic products This observation is in line more comprehensi Chinas exports as a whole e of descriptive ession analyses, Jiang (1993, ch. 11). for instance. identifies three of Chinas export ce in the 1980s; namely: (i) within and electronic products which are characterized by high capita and technology intensity relative to others have experienced the fastest expansion; ()within labour- and textile and clothing, relatively me he fastest exports expansion; and(iii) for most sectors of Chir mainly those with the highest capital and technology intensity in each sector. In view of our Table I and Table 2, it appears that feature()is even more pronounced in the 1990s 01998 John Wiley Sons, Ltd J.mnt.Dev.10,733-74901998)

27 per cent in 1996. It is of course well known that machinery and transport equipment is a broad category encompassing both labour-intensive and capital-intensive products. Yet the critical issue is whether these products are more, or less, labourintensive than other manufactures. In this regard, the fact that major world exporters of machinery and transport equipment are unexceptionally developed countries, with the top ®ve (Japan, the US, Germany, France and the UK) accounting for almost 60 per cent of the world total in 1994 (World Trade Organization, International Trade 1995, p. 102), should be taken for reference. Moving further down, Table 2 gives more disaggregated data classi®ed under China's statistical system. Note that within the total of manufacturing exports the share of mechanical and electronic products increased from 19 per cent in 1988 to 37 per cent in 1996. Put another way, in this period, the former category increased by three times while the latter by seven times. Meanwhile, electronic exports increased by nine times while computer exports, which are a further disaggregated category under electronics, by 59 times. Again, it is noted that major word exporters of electronics (and computer products) are either developed or newly industrializing economies, which conceivably have international comparative advantage very dierent from China's. Thus what Table 1 and Table 2 indicate, at least at an intuitive level, is an export pattern where more capital-intensive (or more `high-tech') industries tend to expand faster.2 2 This observation is in line with some more comprehensive studies of the commodity composition of China's exports as a whole. Based on a range of descriptive and regression analyses, Jiang (1993, ch. 11), for instance, identi®es three `anomalies' of China's export performance in the 1980s; namely: (i) within total manufacturing exports, mechanical and electronic products which are characterized by high capital and technology intensity relative to others have experienced the fastest expansion; (ii) within labourintensive sectors like light industry and textile and clothing, relatively more capital-and technologyintensive products have experienced the fastest exports expansion; and (iii) for most sectors of Chinese industry, the exported products are mainly those with the highest capital and technology intensity in each sector. In view of our Table 1 and Table 2, it appears that feature (i) is even more pronounced in the 1990s. Table 2. China's manufacturing exports and imports (alternative classi®cation), 1988±96. Sources: Lo (1997, ch. 1); ZTN (1996, pp. 581±582); ZTZ (1997, p. 127); ZDB (4 March 1997, p. 1); and (1 April 1997, p. 1); GJSB (11 January 1997, p. 1). (a) 1988 (b) 1996 (b)/(a) Exports (US$100 m) Total manufactures 331.10 1291.41 3.90 Mechanical and electronic products 61.50 482.10 7.84 Electronic products 19.15 184.10 9.61 Computer products 1.08 64.70 59.90 Imports (US$100 m) Total manufactures 452.07 1133.98 2.51 Mechanical and electronic products 207.00 613.60 2.96 Electronic products 52.79 158.30 3.00 Computer products 6.90 34.30 4.97 Exports±imports ratio Total manufactures 0.73 1.14 Mechanical and electronic products 0.30 0.79 Electronic products 0.36 1.16 Computer products 0.16 1.89 #1998 John Wiley & Sons, Ltd. J. Int. Dev. 10, 733±749 (1998) 736 D. Lo and T. M. H. Chan

Machinery and Chinas Trade-Growth Nexus 737 A further observation that can be made from table i and table 2 concerns the exports-imports ratio. It can be seen that, between 1988 and 1996, for the indicated trade categories the general trend is the larger the increase in the ratio the faster the expansion of exports. Whilst the exports-imports ratio of all manufactures increased by 41 percentage points, that of mechanical and electronic products, electronic products alone and computer products increased by 49, 80 and 173 percentage points respectively. To some extent, this development is related to the growth of foreign funded processing or assembling operations in China. However, on a priori grounds one cannot rule out the validity of inferring from it that the Chinese economy is heading towards a pattern of intra-industry, rather than inter-industry, international specialization. This presents yet another anomaly for the standard theory of compara tive advantage Sceptics of course can go on to argue that the intra-industry specialization might still be divided in a way that China is engaged mainly in the labour-intensive opera tions of each industry, even if the industry concerned is on average characterized by high capital and technology intensity. Constrained by the availability of data, it is difficult to investigate into the trade and production characteristics of the intra- industry division of labour. Nevertheless, focusing on the output composition Chinas electronics industry, Figure I presents a trend of increasing vertical integr tion which runs counter to such argument. Between 1988 and 1995, the output share of industrial equipment and components -both characterized by higher capital and technology intensity-increases by 14 and 5 percentage points, respectively. As result, the output composition in 1995 contrasts sharply with that of the mid-1980 hen the Chinese electronics industry was overwhelmingly skew to the production of downstream consumer produc One possible way of explicitly examining the factor intensity of the trade categories is to check the respective capital-labour ratios on the production side. Lardy(1992) connection, it is of note that the expansion of electronics exports, and ex of mechanical in general, has been a salient feature of East Asian industrialization, both in the four earlier of Hong Kong, South Korea, Singapore and Taiwan, and the second-tier NICs(Malays Thailand, Indonesia and the like). Chinas expanding exports of these products might therefore seem to be just in line with the ' normal pattern of inter disagreement among scholars as to whether the East Asian experience is indeed normal in the world cale-particularly in the case of electronics production, where outside this handful number of countries orld has been excluded from engaging Istitutional environment (especially the role of state policy vis-a-vis the world market) within which electronics production has developed in East Asia is even a matter of disagreement( Ernst and OConnor 4 Citing a World Bank study, Lardy (1994, p. 32), for instance, infers that( Chinas)exports of telecom- unications equipment and domestic electrical equipment, which might appear to be more capital. intensive, are based overwhelmingly on processing and assembly-type activity. Thus they too are labor. intensive since imported components account for about four-fifths of the value of exports in total merchandise exports: 7I per cem r orts is much higher in mechanical and elec is true that the ratio of pr roducts than in processing exports, measured as the ratio of net-to-gross processing exports, is also much higher in the her than in the latter: 43 per cent versus 17 per cent in the same year. Moreover, both of the two have ad substantial increases in recent years: for total processing exports, the ratio increased to 21 per cent in 26 per cent in 1996, for pi ng exports of mechanical and electronic products, the ratio 45 per cent in 1995(data from ZWZB, 19 March 1995, p 3: ZJGN 1995, pp 1-66: GJSB 1997, p 1: ZJGN 1996, pp 1-50). Hence, it is premature to judge that Chinas exports of and electronic products are labour-intensive, at least in comparison with the rest of Chinese exports. 01998 John Wiley Sons, Ltd J.mnt.Dev.10,733-749(1998)

A further observation that can be made from Table 1 and Table 2 concerns the exports±imports ratio. It can be seen that, between 1988 and 1996, for the indicated trade categories the general trend is the larger the increase in the ratio the faster the expansion of exports. Whilst the exports-imports ratio of all manufactures increased by 41 percentage points, that of mechanical and electronic products, electronic products alone and computer products increased by 49, 80 and 173 percentage points, respectively. To some extent, this development is related to the growth of foreignfunded processing or assembling operations in China. However, on a priori grounds, one cannot rule out the validity of inferring from it that the Chinese economy is heading towards a pattern of intra-industry, rather than inter-industry, international specialization. This presents yet another anomaly for the standard theory of comparative advantage.3 Sceptics of course can go on to argue that the intra-industry specialization might still be divided in a way that China is engaged mainly in the labour-intensive operations of each industry, even if the industry concerned is on average characterized by high capital and technology intensity.4 Constrained by the availability of data, it is dicult to investigate into the trade and production characteristics of the intraindustry division of labour. Nevertheless, focusing on the output composition of China's electronics industry, Figure 1 presents a trend of increasing vertical integration which runs counter to such argument. Between 1988 and 1995, the output share of industrial equipment and components Ð both characterized by higher capital and technology intensity Ð increases by 14 and 5 percentage points, respectively. As a result, the output composition in 1995 contrasts sharply with that of the mid-1980s when the Chinese electronics industry was overwhelmingly skew to the production of downstream consumer products. One possible way of explicitly examining the factor intensity of the trade categories is to check the respective capital-labour ratios on the production side. Lardy (1992) 3 In this connection, it is of note that the expansion of electronics exports, and exports of mechanical products in general, has been a salient feature of East Asian industrialization, both in the four `earliermovers' of Hong Kong, South Korea, Singapore and Taiwan, and the second-tier NICs (Malaysia, Thailand, Indonesia and the like). China's expanding exports of these products might therefore seem to be just in line with the `normal' pattern of international development. Nevertheless, it is a matter of disagreement among scholars as to whether the East Asian experience is indeed normal in the world scale Ð particularly in the case of electronics production, where outside this handful number of countries the vast majority of the developing world has been excluded from engaging in the production. And the institutional environment (especially the role of state policy vis-aÁ-vis the world market) within which electronics production has developed in East Asia is even a matter of disagreement (Ernst and O'Connor 1992; Hanna et al. 1996). 4 Citing a World Bank study, Lardy (1994, p. 32), for instance, infers that `(China's) exports of telecommunications equipment and domestic electrical equipment, which might appear to be more capitalintensive, are based overwhelmingly on processing and assembly-type activity. Thus they too are laborintensive since imported components account for about four-®fths of the value of exports'. It is true that the ratio of processing exports is much higher in mechanical and electronic products than in total merchandise exports: 71 per cent versus 47 per cent in 1994. But the ratio of domestic value-added in processing exports, measured as the ratio of net-to-gross processing exports, is also much higher in the former than in the latter: 43 per cent versus 17 per cent in the same year. Moreover, both of the two have had substantial increases in recent years: for total processing exports, the ratio increased to 21 per cent in 1995 and 26 per cent in 1996, for processing exports of mechanical and electronic products, the ratio increased to 45 per cent in 1995 (data from ZWZB, 19 March 1995, p. 3; ZJGN 1995, pp. 1±66; GJSB, 11 January 1997, p. 1; ZJGN 1996, pp. 1±50). Hence, it is premature to judge that China's exports of mechanical and electronic products are labour-intensive, at least in comparison with the rest of Chinese exports. #1998 John Wiley & Sons, Ltd. J. Int. Dev. 10, 733±749 (1998) Machinery and China's Trade±Growth Nexus 737