World Bank Reprint Series: Number 114 Bela Balassa The Changing International Division of Labor in Manufactured Goods Reprinted with permission from Banca Nazionale del Lavoro Quarterly Review, no. 130 (September 1979), pp. 243-85 The Changing International Division of Labor in Manufactured Goods* Introduction This paper will examine the changing pattern of the inter- national division of labor in manufactured goods between deve- loped and developing countries. It will consider recent and pros- pective changes in trade flows between the two groups of countries and analyse the employment implications of these flows for the developed countries. Section 1 of the paper will describe changes in trade in man- ufactures between developed and developing countries following the oil crisis of 1973. The employment effects of this trade for the developed countries will be analysed in Section II, with further consíderation given to the skill-composition of employment. In- formation will also be provided on the physical and the human capital intensity of the manufactured exports and imports of the developed countries in trade with the developing countries. Section 111 will consider possible future changes in manufactur- ed trade between the developed and the developing countrius ancl indicate the employment implications of the projected trade flows. In so doing, distinction will again be made among skill classes. Finally, the policy implications of the findings will be discussed. The author is Professor of Political Economy at the Johns Hopkins Uni- versity and Consultant at the World Bank. He prepared this paper as a back- ground study for World Developrent Report, 1979. However, thc paper expresses the author's views alone and should not be interpreted to reflect those of che Bank. The author is indebted to Robert E. Baldwin for useful comments on an carlier draft and, to participants of a Worid Bank seminar for helpful discussions. Thanks are also due to Joung-Yong Lee, Michel Noel, and Kishore Nadkarni for research assistance. 244 Banca Nazionalc del Lavoro 1. Recent Trends in I'rade in Manufactured Goods belween the Developed Industrial and the Developing Countries' The Time Pattern of Trade In popular discussions, as well as in intertnational fora, much attention has been given in recent years to the growth of the manufactured imports of the developed countries fromn the deve- loping countries. In particular, questions have been raised con- cerning the effects of increased imports on output and emoloyment in the forimer group of countries. Changes in the imports of the developed countries from the developing countries cannot, however, be considered independently frorn changes in their exports to them. Correspondingly, in the following, we will examine the time pattern of manufactured exports and imports in trade between developed and developing countries. This will be done for the period 1973-77, fturther disaggregating the results according to major country and industry groups. Unless otherwise noted, all data have been expressed i'n current U.S. dollars. Use has been made of data reported by the developed in- dustrial countries (for short, the industrial counturies), including the United States, Canada, the European Common Market, the Eu- ropean Free Trade Association, and Japan. In turn, among the developing countries, distinction has been made between oil-export- ing and oil-importing countries. Australia, New Zcaland, and South Africa, the countries of Southern Europe, anid the Communist coLuntries have not been included in either grouping. Tables 1 anld 2 provide annual data for the period 1973-77 on trade between industrial and developing countries in manufactured goods (defined as commodity classes 5 to 8 in the U.N. Standard Internatial 7rade Classification less 68 nonferrous metals) in a geo- graphical and in an industry breakdown, respectively. Within the manufa¿turing category, -- .ition has been made among the following commodity groups: nonelectrical and electrical machinery and transport equípment; chemicals, iron and' stecl; textiles; clo- thing; and consumer goods other than clothing and their inter- mediate inpLuts. The latter caregory inclucle, leather and leather Statistical work for thís section f the paper was done by Michel Noel. The Changing International Division of Lalbor in Manufactured Goods 245 products, rubber products, wood and cork products, paper and paper products, non-metallic mineral manufactures, sanitary plumbing, heating, and lighting equipment, furniture, shoes, and miscellaneous manufactures. Little needs to be said of the industríal countries' trade in manufactured goods with the oil-exporting developing countries. The latter group of countries export few manufactured goods while their imports have been rising rapidly since the quadrupling of oil prices in late 1973. As a resilt, the industrial countries increased their exports of manufactured goods to the oil-exporting developíng countr.es from $12.9 billíon in 1973 to $55.5 billion in 1977. Several phases may be distinguished in regard to trade in manufactures between the industrial countries and thc oil-impor t- ing developing countries. In the years of the world recession, 1974 and 1975, the industrial countries' imports of manufactured goods from the developing countries increased by 27 percent. Adiusáing for price ncreases of 31 percent,2 the volume of these imporí.s declined by 3 percent between 1973 and 1975. In the same period, the mnanufactured exports of the industrial countries to the oil- imnporting developing countries rose by 59 percent, represeriting price incrcases of 37 percent and a volume increase of 16 perccnt. These changes in trade flows led to an increase in the export surplus of the industrial countries in their manufactured trade with the oil-importing developing countries from $23.9 billion in 1973 to $42.9 billion in 1975, The United States accounted for two- thirds of the rise in the surplus, with an export increase of 71.5 percent as against average increases of 54.2 percent for the other industrial countries, reflecting largely an improvenenct of the U.S. competitive position following the devaluation of the dollar. The described developments were associated with the conti- nued growth of the national economies of oil-irmporting developing countries in the midst of the world recession. The combined GDP of these countries increased by 9.7 percent between 1973 and 1975 while the 1975 GDP of the industrial countries barely reached the 1973 level. 2 The unlit value indices used in deflating the currcnt dollar value of the malnufactured CxportS of thc developing countrics have becen derived from dalt published in GATT, ¡Vetworks of W'or/d Trade by Areas and Commodiív CGlasses, 1955 1976, CG;I1C\na. 1978. Thc corresponding data for the developed counicries are published in United Nations, Mrowbi'y BuíaieN of Staiistics, December 1978. 246 Baíica Nazionale del Lavoro TRADE BETWEEN THE INDUSTRIAL COUNTRIES AND THE DEVELOPING (Billion dol Un,ted SmatCs Canada EEC Expont lmport |Balance Er rt Import Balance Export Imponr Balantce Oil-Exporting Developing Countries 1973 .... ....... 2.43 0.06 2.37 0.17 0 0.17 7.00 0.42 6.58 1974. 8 ............4.28 0.10 4.18 0.31 0 0.31 11.63 0.42 11.21 1975. 0 ............8.00 0.09 7.91 0.57 0 0.57 20.00 0.46 19.54 1976. 0 ............9.90 0.09 9.81 0.76 0.01 0.75 23.10 0.53 22.57 1977 ............. 10.50 0.10 10,40 0.84 0 0.84 29.43 0.67 28.,6 Oil-Importing Developing Countrifls 1973 . 5 ............10.35 7.33 3.02 0.57 0.53 0.04 16.01 4.68 11.33 1974 ............ 16.10 9.73 6.37 0.87 0.76 0.11 23 03 5.83 17.20 1975 ............ 17.75 8.89 8.86 1.06 0.76 0.30 25.82 6.85 18.97 1976 ............ 18.25 12.95 5.30 I.06 1.21 -0. 1 5 25.35 8.92 16.43 1977 .......... 18.84 15.61 3.23 1.08 0.16 -0.08 29.98 10.87 19.11 Developing Countríes ¡ogether 1973 ............ 12.78 7.39 5.39 0.74 0.53 0.21 23.01 5.10 17.91 1974 ...... 20.38 9.83 10.55 1.18 0.76 0.42 34.66 6.25 28.41 1975 ............. 25.75 8.98 16.77 1.63 0.76 0.87 45.82 7.31 38.51 1976 ............. 28.15 13.041 15.11 1,82 1.22 0.60 41.45 9.45 39.00 1977 ....... .... 29.34 15 S1 13.63 1.92 1. 16 0.76 59.41 11.54 47.87 Source, GA1T, lIoernautona/ Trade 19 '7/ 8 Thae oil-importing developing countries financed the expansion of tnheir manufactured imports by drawing down foreign exchange reserves and borrowing in international financial markets. In a recessionary situation, the increased imports of these countries be- nefited the industrial countries through a multiplicative effect on their domestic incomes. The situation changed in 1976 when rernewcd economic expan- sion on the industrial countries gave impecus to thcir impotts of manufactured goods from the oil-importing developing countries. At the same time, the 19 percent increase in the volume of these The Chanigng International Division of Labor in Manufactured Goods 247 TABLE 1 COUNTRIES IN MANUFACTURED GOODS (EXCLUDING NON-FERROUS METALS) lars, f.o.b.) FTA Japan Indusurial Countrics F.xpot Inxport B21.nce Export 1mpon B.aInce Exporn Import BaLocC 0.69 0.05 0.64 2.57 0.02 2.55 12.86 0.55 12.31 1.19 0.07 1.12 5,16 0.03 5.13 22.59 0.62 21.95 1.97 0.05 1.92 8.12 0.01 8.11 38.66 0.61 38.05 2.80 0.06 2.74 8.93 0.02 8.91 45.49 0.71 44.78 3.21 0.08 3.13 11.51 0.04 11.47 55.49 0.89 54.60 2.56 0.75 1.81 9.96 2.22 7.74 39.45 15.51 23.94 3.54 1.01 2.53 14.90 2.52 12.38 58.44 19.85 38.59 3.74 1.18 2.56 14.25 2.00 12.25 62.62 19.65 42.54 3.82 1.38 2.44 16.05 2.89 13.16 64.53 27.35 37.18 ,4.57 1.81 2.76 20.14 2.97 17.17 74.61 32,42 42.19 3.25 0.80 2.45 12.53 2.24 10.29 52.31 16.06 36.25 4.73 1.08 3.65 20.06 2.55 17.51 81.03 20.47 60.54 5.71 1.23 4.48 22.37 2.01 20.36 101.28 20.29 60.99 6.62 1.44 5.18 24.98 2.91 22.07 110.02 28.06 81.96 7.78 1.89 5.89 31 65 3.01 28.64 130.10 33.31 96.79 imports, derived by adjusting the 39 percent increase in import value for an average price increase of 8 percent, exceeded the 5.3 percent rise in the combined GDP of the industrial countries by a considerable margin. This result reflects in part the high marginal propensity to import in the developed countries and in part the export prormotion efforts of several developing countries, in par- ticular, Korea and Taiwan. In turn, the value of the manufactrured exports of tic in- dustrial courntries to the oil-importing developing countries rose by only 3 percent, with export prices remaining at the 1975 level. 248 Banca Nazionale del Lavoro THE INDUSTRY COMPOSITION OF TRADE BETWEEN THE INDUSTRIAL COUNTRIES AND THE DEVELO (Billion dol Non-electracal aud Electric al Mlachanleey Irm, atad Stecl and T~r.nsport b,quiprnnaCemcasTeti Exp.rt Impor Bala1c Ekxponr | Itmpon Balane r",p,c,t I -Pr 13alan1ce Oil- Fxporting Deve/oping Countr¡es 1973 ............. ...8.04 0.07 7.97 2.89 0,06 2.83 0.78 0.31 0.47 1974 ............. ..13.25 0.08 13.17 6.15 0.13 6f02 1.11 0.28 0.83 1975 .......... .... 25.90 0.14 25.76 8.30 0.10 8.20 1.35 0.28 1.07 1976 ............. .32.88 0.15 32.73 6.93 0.10 6.83 1.56 0.32 1.24 1977 ............. 40.09 0.22 39.87 7.83 0.11 7.,72 1 84 0.34 1.50 Oil lmporting Devte/oping Countr¡es 1973 .......... .... 22.71 3.70 19.01 10.02 1.47 8.55 2.38 2.02 0.36 !974 ............. 31.85 5.30 26 55 17.75 2.35 15.40 2.88 2.40 0.48 1975 ......... .... .37.55 5.31 32.24 16.24 2,07 14.17 2 55 1.96 0.59 1976 ............. .40.04 a .45 32. 59 14.91 2.40 12.51 2 75 2.72 0.03 1977 ..... ....... 46.01 9.23 36.78 16.88 2.h1 14 07 3 25 2.93 0.32 Da :a/'ns, Countrzes together 1973 ............. ..30.75 3.77 26.98 12.91 1.53 11.38 3.16 2.33 0.83 1974 ............. .45.10 5.38 39.72 23.90 2.48 21.42 3.99 2.68 1.31 1975 ............. .63.45 5.45 58.00 24.54 2.17 22.37 3.90 2.24 1.66 1976 ............. .72.92 7.60 65.32 21.84 2.50 13.34 4.31 3.04 1.27 1977 ............. ..86.10 9.4 5 76.6, 24.71 2.92 21.79 5.09 3.27 1.82 Source; GATr, Internato.nal Trade ¡ 977/'8. Within this total, the exports of iron and steel and chemicals declined by 8 percent while other commodity groups showed in- creases of 7 to 8 percent. These developments may be explained by a slowdown in the economic growth in some of the oil-importing developing countrics and in the efforts made to improve their balance-of-payments position, in particular through import substi- tution in steel. The described changes in trade flows led to a decline in the net export balance of the industrial countries in their manufactured trade with the oil-importing developing countries from $42.9 bilí- The Changing International Division of Labor in Manufactured Goods 249 TABLE 2 9ING COUNTRIES IN MANUFACTURED PRODUCTS (EXCLUDING NON-FERROUS METALS), 1973/77 ars, f.o.b.) Consumer Goods other Mat,ufattured Products Clothing than Clothing and 1xcludínR Intcrmcdiac Inputs Non.ferrous Mc(als Export Import Balance Export Import Bil.ance Expon Impont B.a¡.,,c 0.08 0 0.08 I.06 0.08 0.98 12.86 0.55 12.31 0.13 0 0.13 1.90 0.09 1.81 ,22.59 0.62 21.95 0.23 0 0.23 2.86 0.08 2.78 38.66 0.61 38.05 0.31 0 0.31 3.81 0.1? 3,69 45.49 0.71 44.78 0.45 0 0.4s 5.26 0,18 5.08 55.49 0.89 54.60 0.39 3.42 -3.03 3.97 4.9i -0.94 39.45 15,51 23.94 0.52 4.30 -3.7 5.45 5.51 -0.06 58.44 t9.85 38.59 0.51 4.97 -4.46 1 5.75 5.37 0.38 62.62 19.68 42.94 0.55 7.08 -653 6.22 7.69 -1.47 64.53 27.35 37.18 0.72 7.78 -7.06 7.76 9.70 -1.94 74.61 32.42 42.19 0.47 3.42 -2 95 5.03 4.99 0.04 52.31 16.06 36.25 0.65 4.30 -3.65 7.35 5.60 1.75 81.03 20.47 60.56 0.74 4.97 -4.23 8.61 5.45 3.16 101.28 20.29 80.99 0.86 7.08 -6.22 10.03 7.81 2.22 110.02 28.06 81.96 1.17 7.78 -6.61 13.02 9.88 3.14 130.10 33.31 96.79 ion in 1975 to $37.2 billion in 1976. The brunt of the decline wvas borne by the United States, whose competitive position deterio- rated at the time. The export surpluses of European countries decreased to a lesser extent while Japan experienced a small increase in Its surplus. Predictions made as to the continuation of this trend proved erroneous, however, and the industrial countries' export surplus in manufactures traded with the oil-importing dcveloping countries íncreased again the following year. As a result, the trade balance in manufactured goods between the two groups of countries returned 250 Banca Nazionale del Lavoro to approximately the 1975 level, exceeding the level reached ¡i 1973 by more than thrce-fourths. The improverent in the industrial countries' trade balance in manufactured goods with the oil-importing developing countrics between 1976 and 1977 reflects a deceleration in the growth of their imports and an acceleration in the rise of their exports. The rates of iricrease were 19 percent and 16 percent, respectively, when the average prices of manufactured goods in world trade increased by 9 percent. The slowdown in the imports of the industrial countries was concentrated in Japan, whose manufactured imports from the oil- importing developing countries rose by only 3 percent in 1977, as wefl as it1 Canada, wh(ose imporcs fell in absolute terms. In both cases, a decline in the rate of economic growth and continued protection against developing country products contcributed to the outcome. Despite increased protection of textiles and clorhing and a deceleration of econiomic growth, U.S. and European imports of manufactured goods from the non-oil producing developing coun- tries rose by more than one-fifth becwccn 1976 and 1977. With a continued deterioration ín its competitive posit on, however, U.S. exports of manufactured goods to the oil-imnporting developing countries increased by only 3 percent. As a result, the manLifactured trade surplus of the United States with thesc coun- tries declined further, from $5.3 billion in 1976 to $3.2 billion in 19777. In turc, slow inercases in imports, accompanied by a one- fourth rise in exports, led to an incrcase in Japan's trade surplus in manufactured goods with the oil-imnporting developing countries from $13.2 billion in 19-76 to S.2 billion in 197,7. In the same period, the trade surpluses of the European countries with the oil-imnporting developinig countries increased by about one-sixth. Amonog individual product groups, the slowdown in the in- du.striai councries' iniporís was the most pronounced in textiles and clochinng, with imercases of 8 and 10 percent, respectively, reflecting in part a ¿Osoll¡d,ciorin a fter the unsustainable increase of about 40 perecll in 1976 and in part the effects of protectionist' measures takecn. By contrast, the imports of nonelectrical and clectrical ma- chinery and tranisport equípicnt, and of consumer goods other thanc tcxtileis and their intermediate inputs, incretas(d by ahout onc-fourth and imports of iron and steel and chemicals by one- sixth. The Changing International Division of Labor in Manufactured Goods 251 In the exports of the industrial countries to the non-oil im- porting developing countries an increase by one-fourth occurred in consumer goods other than textiles and their intermediate inputs. The exports of clothing rose at a somewhat faster (31 percent), and textile export3 at a somewhat slower (18 percent), rate. Finally, the exports of nonelectrical and electrical machinery and transport equipment and of iron and steel and chemicals increased by 15 and 13 percent respectively. The Changíng Position of Individual Countries and Country Groups For the 1973-77 period as a whole, the manufactured trade surplus of the industrial countries with the oil-exporting developing countries increased nearly four-and-a-half times, from $12.3 billion in 1973 to $54.6 billion in 1977. There were relatively small va- riations in percentage increases of the export surplus among the industrial couneries, with Canada (394 percenlt) at the upper, and the EEC countries (337 percent) at the lower end of the range. In turn, the manufactured trade surplus of the industrial coun- tries with the oil-importing developing countries rose from $23.9 billion in 1973 to $42.2 billion in 1977, with substantial variations shown among the major industrial countries andc country groups. The U.S. surplus increased from $3.0 billion in 1973 to $3.2 billion in 1975, with practically no change in Canada, increases in the surpluses of the EEC and the EFTA countries from $l1.3 billion to $19.1 billion and from $1.8 billion to $2.8 billion, respectively, and a rise from $7.7 bíllion to $17.2 billion in Japan. The observed changes in net balance of the Common Market countries and the United States in their manufactured trade with the oil-importing developing countries are largely explained by re- ference to the initial situation in 1973. The EEC countries in- creased their imports of manufactured goods from the oil-importing developing countries (132 percent) more rapidly than the United States (113 percent) and experienced only a slíghtly larger rise in their exports to these countries (87 percent as compared to 82 percent in the U.S.). However, they benefited from the fact that their export-import ratio in manufactured trade with the oil-irn- porting developing countries was 3.4 in 1973 as compared to 1.4 in the United States. 252 Banca Nazionale del Lavoro Similar considerations apply in United States-EFTA relation- ships. Thus, the imports of manufactured goods b- the EFTA countries from the oil-importing developing countries rose more (141 percent) than imports into the United States, while their exports increased less (79 percent) than U.S. exports. But, the EFTA countries had an export-import ratio of 3.4 in ma.nufactured trade with the oil-importing developing countries in 1973, just as the Common Market. Differences in the initial situation, in turn, are largely explained by the more liberal import policy followed vis-a-vis the developing countries by the United States, as compared to European countries, in. the early seventies. The contrast is even greater with Japan, whose export-import ratio in manufactured trade with the oil-im- porting developing countries was 4.5 in 1973. At the same time, Japar,'s imports from these countries rose by only 34 percent be- tween 1973 and 1977, represenring an absolute decline in volume terms. The protectionist policies followed by Japan, together with the greater similarity of its economic strLctnure to that of the more advanced developíing countries, may, áccount for these results. At the same time, the 102 percent rise in Japan's exports was aided by its increased competitiveness. As a result, japan's manufactured trade surplus with oil-imporcing developing countries more than doubled between 1973 and 197 7. The Commodity Composition o/ Tr.¡de The manufactured exports of the industrial countries to the oil-exporting developing countries are increasingly dominated by nonelectrical and electrical mnachinery and transport equipment, which accounted for 62.5 percent of the total in 1973 and for 72.3 percent in 1977. Other commodity groups of importance, with their 1977 share in parenntlesis, are iron and steel and chemicals (14.1 percent) and consumer goods other than clothing and their inter- mediate inputs (9.5 percent). Noneletrical and electrical machinery and transport equipment also assurned inereasíng importance in the manufactujr There are a few additional exceptions for individual developed countries and country groups but the differences exceed 10 percent only for transportation equipment in the United States (a ratio of 1.13), leather and leather products in the EEC (a ratio of 1.25), and miscellaneous manufactured products in Japan (a ratio of 1.36). In the United States. the result has been due to the relatively high share of th imports of auromobile parts and accessories that are combined with capital-intensive motor vehicles; in the EEC, the high export share of capital-intensive leather has contributed to the outcome; while in Japan the exportation of relatively labor-intensive toys and sports goods is largely responsible for the result obtained in regard to miscellaneous manufactured products. The numerical magnitudes of average betwveen and w¡thin in- dustry group differences on labor-input coefficients are of further interest. The5e can be shown by comparing the ratios of average labor-input coefficients for exports and imports reported above with the ratios derived on the assumption that labor-input coefficients are invariant within each of the eighteen industry groups. The relevant results are .65 and .71 for the OECD, taken as a whole, .61 and .66 for the United States, .65 and .70 for the EEC, and .73 and .81 for Japan. It appears, then, that within-industry group differences in labor-input coefficients account for about one-sixth of '> Labor-input cocfficierits are the same for exports and for rrnpcts ¡ti the case of furniture and pctrolcurn produccs where, due to data limitations, further disaggregation was not feasible The Changing International Division of Labor in Manufactured Goods 265 the difference in average labor-input coefficients for exports and imports, estimated in a 184 commodity category breakdown. It may be suggested that additional disaggregation would further increase differences in the coefficients. Thus far, we have compared average labor-input coefficients for manufactured goods traded between developed and developing countries. These coefficients will be relevant in the event of a balanced expansion of manufactured trade between the two groups of countries. In such an eventuality, the number of jobs lost through increased imports would be about one-half higher than the number of jobs gained through exporting, for the OECD taken as a whole. An alternative hypothesis postulates equiproportionate increa- ses in the exports and imports of manufactured goods in trade between developed and developing countries (i.e. identical rates of change for exports and imports); for short, a proportional expansion of trade. Assuming further unchanged export and import structures as beforehand, the employment effects of a proportional expansion of trade are indicated in Table 4 that provides information on the labor content of manufactured exports, and of products competing with manufactured imports, in the trade of the developed countries with the developing countries. The estimates have been derived by multiplying export and import values for 1976 by the labor-input coefficients estimated in the 184 industry breakdown. The results show substantial positive employment effects for the OECD countries in their trade in manufactured goods with the developing countries. Similar conclusions apply to the United Sta- tes, the EEC, and Japan. Thus, in the event of a proportional expansion of trade, the ratio of jobs gained through exports to jobs lost through imports would be 2.8 in the OECD taken as whole, 1.4 in the United States, 3.4 in the EEC, and 7.9 in Japan. It is apparent that differences in the results shown in Tables 3 and 4 are explained by intercountry variations in export-import ratios in man- ufactured trade with the developing countries. 1 The proportional expansion of manufactured trade between developed and developing countries would create new jobs in most industry groups in the developed countries. Exceptions are textiles, clothing, lumber and wood products, rubber and plastic products, leather and leather products, and miscellaneous manufactures. For the OECD countries, taken together, a 10 percent proportionial 266 Banca Nazionale del Lavoro TABLE 4 TOTAL EMPLOYMENT EFFECTS OF MANUFACTURED TRADE BETWEEN DEVELOPED AND DEVELOPING COUNTRIES, 1976 (Thousand iobs) OECD Expon I..pon Die 1. Textile Mill Products ............... . ..138.9 192.2 - 53,3 2. Apparel & Other Textile Products ..-. 40.8 218.2 -177,.4 3. Lumber & Wood Produccs 0.......... 1.4 26.3 - 15.9 4. Furniture & Fixtures ................ 19.7 10.6 9.1 5. Paper & Allied Products ... - ..... 33 3 2.8 30.5 6. Printing & Publishing ............... 25.0 2.7 22 3 7. Chcmical & Allied Products 163.4 23.4 140.0 8. Petroleum Products ................. ..... 1.1 0.0 1 1 9. Rubber & Plastic ProducCs . . . 32.2 63.1 30.9 10. Leather & Leather Products . ...... 10.1 36.2 - 26 2 II, Stone, Clay & Glass Products ........ 51.6 10.9 40? 12. Primary Metal & Allied ............ 194.2 27.7 166.5 13. Fabrica ted Metal Products .......... 91.9 7 3 84.5 14. Nonelectrical Machinery .......... .. .. 527.4 18. 508.7 15. Electrical Equipment & Supplies 417.3 122.7 294.6 16. Transportation Equipment ..... ... 486.2 13.7 472. 5 17. Instrumenrs & Related Products . . .72 17.8 54.9 18. Misc. Maniufactured Produces 'Y'. ,7 . 58.0 - 103 Total .................... 2 563 R8 852.5 1511 3 of whlich stim of positive balances 1825 4 negative balances . . . 314.1 USA 1. Textile Mill Products .............. 15.4 62.5 -47.1 2. Apparel & Other Textile Proclucts . . 13. 0 81.5 -68.5 3. Lumber & Wood Products 2.0 11.5 - 9.5 4. Furniture & Fixtures .............3.4 4.1 - 0.' 5. Paper & Altied Products ....... . . 8.6 1.0 7.6 6. Printing & Publishing .............. . 1.5 0.8 3.7 7. Chemical & Allied Products .375 10.4 27.1 8. Pectroleurn Products .......... - . 0.0 0.0 0.0 9. Rubbcr & Plastic Products 4.4 41.8 -37.4 10. Leather & Leather Products 2..... . 25 15.1 -12.6 I1. Stone, Clay & Glass Products 6....... 6. 4.5 2.1 12. Primarv Metal & Allied ............ 25.' 11.9 13.8 13. Fabricated Metal Products ... ........ 15.9 2,8 13.1 14. Nonlecetrical MRachinery ............ 142.7 6.4 136.3 15 Electrical Eqtuipment & Supplies .. ,. 106.3 77.9 28.4 16. Transportation Equipment .......... 108.0 2.6 105.4 17. Instruments & Related Products 16.3 9.5 6.8 18. Misc. Manufactured Products 9.1 27.2 -18.1 Total. 521.9 371.5 150.4 of which surm of positivc balanccs 344.3 negacívc balances . . 193.9 The Changing International Division of Labor in Manufactured Goods 267 L'ABLE 4 (cont.d) TOTAL EMPLOYMENT EFFECTS OF MANUFACTURED TRADE BETWEEN DEVELOPED AND DEVELOPING COUNTRIES. 1976 (Thousand jobs) EEC Export Imporr B.I.,ue 1. Textile Mill Products ............... 48.7 74.3 -25.6 2. Apparel & Other Textile Products . .. 19.8 90.1 -70.3 3. Lumber & Wood Products .......... 5.2 8.9 - 3.7 4. Furniture & Fixtures ............... 12.8 3.4 9.4 5. Paper & Allied Products 6........... 6.5 1,2 5.3 6 6. Printing & Publishing .................... 15.3 0.8 14.5 7. Chemical & Allied Prcducts ......... 73.8 6.6 67.2 - 8. Petroleum Products .... ......... ..... 0.7 0.0 0.7 9. Rubber & Plastic Products .......... 16.0 11.5 4.5 10. Leather & Leather Products ......... 3.7 13.7 -10.0 11. Stone, Clay & Glass Products ........ 25.7 3.2 22.5 12. Primary Metal & Allied .7.4.......... ---A 9.8 67.6 13. Fabricated Metal Products .......... 49.5 2.2 47.3 14. Nonelectrical Machinery ............ 261.7 6.6 255.1 15. Electrical Equipment & Supplies .... 187.1 29.9 157.2 16. Transportation Equipment .......... 145.1 5.0 140.1 17. Instruments & Relatcd Products ..... 22.3 4.5 17.8 18. Misc. Manufactured Products ........ 25.3 18.3 7.0 Total .............................. 996.7 289.9 706.8 of which sum of positive balances 816.3 negative balances .... 109.5 JAPAN 1. Textile Mill Products ............... 61.7 18.8 42.9 2. Apparel & Other Textile Products .... 3.4 13.7 -10.3 3. Lumber & Wood Products .......... 0.3 2.2 - 1.9 4. Furniture & Fixtures ............... 1.7 1.6 0.1 5. Paper & Allied Products ............ 3.6 0.1 3.5 6. Printing & Publishing ...... ............... 1.4 0.2 1.2 7. Chemical & Allied Products ......... 28.8 2.8 26.0 8. Petroleum Products ................ 0.0 0.0 0.0 9. Rubber & Plastic Products ........... S1 3,2 4,9 10. Leather & Leather Products, ........ 3.1 2.0 1.1 11. Stone, Clay & Glass Products 1......... l0 1.3 9.7 12. Primary Metal & Allied ............ 73.1 2.8 70.3 13. Fabricated Metal Products ........ 13.5 0.3 13.2 14. Nonelectrical Machinery ............ .... 71.5 1,6 69.9 15. Electrical Equipment & Supplies ..... 88.6 10.8 77.8 16. Transportation Equipnment ..........r 177.4 3.0 174.4 17. lnstruments & Related Products ..... 17.2 1.5 15.7 18. Misc. Manufactured Products . . . ..... 9.4 6.3 3.1 Total ......................... .. . 573.8 72.3 501.6 of which sum of positive balances 5 13.8 negative balances .. . 12.2 Source: Set text 268 Banca Nazionale del Lavoro increase in trade flows would entaiíl the loss of altogether 31 thousand jobs in these industries as compared to a total gain of 183 thousand jobs in the other industry groups. Comparable figures are 19 thousand and 34 thousand for the United States and 11 thou- sand and 82 thousand for the EEC. Finally, Japan would experience a gain of 51 thousand jobs in its export industries, with practically no loss of jobs elsewhere.'°1 The Sk/ll-Intensity of Trade We have seen that average labor inpu.-coefficients are about one-half greater for the manufactured imports of the developed countries from the developing countries than for their exports to these countries. The opposite conclusion obtains if we consider professional and technical labor alone. In this occupational group, the ratio of labor-input coefficients for exports as compared to imports is 1.20 in the OECD taken as a whole, 1.20 in the Unired States, 1.29 in the EEC, and 1.12 in Japan (Table 5A). Thus, there would be a gain in the employment of professional and technical labor in the developed countries, even if one assumed a balanced expansion of their trade in manufactured goods with the developing countries. On the OECD level, as well as for the EEC and Japan, there would also be gains in the employment of foremen and skilled workers, while a small loss in this category is shown for the United States. In turn, the largest losses would be incurred in regard to semi-skilled and unskilled production workers, where the ratio of labor-input coeffipients for exports as compared to imports is .44 for the OECD taken as a whole, .42 for the United States, .45 for the EEC, and .50 for Japan." The resulrs cited in Table 5A point to the existence of con- siderable differences in the occupational structure of production for exports and for import substitution in the trade of the developed countries with the developing countries. These differences are put rito focus in Table 5B that provides information on the average lO In al] cases, we have taken one-tenth of the sum of positive and negative employment balances shown in Table 4. It is apparent from the table that the net employment effects for particular industries were generally of the same sign in the individual developed countries and country groups as for the OECD as a whole. " Results for the individual occupational groups have been obtained by utilizing U.S. statistics on Occupation by Industry for the year 1970. The Changing International Division of Labor in Manufactured Goods 269 TABLE 5 EMPLOYMENT EFFECTS OF TRADE IN MANUFACTURED PRODUCTS BETWEEN DEVELOPED AND DEVELOPING COUNTRIES ACCORDING TO OCCUPATIONAL CATEGORIES OECD USA EEC JAPAN Exporw Impon Exporr Import Expon Import Expon,v Imponr A. Labor-Input Coe/ficients (No.ojJobs per $ Ml/ion of Output) 1. Professional and Technical Workers ..................... ...07 1.72 2.36 1.97 2.03 1.57 1.93 1.'2 2. Managersand Administrators 1.03 1.40 1.06 1.47 1,08 1.36 0.91 1.32 3. Sales Workers . ............... 0.28 0.50 0.28 0.51 0.37 0.51 0.22 0.47 4. Clerical Workers ...............29 2.93 2.35 3.08 2.32 2.85 2.23 2.72 5. Foremen and Skilled Workers (Craftsmen) 4.28 3.84 3.69 3.89 4.08 3.69 5.26 3.94 6. Production Workers, Unskilled and Semi-skilled (Operators) 7.32 16.57 7.08 16.68 7.53 16.81 7.36 14.63 7. Workers in Construction, Tran. sportation, Material Handling, etc. 0.39 0.50 0.36 0.50 0.38 0.50 0.42 0.48 8. Laborers .................. 0.78 1.06 0.60 1.02 0.76 1.11 0.99 1.04 Total ............18.44 28.53 17.77 29.12 18.50 28.40 19.30 26.32 B. Employment Shares (percent) 1. Professional and Technical Workers ................... 11.22 6.04 13,26 6 77 10.98 5.52 9.99 6.53 2. Managersand Adminisrrators 5.57 4.91 5.96 5.04 5.83 4.80 4.73 5 01 3. Sales Workers . . ...............54 1.76 1.56 1.75 1.72 1,78 1.14 1.78 4. Clerical Workers ............12.43 10.27 13.20 10.59 12.55 10.03 115.3 10.32 5. Foremen and Skilled Workers (Crafrsmen) .................23.23 13.45 20.76 13.37 22.07 12.99 27.25 14.96 6. Production Workers, Unskilled and Semi-skilled (Operators) 39.71 58.08 39.84 57.27 40.58 59.19 38.11 55.60 7. Workers in Construction, Tran- sportation, Material Handling, etc. 2.09 1.76 2.04 1.7 3 2.07 1.77 2.15 1 84 8. Laborers ...................4.23 3.73 3.39 3.49 4.11 3.9)2 5.14 3 95 Total ........... 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 C. Employmen: Ejfects of Trade ín 1976 Thousands of/Jobs) 1. Professional and Technical Workers ... ............... 265.2 51.5 69.2 25.2 109.4 16.0 57 3 4.7 2. Managersand Administrators . 131.8 41.9 31.1 18.7 58.2 13.9 27.1 3.6 3. Saics Workers . ...............36.4 15.0 8.1 6.5 17.2 5.2 6 5 1.3 4. Clerical Workers......... .. 294.0 87.6 68.9 39.3 125 1 29.1 66.2 - 5 5. Foremen and Skilled Workers (Craftsmen) .................549.0 114.7 108.4 49.7 220.0 37, .7 156.3 10.8 6. Production Workers, Unskilled and Semi-skilled (Operators) 938.6 495.1 207.9 212.7 405.5 171.6 2 18. 4 0 2 7. Workers in Construction, Tran. sportation, Material Handling, etc. 494A 15.0 10.6 6.4 20 6 5.1 12.4 1 1 8. Laborers .................. 99.8 31.8 17.6 13.0 40.9 11.4 29.5 2 9 Total ...........2363.8 852.5 521.9 371.5 996.7 289.9 573.8 72.2 270 Banca Nazionale del Lavoro employment shares of the various occupational categories in the manufactured exports and imports of the OECD countries, taken together, the United States, the EEC, and Japan. Among occupational categories, the share of professional and technical workers in total einployment is nearly double for the exports than for the imports of manufactured goods in the OECD countries' trade with the developing countries. The relevant ratios are 1.96 for the United States, 1.99 for the EEC, and 1.53 for Japan, reflecting a less sophisticated export structure in the latter case. Also, the share of foremen and skilled workers in total em- ployment is about three-fourths higher for the exports than for the imports of the OECD countries in their manufactured trade with the developing countries, with Japan at the top (1.82) and the United States (1.55) at the bottom of the range. The differences go in the same direction,12 but they are numerically smaller, for man- agers and administrators, clerical workers, service workers, and for workers in construction, transportation and material handling. Differences in the opposite direction are observed in regard to sales workers and, in particular, for the largest category, unskilled and semi-skilled production workers. In the latter case, the average employment share of developed country manufactured imports fr- om the developing countries is about two-fifths lower than that of their exports, with the United States (36 percent) and Japan (45 percent) being at the opposite ends of the range. Again, these results conform to the Heckscher-Ohlin theory of international specialization. By comparison, Keesing found that in 1957 the exports of the United States, Germany, Sweden, and the United Kingdom were relatively skill-intensive whereas skill intensity was higher for im- ports than that for exports in the Netherlands, Belgium, Italy, France, and Japan (1965, p. 291). In interpretating Keesing's es- timates, it should be recalled, however, that they pertain to a relatively early postwar year and concern the total manufactured trade of the countries in question that was dominated by trade among the developed countries themselves. In turn, Baldwin and Lewis estimate that a 50 percent multi- lateral tariff cut in manufacturing (textiles excluded) would provide 12 Exceptions are clerical workers in Japan and workers in construction, tranisportation. and material handling in the United States. The Changing International Division of Labor in Manufactured Goods 271 a net employment gain in the professional and technical labor as well as in the management and administration categories, as against losses in all other labor categories, in the United States (1978, p. 253). A comparison of the Baldwin-Lewis estimates with those reported in Table 5C shows that, as expected, increased trade with the developing countries would lead to a larger shift from low-skill to high-skill occupations than that resulting from an increase in U.S. overall trade. While the estimates reported in Table 5A indicate the em- ployment effects of a balanced expansion of manufactured trade between developed and developing countries in the various occu- pational categories, the estimates of Table SC show the effects of proportional change in this trade. kt is apparent that an equi- proportionate increase in OECD exports and imports in trade in manufactured goods with the developing countries would have a net employment-creating effect in all eight occupational categories. The same results have been obtained for the EEC and Japan and, with the exception of unskilled and semi-skilled production wor- kers, for the United States. For the latter category, the U.S. shows a slight deficit. The Human and the Physical Capital Intensity of Trade We have considered the effects of trade in manufactured goods between developed and developing countries on employment in the former group of countries for various skill classes. In order to express the skill intensity of trade in a singlc number, Kccsing (1965, 1968, 1971)'3 has calculated ratios of jobs in high-skilled ro low-skill occupations. Given the arbitrariness involved in making such a distinction, we have instead followcd Kenen (1965) in es- timating the human-capital intensity of trade. This procedure also permits making comparisons between the physical and the human capital intensity of the exports and imports of the dcveloped countries. 15 In bis carlier work (1965 and 1968), Kccsiig also usedi l.S. labor onp coefficients for examining the occupational effecis of thc exports of various in. dustrial countries. In a later paper (1971), Kcesing has made use of data obtamned for the individual industrial countries. This has tlot becoi done ¡i the present papcr for lack of comparable data on the appropriate brcakdown, 272 Banca Nazionale del Lavoro Physical capital and human capital coefficients have been taken from "A 'Stages Approach' to Comparative Advantage,- (Balassa, 1979) where the derivation of these coefficients is described in detail. Two sets of coefficients have been calculated: "stock" and "flow" coefficients. The former have been defined as the value of fixed capital per worker (physical capital) and the discounted value of the difference between average wage and the unskilled wage (human capital), and the latter as nonwage value added per worker (physical capital) and the average wage (human capital). Table 6 provides information on average physical and human capital coefficients for exports and imports of manufactured goods in the trade of developed countries with the developing countries. As is apparent from the table, for the OECD taken as a whole, capital coefficients for exports are substantially higher than for irnports. The average ratios of the coefficients for exports and for imports are 1.58 for physical capital and 1.44 for human capital utilizing the stock measure, and 1.48 and 1.28, respectively, utilí- zing the flow measure. Similar results have been obtained for the United States and the EEC, except that in the latter case tie stock measure shows the same ratios for physical and for human capital. Note further that ini the EEC the ratio for human capital under the flow measure (1.28) is roughly the same as that obtained for Belgium, where this was estimated at 1.26 for 1970 (de Grauwe, et al., 1977, p. 8). The ratios of physical and human capital coefficient for exports and for imports are, however, substantially lower for Japan than for the other developed countries. This result rnay be taken to reflect the fact that capital accumulation in Japan has not been com- mensurate with incomes per head that have riseri very rapidly in recent years. The cited estimates indicate the relatively high physical capital and skill-intensity of the n;anufactured exports of the developed countries to the developing countries as compared to their imports from these countries. It is thus apparent that, in their manufactur- ed trade with the developing countries, the developed countries exchange physical and human capital for unskilled labor. The Changing International Division of Labor in Manufactured Goods 273 TABLE 6 PHYSICAL AND HUMAN CAPITAL COEFFICIENTS FOR MANUFACTURED EXPORTS AND IMPORTS OF DEVELOPED COUNTRIES (Thousand dollars per worker) OECD Export Import Ratio Stock Measures Physical Capital ...................... 16.6 10.5 1.58 Human Capital ....................... 29.7 20.6 1.44 Together ............................... 46.3 31.1 1.49 Flow Measures Physical Capital ............................ 8.6 5.8 1.48 Human Capital ....................... 9.6 7.5 1.28 Together ........................... .18.2 13.4 1.36 USA Stock Measures Physical Capital ............................ 15.6 9.2 1.68 Human Capital ....................... 30.6 19.9 1.54 Together ............................... 46.2 29.2 1.58 F/ow Meas.ures Physical Capital ............................ 8.8 5.5 1.60 Human Capital ....................... 9.8 7.6 1.29 Together ............................... 18.7 13.1 1.42 EEC Stock Measures Physical Capital ............................ 15.3 11.0 1.39 Human Capital ....................... 29.1 21.0 1.39 Together ................. ........... 44.4 32.0 1.39 F/ow Measures Physical Capital ........................... 9.0 5.9 1.53 Human Capital ............................. 9.5 7.4 1.28 Togethcr ............................ 18.5 13.3 1.39 JAPAN Stock Measures Physical Capital ....................... 14.1 11.9 1.18 Human Capital ....................... 24.8 22.0 1.13 Together ............................ 38.9 33.9 1.15 ,Fow Measures Physical Capital ....................... 8.1 6.7 1.21 Human Capital ....................... 8.4 7.6 1.10 Together ............................ 16.5 14.3 1.15 Source See tcxt. 274 Banca Nazionale del Lavoro III. Employment Implications of Prospective Trends in Trade in Manufactured Goods between Developed and Developing Countries14 Projections for Growth and Trade In the preceding section, we have examined the employment implications for the developed countries of a balanced and a pro- portional expansion of their trade in manufactured goods with the developing countries. Consíderation has further been given to the occupational structure and to the physical and human capital-in- tensity of this trade. Neither a balanced nor a proportional expansion of trade in manufactured goods between developed and developing countries is a likely occurrence. Rather, these alternatives represent extreme cases, assuming a zero trade balance in manufactured goods on the margin in the first case, and an equiproportionate expansion of all manufactured trade flows in the second. Correspondingly, in order to evaluate the employment impli- cations of future trade in manufactured goods between the de- veloped and the developing countries, projections of trade flows would have to be rnade. In projecting trade flows, we have taken the estimates of the Worid Development Report as a point of departure. The Report forecasts the followíng growth rates of the total exports of manufactured goods from the developing countries be- tween 1975 and 1985: textiles, 6.2 percent; clothing, 8.3 percent; chemicals, 13.0 percent; iron and steel, 14.5 percent; machinery and transport equipment, 17.3 percent; and othcr manufactures 10.0 percent; averaging 12.2 percent (1978, p. 29). The same estimates are employed in regard to the developing countries' exports to the developed countries, which account for about three- fourths of the total. We have applied the export growth rates assumed in the Wlorld Development Report for 1975-85 to project the future expor- ts of the developing countries to the developed countries in in- '4 The calculations reported in this section were performed by joung-Yong Lee. The Changing International Division of Labor in Manufactured Goods 275 dividual industry groups for the 1976-86 period. The resulting estimates, reported in Table 7, entail an average growth rate for manufactured exports of 12.6 percent, slightly higher than the 12.2 percent shown in the Report (p. 29). The difference is largely due to the fact that nonelectrical and electrical machinery and transport equipment, which have the highest projected growth rate, had a larger share in the developing countries exports of manufactured goods in 1976 than in 1975. Our estimates for textiles and clothing represent a departure from the projections of the World Development Report that cali for deyeloping country exports to increase at an arnnual average rate of 4.5 percent in the first case, and 5.5 percent in the second, be- tween 1976 and 1986 (p. 28). The latter figures appear overly low. For one thing, they assume no change in the market share of the developing countries in the domestic consumption of textiles and clothing in the developed countries, although the developing coun- tries may further increase their share at the expense of developed country exporters such as Japan. For another thing, the estimates do not take account of the continuous upgrading of the exports of textiles and clothing from the developing countries that has raised the unit value of these exports at a rate much exceeding average prices in world trade (Keesing, 1978). More generally, the question needs to be answered if the 1975-85 export growth rates projected in the World Development Report could be applied to the 1976-86 period, given that the manufactured exports of the developing countries to the developed countries rose by 29 percent in volume terms in 1976 (cf. p. 4 above). The results for 1977, an increase of 9 percent in volume terms, would point to the need for a downward adjustment. How- ever, this decline has proved to be temporary and GATT reports that ''in volu.me terms... exports from the oil-importing deve- loping countries, which had only a small increase in 1977, appear to h,ave increased at a much faster pace, mainly on account of manufactures" in 1978 (GATT/1231, February 9, 1978, p. 4). In fact, among the major developing country exporters of manufactur- ed goods, the volume of total exports rose by 14 percent in Korea and by 24 percent in Taiwan in 1978 while manufactured exports, including processed food, increased by 28 percent in Brazil.15 The 15 International Financial Siatistics, May 1979 and Conjuntuira Económica, April 1979. 276 Banca Nazionale del Lavoro TABLE 7 PROJECTED EXPANSION OF EXPORTS AND IMPORTS IN MANUFACTURED GOODS BETWEEN THE OECD AND DEVELOPING COUNTRIES (S million; percent) Increment Increment Annual Projected in Annual Proiected in Exports Gro.th Exports, Expoats, mports Growrh Impots, Impoms, 1976 Rate 1986 1976-86 1976 Rate 1986 1976 1. Textíle Mill Pro- ducts ........... 4846 3.0 6512 1666 5874 6.2 10719 4845 2. Apparel & Other Textile Products. . 1016 3.0 1365 349 5060 8.3 11231 6171 3. Lumber & Wood Products ......... 414 3.0 556 142 1025 10,0 2659 1634 4. Furniture & Fix- tures ............ 601 3.0 808 207 324 10.0 839 515 5. Paper & Allied Products ......... 2819 5.7 4907 2088 228 10.0 592 364 6. Princing & Pu- blishing ....... 937 5.7 1631 694 110 10.0 285 . 175 7. Chemical & Allied Products .........15316 5.7 26662 11346 2058 13.0 6986 4928 .8. Petroleum Pro- ducts ......... 107 3.0 143 36 4 10.0 9 5 9. Rubber & Plastic Products ......... 1407 3.0 1890 483 1458 10.0 3782 2324 10. Leather & Leather Products ......... 357 3 9 479 122 1171 10.0 3036 1865 11. Stone, Clay & Glass Products .. . 2112 3.0 2839 727 344 10.0 892 548 12. Primary Metal & Allied Products 14289 3.0 19203 4914 2076 10.7 5736 3660 13. Fabricated Metal Products ......... 4318 3.0 5804 1486 320 10.7 884 564 14. Nonelectrical Ma- chinery ..........26731 7.0 52583 25852 916 20.3 5812 4896 15. Electrical Equip- ment & Supplies 16971 7.0 33384 16413 4970 20.3 31550 26580 1 6. Transportation Equipment 30652 5.7 53360 22708 725 20.3 4601 3876 17. Instruments & Re- lated Products . .. 3281 7.0 6454 3173 768 10.0 1991 1223 18. Misc. Manufactu- red Products .... 2036 5.7 3545 1509 ' 2454 10.0 6365 3911 Total ........ ..128210 5.7 222125 93915 29881 12.6 97968 68087 The Changing International Division of Labor in Manufactured Goods 277 increases would be larger if we considered manufactured goods alone. IG Furthermore, the Korean 15-year plan calls for manufactured exports to rise at an average annual rate of 14.9 percent between 1977 and 1991 (1978, p. 53); the corresponding projections of Taiwan's 6-year plan are 15.8 percent for the period 1978-81 (1978, p. 12); and unofficial projections for Brazil cali for increases of slightly over 10 percent for total exports, with substantially higher growth rates for manufactured exports. Note finally that the es- timates of this study exclude, and those of the World Development Report include, the countries of Southern Europe that have exper- ienced considerably lower growth rates of manufactured exports than all other developing countries, taken as a group. In estimating the future exports of manufactured goods from the developed countries to the developing countries, we have app- lied the GDP growth rate of 5.7 percent projected in the World Development Report for the developing countries in the 1975-85 period (1978, p. 32) to the 1976-86 period.,' Next, we have as- sumed varying income elasticities of demand in demand in the developing countries for different groups of manu1actures, con- strained to an average of 1.0 for all manufactured goods combined. The latter figure has been derived from the overall average elast- icity of .88 assumed in the Report (PP. 30, 32), with adjustment made for past relationships between the relevant elasticities. Thus, in the 1965-76 period, the ratio of the elasticity of import demand for manufactured imports to that for all imports by the developing countries was 1.15. In conformity with the experience of this peri- od, it has further been assumed that share of the developed coun- tries in the manufactured imports of the developing countries would remain unchanged during the period of projection. An income elasticity of import demand of approximately 1.2 16 Manufactured goods account for 85 percent of total exports in Korea and 88 percent in Taiwan while processed food rcpresents 38 percent of Braziliaii exports of manufactured goods, inclusive of processed food. At thc same time, historically, manufactured exports have been rising more rapidly than the exports of unprocessed and processed prímary commodities. 11 Again. the definition of developing countries used in this study diffcrs from that employed in the Reporr. as wc exciude the tountries of Soulhem Europe; in turn, we include Kuwalt, L¡bya, and Saudi Arabia in the group. Thcse adjustments are unlikely to modify the results by more chan one or two tenths of a percentage point, however. 278 Banca Nazionale del Lavoro has been assumed for nonelectrical and electrical machinery and for instruments and related products that are required by the invest- ment effort developing countries would have to make in order to reach high rates of economic growth, In turn, an income elasticity of import demand of 1.0 has been assurned for paper and paper products, printing and publishing, chemicais, and transport equip- ment, where a high income elasticity of domestic demand is coun- ter-balanced by ímport substitution. Finally, an elasticity of import demand of approximately 0.5 has been assumned for all other in- dustry groups, where income elasticities of domestic demand are relatively low and the import substitution effort tends to be con- siderable. These estimates would give rise to an average rate of growth of 5.7 percent for the exports of manufactured goods from the de- veloped countries to the developing countries between 1976 and 1978 as compared to the projected growth rate of 12.6 percent for their imports. Expressed in 1976 prices, there would be an absolute increment of $94 billion in the exports, and $68 billion in the imports, of manufactured goods by the OECD countries in their trade with the developing count.ries during the 1976-86 period (Table 7). The ratio of exports to imports would correspondingly decline from 4.3 in 1976 to 2.3 in 1986, with the incremenltal ratio being 1.4. The Employment Effects of Projected Trade Flows The employment implications of projected trade in manu- factured goods between developed and developing countries will be considered in the following for the OECD taken as a whole. The estimates have been made on the assumption that labor-input coef- ficients would remain unchanged during the 1976-86 period. Cor- respondingly, changes in employment resulting from trade flows have becn overestimated by neglecting future increases in labor productivity and the substitution of capital for labor in response to increases in real wages. As shown in Table 8A, the projected expansion of trade in manufactured goods between the developed and the developing countries would have practically no net effect on employmnent in the former group of countries. The increase of employment in The Changing International Division of Labor in Manufactured Goods 279 exports is estimated at 1747 thousand and the decline of employ- ment in imports at 1736 thousand. The average labor-input coefficients for increases in exports and in imports are estimated at 18.6 and 25.5, respectively, with their ratio being .73. It is thus apparent that the projections would entail a slight reduction in the labor-intensity of imports from the de- veioping countries that would result from changes in thc com- modity composition of this trade. Similar considerations apply to the capital coefficients shown in Table 8C. Nevertheless, substantial differences in labor-input cocffic eits and in capital intensity remain as between projected exports and imports in manufactured trade between developed and developing countries. Also, there are cotisiderable differences in projected chan- ges in employment among occupational categories, involving a shift from low-skill to high-skiil employment. The major loser is the unskilled and semi-skilled production workers category, with a net decline of 197 thousand jobs. These results are not surprising, given the occupational distribution of labor in production for exports and for imports in the developed countries. Prima facie, it is surprising, however, that the largest projected gain among the skilled and technical labor categories is slhown for foremen and skilled workers, with an incremental export-import ratio of 1.54, rather than for professional and technical workers, where the incremental ratio is 1.18. This result is explaialed by tihe fact that in electrical machinery and equipment, the industry groLup with the highest professional and technical labor coefficients, irn- ports from developing counttries are projected to rise at a rapid rate, with the absolute increment in imports exceeding the increment ¡i the exports of these commodities between 1976 and 1986. At the same time, projected increases of imports of electrical machinery and equipment from th,e developing countries are likelv to involve the increased importation of parts, components, and accessories, coupled with the eventual dominance of these countries in the world export of radios, television sets, automotive c lcctrical equipment, and simple electronics. The products iti quCstio0n havc lower professional and technical labor requirements thani electrical machinery and equipment in general. This is not shown by tlbc results, however, because the occupational statistics werc not ava¡- lable in sufficient detail, so that a distinction between the two groups of electrical machinery and equipment could not be made. 280 Banca Nazionale del Lavoro TABLE 8 EMPLOYMENT IMPLICATIONS AND CAPITAL COEFFICIENTS FOR INCREASES IN MANUFACTURED TRADE BETWEEN DEVELOPED AND DEVELOPING COUNTRIES, 1976-86 Exporrs Imports Balanice Ratios A. Employment Effects by Occupational Category (thousands of jobs) 1. Professional and Technical Workers ............ 215.1 182.2 33.0 1.18 2. Managers and Administrators ................. 101.5 94.0 7.5 1.08 3. Sales Workers .............................. 27.5 28.3 -0,8 0.97 4. Clerical Workets ...........................226.1 206.5 19.6 1.09 5. Foremen and Skilled Workers (Craftsmen) , ..... 402.3 261.9 140,4 1,54 6. Production Workers, Unskilled and Semi-Skilled (Operators) ................................ 677.5 874.4 -196.9 0.77 7. Workers in Construction, Transportation, Material Handling, etc .................... .......... 35.8 31.4 4.4 1,14 8. Laborers ..................................61.6 58.0 3.6 1.06 Total ................................. 1,746.8 1,736.0 10.8 1,01 of which sum of positive balance 208,5 negative balance 197.7 B. Employment Effects by Industrial Groups (thousands of jobs) 1. Textile Mill Products ......................... 47.8 158.6 -110.8 0.30 2. Apparel and Other Textile Products .......... 14.0 266.1 -252.1 0.05 3. Lumber and Wood Products ... ............... , .3.6 41.9 - 38.3 0.09 4. Furniture and Fixtures ..................... . 6.8 16.8 - 10.0 0.40 5. Paper znd Allied Products .................. , . 24.6 4.5 20.1 5.47 6. Printing and Publishing .................... , .18.5 4.3 14.2 4.30 7. Chemicals and Allíed Products ............... 121,1 56.0 65.1 2.16 8. Petroleum Products . ...................... 0.4 0.1 0.3 4.00 9. Rubber and Plastic Products ................. 11.0 100.6 - 89.6 0,11 10. Leather and Leather Products ............... ....3.4 57.7 - 54.3 0.06 11. Stone, Clay and Glass Products .... ........ .... 17.8 17.4 0.4 1.02 12. Primary Metal and Allied Products ............... .66.8 48.9 17.9 1.37 13. Fabricated Metal Products ...................6..31. 12.9 18.7 2.45 14. Nonelectrical Machinery .................... 510.1 100.0 410.1 5.10 15. Electrical Equipment and Supplies . ................ 403.6 656.1 -252.5 0.62 16. Transportation Equipment .................. 360.1 73.2 286.9 4.92 17. lnstruments and Related Products .... ......... 70.3 28.4 41.9 2.48 18, Miscellancous Manufactured Products ......... 335.4 92.5 - 57.1 0.38 Total .................................1,746.8 1,736.0 10.8 1.01 of which positive balance 875.6 negative balance .......... 864.7 C. Capital Coefficients (thousand dollars per job) Stock Measures Physical Capital .............................. 15.2 10.6 - 1.43 Human Capital ............................... 30.2 25.3 - 1.19 Together ................. ................ 453 35.9 - 1.26 FIow Measures Physical Capital ...............................8.7 6.3 - 1.38 Human Capital ............................... 9.6 8.4 - 1.14 Together .................................. 18.4 14.7 - 1.25 The Changing International Division of Labor in Manufactured Goods 281 It follows that the results shown in Table 8A understate the gain in professional and technical employment in the developed countries that would result from increased trade in manufactured goods with the developing countries. The same conclusion is likely to apply to other skilled categories, all of which show a net em- ployment gain. Thus, the shift from low-skill to high-skill occu- pations in the developed countries, resulting from their increased trade in manufactured goods with the developing countries, would be greater than estimated here. Among the 18 industry groups, total employment in the de- veloped countries would rise in paper and allied products, printing and publishing, chemicals and allied products, primary metals and fabricated metal products, with the largest gains experienced in nonelectrical machinery, transport equipment, and instruments and related products (Table 8B). In turn, there would be employment losses in textiles, clothing, lumber and wood products, furniture, rubber and plastic products, leather and leather products, electrical machinery and equipment, and miscellaneous manufactures, wher- eas practically no change is shown for petroleum and coal products and stone, clay, and glass products. The implicatrons of these results will be examined in the concluding section of the paper. Conclusion and Poiicy Implications We have seen that the export surplus of the developed in- dustrial countries in manufactured trade with the developing coun- tries increased from $36.3 billion in 1973 to $96.8 billion in 1977. And while the bulk of this increase occurred in trade with the oil-exporting developing countries, the export surpius in manu- factured trade with the oil-Importing developing countries also rose from $23.9 billion in 1973 to $42.2 billion in 1977. Alt major industrial countries and country groups experienced an increase in their trade surplus in manufactured goods with the oil-importing developing countries, with the largest increases shown for Japan and the smallest for the United States. These diffrcices have been explained by reference to differences in the initial sí- tuation in 1973 and in the trade policies followed. We have further considered the employment implications for the developed countries of their manufactured trade with the de- 282 Banca Nazionale del Lavoro veloping counttries. The employment effects of a balanced expan- sion of trade have been indicated by the use of labor-input coef- ficients that are, on the average, one-half larger for the imports than for the exports of manufactured goods in the developed countries. In turn, due to differences in trade volume, the ratio of jobs gained chrough exports to jobs lost through imports in the OECD countries would be 2.8 in the event of p,oportional increase in their man- ufactured exports and imports in trade with ilhe developing coun- tries, taking 1976 trade flows as a basis. Neither a balanced increase nor a proportionate increase in manufactured trade between developed and developing countries is a likely occurrence, however. We have, therefore, estimated the employment implications of the expected expansion of this trade in the period 1976-86, by taking the projections of the W'orld De- velopment Reporl as a starting point. The resulrs s}how no net employment effects for the OECD countries as well as a continuing shift of labor from low-skill to high-skill occupations in these coun- tries. Particular interest attaches to the effects of trade in man- ufactured goods on the skill-composition of labor, since one should look to general economic policies rather than tradce policies to ensure a satisfactory level of overall employrnent. The upgrading of the labor force, as well as the exchanige of physical and human capital for unskilled labor, would have fa- vorable effects on resource allocation and on economic growth in the developed countries.1Y At the same time, welfare gains are not limited ro the exchange of productive factors through trade. Thuls, additional gains would be forthcoming as a result of increased competition that provides inducements to tcchnological change. The shift from low-skill to high-skill occupations would be n.ssociated with a shift of labor from unskilled-labor intensive to skill-and physical capital-intensive industries. The sum of iob losses in the former group of industries would not, however, reach one million, accounting for a small proportion of employmnent in these industries. In the case of textiles, clothing, leather and leather products, the 'trimated job loss of 416 thousand compares to a total employrment of 9213 thousand in this sector in the OECD couriiries. taken together, 1976. For the nisccll:Luniit manu- 1< Ncverí hclss. in this studs tc rate of GiNP gro\s,lw lihas bccn tak-e as exogecnous and CsLimiltCs of the cffcts of trade otn conomci growth have not bcen made. The Changing International Division of Labor in Manufactured Goods 283 facturing product group, the comparable figures are 57 thousand and 1421 thousand (Table 8B and OECD, Labor Force Statíriics, 1978). 4-ote further that the ten-year time period considered in this study may permit adjustments to take place ámong industries and allow sufficient time for the upgrading of labor. In fact, one mna) envisage larger increases in the imports of manufactured goods by the developed countrics from the developing countries than pro- jected in the World Development Report without appreciable ad- verse effects on employment in import-competing industries. For one thing, the Report foresees a decline in the incremental share of manufactured goods imported from' the developing countries in developed couintry markets from 7.1 perce.nt in 1970-75 to 5.4 percent in 1975-85 (1978, p. 28), despite the large increase in this share that occurred between 1975 and 1976. For another thing, there are possibilities for additional increases in the share of exports to the developing countries in the manufactured OL1tpUt of the developed countries, which is projected to rise only slightly fromn the 1975 level in the Report.'' The above considerations indicate the advantages of liberal trade policies for the developed countries. This conclusion is stren- gthened if we consider that the import-competing industries of tWe developed countries may not experience job losses in absolute terms inasmuch as increased demand associated with economic growNth leads to higher domestic consumption and production. Furthermo- re, increased imports of labor-intensive products rnay be a sub- stitute for the immigration of unskilled and semi-skilled labor, whether legal or illegal.2'u By disregarding labor migration, the above calculations overestimate job losses to the nationais of the developed countries in import-competing industries. Reference should further be made to the incre;asing importance of developing country markets for the developed industrial coun- tries. Excluding U.S.-Canada trade that is in large part intra-firm trade and one-third of which does not involve the payment of duties in the framework of rhe automortive agreement, as well as trade in manufactured goods within the Europeao industrial free- '> Results based on trade data and on information provided n the 11 orIt! Development Reporí (p. 28). 211 In the case of Germany, one-fifch of the projected job loss would be collm- pensated by a decline in irnmigration (H¡1MI.NZ and S(& ;.Niz, 1979, p. 50). 284 Banca Nazionale del Lavoro trade area, the developing countries offer larger markets to the exports of the developed industrial countries than does trade among groups of these countries. In fact, the ratio of the sales of man- ufactured goods by the developed industrial countries in developing country markets to their exports to other developed industrial coun- tries increased from 1.03 in 1973 to 1.55 in 1977. And while increases were much larger than the average in manufactured expor- ts to the oil-exporting developing countries, with the ratio of these exports to trade among groups of developed industrial countries rising from .25 to .66, increases in this ratio are shown with respect to exports to the oil-importing developing countries as well (from .78 in 1973 to .89 in 1979). Concentrating our attention to the oil-importing developing countries, it may be suggested that increases in the manufactured exports of these countries have multiplicative effects on their na- tional incomes and, by increasing their creditworthiness for bor- rowing in international markets, on their imports. lt follows that, apart from adverse effects on the economic growth of the deve- loping countries, protectionist actions taken in the developed coun- tries would adversely affect economic growth in the developed coun- tries therrselves by foregoing employment gains in high-skill oc- cupations and in technologically advanced industries that can be obtained through increased trade in manufactured goods. Washington, D.C. BEI A BALASSA REFERENCES BALASSA, BELA, 'A 'Stages Approach' to Comparative Advantage.," in Economic Growih and Resoatrces, Proceedings of the 5th World Congrcss of chc International Economic Association held in Tokyo in September 1977, Lon- don, Macmillan, 1979. BALDWIN, ROBERT E., "Trade and Errployment Effeccts in che United Statcs Multi- lateral Tariff Reductions,' American Economic *Review, Papers and Pro- ceedings, May 1976, pp. 142-48. BALDWIN, ROBERr E. and WAYNE E. LEwIS, "U.S. Tariff Effects on Trade and Em- ployment in Detaíled SIC Industries,' The Impact oj Internatlonal Trade and Investment iín EJmployment, Proceedings of a Conference on the De- partment of Labor Rcsearch Results, held in December 2 and 3, 1970, Washington, D.C., U.S. Department of Labor. 1978. The Changing International Division of Labor in Manufactured Goods 285 BIRNBERG, THOMAS, "Economic Effects of Changes in Trade Relations between De- veloped and Less Developed Countries," NIEO Monograph Series, Overseas Development Council, Summer, 1978. CUNE, WILLIAM, et al, Trade Negotiations in the Tokyo Round: A Quantitative Assessment, Washington, D.C., The Brookings Institution, 1978. COUNCIL FOR ECONOMIC PLANNING and DEVELOPMENT, Revised Plan Jor the Second Three-Year Period (1978-1981) of the Six-Year Plan Jor the Economic Development of Taiwan, Republic of China, Taipei, 1978. DE GRAUWE P., W. KENNES, T. PEETERS, and R. VAN STRAELElN, -Trade Liberalization with the less Developed Countries: A Case Study of Belgium', Buí/etin de l'IRES, No. 44, 1977, pp. 1-16. FELS, GERHARD, "The Choice of Industry Mix in the Division of Labor between Developed and Developing Countries," Weltwirtsch4ftliches Archiv, 1972 (1), pp. 71-121. GRINOLS, ERROL and ERIK THORBECKE, -The Effects of Trade between the U.S. and Developing Countries in U.S. Employment," Working Paper No. 171, De- partment of Economics, Cornell University, 1978. HIEMENZ, U. and K.W. SCHATZ, Trade in Place of Migration, Geneva, International Labour Office, 1979. HsIEH, C., "Measuring the Effects of Trade Expansion on Employment: A Rcview of Some Research," International Labor Review, January 1973, pp. 1-79. KEESING, DONALD B., 'Different Countries' Labor Skill Coefficients and the Skill Intensity of International Trade Flows," Journal of International Economics, November 1971, pp. 443-52. KEESING, DONALD B., 'Labor Skills and trSe Stnucture of Trade in Manufactures," in The Open Economy (Peter B. Kenen and Robert Laurence, cd.), New York, Columbia University Press, 1968, Pp. 3-18. KEESING, DONALD B., "Labor SkilJs and International Trade: Evaluating Many Trade Flows with a Single Measuring Device," Review of Economics and Statistics, August 1965, pp. 287-94. KEESING, DONALD B., "World Trade and Output of Manufactures: Structural Trends and Developing Countries' Exports," Washington, D.C., World Bank, Fe- bruary 1978, (mimeo). KENEN, PETER B., "Nature, Capital, and Trade," Journal of Politica¡ Economy, October 1965. KOREA DEVELOPMENT INSTITUTE, Long-term Prospects for Economic and Social De- velopment 1979-91, Seoul, 1978. KRUEGER, ANNE O., "Impact of LDC Exports on Employrnent in American In- dustry," Paper presented to the Annual Conference of the International Economic Study Group, White House, Isle of Thorns, Sussex, England, 1978 (mimeo). LYDALL, H. F., Trade and Employment, Geneva, International Labor Office, 1975. SCHUMACHER, D., "Verstarkter Handel mit der Dritten Welt: Eher Unisetzung als Freisetzung deutscher Arbeitskrafte," DIW Wochenbericht 5/77, February 3, 1977, pp. 35-40. SCHUMACHER, D., "Beschaftigungswirkungen von Importen aus Entwicklungslan- dern nicht drarnatisieren," DIW Wochenbericht 1/78, January, 1978, Pp. 6-11. THE WORLD BANK Headquarters: 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. European Office: 66, avenue d'Iéna 75116 Paris, France Tokyo Office: Kokusai Building, 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan The full range of World Bank publications, both free and for sale, is described in the World Bank Catalog of Publications, and of the continuing research program of the World Bank, in World Bank Research Program: Abstracts of Current Studies. The most recent edition of each is available without charge from: PUBLICATIONS UNIT THE WORLD BANK 1818 H STREET, N.W. WASHINGTON, D.C. 20433 U.S.A.